Assessment of The European Community System of Pharmacovigilance

Assessment of the European
Community System of
Pharmacovigilance
Final Report – Final version

25 January 2006
European Commission
Enterprise and Industry Directorate-General, Unit F2, Pharmaceuticals
Reference: Service Contract No: FIF.20040739
Submitted by the
Fraunhofer Institute for Systems and Innovation Research, Karlsruhe,
Germany
in collaboration with the
Coordination Centre for Clinical Studies at the University Hospital of
Tübingen, Germany
Contact:
Dr. Thomas Reiss
Fraunhofer Institute for Systems and Innovation Research (ISI)
Breslauer Str. 48, 76139 Karlsruhe, Germany
Tel.: +49-721-6809-160, Fax: +49-721-6809-315
E-mail: [email protected]
2 Assessment of the European Community System of Pharmacovigilance
Final Report November 2005
Table of Content
Executive summary.......................................................................................... 5
Overview and aim of the study ..................................................................... 5
General aspects ........................................................................................... 6
Data collection.............................................................................................. 8
Data management...................................................................................... 10
Signal detection.......................................................................................... 10
Safety issue assessment ........................................................................... 10
Decision-making......................................................................................... 11
Communication and action......................................................................... 12
Core recommendations.............................................................................. 13
Abbreviations ................................................................................................. 14
List of figures.................................................................................................. 16
List of tables ................................................................................................... 19
1 Overview ................................................................................................... 22
2 Methods .................................................................................................... 23
2.1 Organisation and implementation of the project ........................ 23
2.2 Tasks and methods ................................................................... 24
2.3 Collection and analysis of data.................................................. 28
3 System analysis and description of the status quo.............................. 35

3.1 Description of the regulatory framework of
pharmacovigilance in Europe .................................................... 35
3.2 Implementation of the European regulatory
framework into practice ............................................................. 66
3.3 Systems of pharmacovigilance in other countries ..................... 67
3.4 Other framework conditions ...................................................... 78

3.5 Resources for pharmacovigilance ............................................. 80
3.6 General quality management .................................................... 90
3.7 Phases of pharmacovigilance ................................................... 92
3.8 Outcomes of regulatory action ................................................ 120
4 Goals in respect of effectiveness and efficiency ................................ 123
5 Critical success factors......................................................................... 124
6 Performance indicators ......................................................................... 127
7 Case studies ........................................................................................... 131

7.1 Statins case............................................................................. 131
7.2 SSRI case ............................................................................... 132
8 Best practice........................................................................................... 134

8.1 Indicator-based identification of best-practice ......................... 134
8.2 Results from the interviews ..................................................... 146
9 Discussion of strengths and weaknesses of the European

system of pharmacovigilance ............................................................... 150
9.1 General factors........................................................................ 150
9.2 Data collection......................................................................... 155
9.3 Data management................................................................... 157
9.4 Signal detection....................................................................... 158
9.5 Safety issue assessment......................................................... 159
9.6 Decision-making...................................................................... 160
9.7 Communication and action to protect public health................. 161
10 Recommendations ................................................................................. 163

10.1 General factors........................................................................ 163
10.2 Data collection......................................................................... 166
10.3 Data management................................................................... 167

10.4 Signal detection....................................................................... 168
10.5 Safety issue assessment......................................................... 168
10.6 Decision-making...................................................................... 169
10.7 Communication and action to protect public health................. 169
10.8 Core recommendations ........................................................... 170
Annex 1: Literature....................................................................................... 172
Annex 2: Questionnaire for Agency survey ............................................... 176
Annex 3: Results of Delphi survey on critical success factors................ 186
Annex 4: Results of Delphi survey on performance indicators ............... 193

Executive summary
Overview and aim of the study

Pharmacovigilance (PhV) is a key public health function. It is defined as the proc-
esses and science of monitoring the safety of medicines and taking action to re-
duce risk and increase benefit. It includes six phases:
1. Data collection
2. Data management
3. Signal detection
4. Safety issue assessment
5. Decision-making
6. Communication and action
The general aim of the present project was to analyse how the European central
and EU Member States' medicines agencies collaborate with each other, the mar-
keting authorisation holders (MAHs) and other stakeholders, in surveilling the ad-
verse effects of pharmaceutical products, and to put forward recommendations to
make the system more robust.
The work was based on a systemic perception of pharmacovigilance and combined
the analysis of different aspects of the system: processes, stakeholders, resource
availability and functional capability, gaps, strengths and weaknesses, as well as
best practice. It was carried out by the Fraunhofer Institute for Systems and Inno-
vation Research, Karlsruhe, Germany, and the Coordination Centre for Clinical
Studies at the University Hospital of Tübingen, Germany, together with Prof.
H.G.M. Leufkens from the Utrecht Institute for Pharmaceutical Sciences, University
of Utrecht, the Netherlands, and Prof. U.M. Gassner, Department of Public Law
and Research Centre for Law of Medicinal Products at the University of Augsburg,
Germany.
The national medicines agencies, as well as the European Medicines Agency as
the competent authorities were included in the research in several ways: interviews
and a written survey were carried out with representatives, mainly the heads, of the
agencies' pharmacovigilance units. Moreover, the Heads of Medicines Agencies
Working Group for the European Risk Management Strategy and other experts
were systematically involved in the design of the study and the discussion of pre-
liminary findings and conclusions. The final results and recommendations, how-
ever, are the sole responsibility of the project team.
The main results of the study for general aspects as well as for the phases of
pharmacovigilance are briefly reviewed in the following paragraphs, and the core
recommendations are presented.
General aspects
The legal framework harmonises regulation, pharmacovigilance practice, product
information, communication and action across the Member States. International co-
ordination lends more power to regular action, this is especially true for the system
for Centrally Authorised Medicinal Products (CAPs).
However, the legal system is also complicated because of the many responsible
authorities involved; different procedures and responsibilities for products under the
centralised and the non-centralised authorisation procedure. The system is very
difficult to oversee despite the existence of detailed guidances.
Different implementation of the framework is caused by e.g. diverging health sys-
tems in the MS and different opinions which tasks should fall under the responsibil-
ity of the national authorities. The new Member States are not yet totally integrated
and existing instruments are not fully applied.
At the moment, the emphasis strongly lies on the collection and analysis of sponta-
neous reports. This will remain important despite the fact that the recent safety
crises have shown that other information and especially independent safety studies
may be even more important to identify safety issues. The new regulatory system
in place from November 2005 on will allow Pharmacovigilance Planning including a
more proactive approach to pharmacovigilance by agencies and MAHs, and should
be rigorously applied.
The analyses have shown that staff numbers and technical resources vary tre-
mendously across agencies.
Table 0.1. Total national staff for PhV per capita
PhV staff NCA1 PhV staff NCA+RC2

[FTE per million capita] [FTE per million capita]
Minimum 0.2 0.2
Median 0.772 1.183
Maximum 4.6 4.6
Staff for pharmacovigilance, scientific and administrative.
Source: Fraunhofer ISI 2005
In some agencies the number of staff seems to be less than the minimum required
to complete the necessary tasks. Sufficient resources are needed in the MS to
reach comparable staff numbers relative to their population sizes. The median of
agencies might be used as a minimum value for all agencies. The completion of all
urgent tasks at every point in time must be guaranteed.
1 National Competent Authority

2 Regional Centre for Pharmacovigilance
The system draws strongly on the combination of expertise: expertise, assess-

ments of safety issues and other documents developed at EU level can be used by
the other agencies. However, the capability to assess safety issues does not exist
in all agencies. As a result of the complex system and lack of experienced staff,
some of the agencies need more support to be able to comply with the require-
ments. In some countries, it is difficult to find the necessary external experts espe-
cially for the assessment of safety issues, which also hampers their full contribution
to the system.
With respect to the collaboration between agencies, the European system offers
good and in general well-functioning structures, including the central role of the
European Medicines Agency (EMEA).
Figure 0.1. Cooperation between national agencies and EMEA
70
60
50
Responses in %
40
30
20
10
0
Very good 2 3 4 Very bad
On the other hand, being dependent on other agencies' work is sometimes a prob-
lem as long as the agencies' work is of different quality. Opinions differ as to what
amount of work should be done at the national level, leading to different assess-
ments of necessary and unnecessary duplication of work, which some of the agen-
cies consider to be relatively high. Communication between MS agencies and
EMEA is sometimes considered problematical.
The collaboration with health-care professionals, especially the physicians who
directly impact on the prevalence of adverse drug reactions through their prescrip-
tion behaviour, could be improved. Regional centres for pharmacovigilance are a
promising approach to effectively communicate with health-care professionals.
The compliance of Marketing Authorisation Holders with the safety regulations
for their products should be checked more rigorously.
Figure 0.2. Compliance of MAHs in analysis of signals

50
45
40
35
% of agencies
30
25
20
15
10
0
Systematic quality management is not implemented in most PhV departments.

The regulatory system does not provide clear goals or provisions in this respect. If
implemented, nearly all agencies state that their audit procedures do not ade-
quately ensure the quality of their work. In particular, only few agencies follow up
the impact of communications on a routine basis. The continuous management and
improvement of the agencies' quality of work is a major area for future action.
Data collection
The European system combines the Individual Case Safety Reports from a large
population in order to increase the statistical power with which signals can be de-
tected; small countries3 with few reports in particular benefit from this.
The agencies are not very well prepared for crises by routine data (spontane-
ous reports coming from health-care professionals or marketing authorisation hold-
ers and Periodic Safety Update Reports (PSURs) from the marketing authorisation
holders), their usefulness is restricted.
3 The terms "small" or "large" for countries refer to the size of their population.
Figure 0.3. Preparation for last crisis by routine data

45
40
35
30
% of agencies
25
20
15
10
5
0
Besides these "routine data", data especially on drug consumption, but also regis-
tries and other data combining drug exposure and outcomes, including ad-
verse drug reactions, are highly relevant. Such registries exist in most of the
countries. However, most agencies only have access to these data in exceptional
cases, and they are quite infrequently used. This situation has to be improved.
Table 0.2. Existence and use of data on the consumption of medicines
Exist Use
N of % in
% %
agencies except.
never routinely
cases
Sales data 24 0% 33.3% 66.7%
Prescription non-hospital 19 25.0% 40.0% 35.0%
Prescription hospital 14 47.1% 23.5% 29.4%
Safety studies and other data that can supplement the routine data played a deci-
sive role in the last safety crises. However, only very few prospective safety studies
were prepared in the last years, and some of them were not performed independ-
ently of the producer of the drug studied. The funding of necessary studies is often
not guaranteed. This open question is tackled by the new regulatory system which
allows more pro-active data collection; its implementation is urgently required.
Research into the safety of drugs for children is disparately lacking, as is a data-
base on products already on the market.
Data management
The system allows for a systematic sharing of work between the involved stake-
holders (MAHs vs. agencies, as well as among different agencies). The databases
that are used in the national agencies to manage case safety reports and other
safety data vary greatly and are not all sufficiently specific to handle the necessary
data.
Some duplication of work related to the handling of the same data exists at dif-
ferent agencies, especially at the EMEA, on the one side, and national agencies on
the other. However, the issue of duplication of work (what is necessary, what is
unnecessary duplication?) is assessed heterogeneously by the agencies. The co-
ordination with other international partners could be improved.
Signal detection
EudraVigilance and the related procedures form the basis for the effective system-
atic pooling of and signal detection from spontaneous reports. The success of the
combination of expertise and resources for signal detection depends on the full
implementation of the provisions; with regard to the dependence on national
resources and priorities, which at the moment cannot be taken as guaranteed and
therefore needs continued supervision and support. This also holds true for the
statistical tools for signal detection, as the tools for small numbers of cases in
particular are still insufficient; improved techniques will have to be developed. As
for data management, it does not seem that the best use is being made of work
that is performed by the European system and by other international partners, re-
spectively.
As hardly any controls are in place, it remains unclear whether the Marketing Au-
thorisation Holders fulfil their role of first-line signal detection.
Safety issue assessment

The system allows sharing work and using assessment reports from other coun-
tries; generally, it comes to comparably good or better conclusions that other inter-
national systems.
The share of work and best use of international collaboration depends to a great
extent on the quality of the work that the single agencies can contribute. Some
agencies, however, admit that they do not yet have the ability to manage safety
issues adequately on their own.
External expertise has not always been adequately used, partially because of
difficulties in accessing external experts that some agencies experience.
Figure 0.4. Receive support from experts routinely

50
45
40
35
% of agencies
30
25
20
15
10
5
0
Very easy 2 3 4 (nearly)
impossible
Decision-making
Decision-making often takes too long, which is partially attributed to complicated
structures within the CHMP and between CHMP and the Commission, especially in
the case of referrals.
Figure 0.5. Decisions for safety issues found in adequate time
NAPs
MRPs
CAPs
0% 20% 40% 60% 80% 100%

% of responses
Always 2 3 4 Seldom
NAP: Nationally authorised product, MRP: Product authorised under Mutual Recognition Procedure;
CAP: Centrally authorised product

Communication and action

The system provides the structures to develop ?? timely and harmonised commu-
nications and actions, especially in those cases where satisfactory agreement has
been obtained between the agencies and where sufficient time is available. How-
ever, the time between the detection of a signal and action was too long in
some cases.
Regulatory transparency is very important to allow for consistent communication
and argumentation of decisions. It is important to adequately represent the Mem-
ber States' particular situation (e.g. regarding consumption patterns and epide-
miology) in CHMP opinions or Commission decisions.
The outcomes of regulatory action are only assessed in exceptional cases.
There is very little information about the reaction of prescribers to label information
and label changes. Moreover, when information is available, the results are not
very encouraging.
Figure 0.6. Influence of agencies' communications on prescription behaviour
35
30
25
% of agencies
20
15
10
0
Very good 2 3 4 Very w eak
More should be done to ensure and supervise that communications and regulatory
action result in the intended effects, especially by doing more research into the
impacts of safety communication and action on prescription behaviour, but also by
more inspections of MAHs with a pharmacovigilance focus.
Core recommendations
From the present research, we derive the following most important conclusions to
make the European System of Pharmacovigilance more robust:
• The relative contribution of the different sources of safety information
(ICSRs, PSURs, registries, consumption data, safety studies etc.) and re-
spective resources for pharmacovigilance should be reviewed. The neces-
sary statistical tools should be developed and specific requirements of small
countries should be kept in mind.
• The new legislation strengthens the potential impact of tackling safety is-
sues more pro-actively. This opportunity should be extensively used.
• The decision-making process should be reviewed; opportunities to stream-
line and fasten it should be identified.
• The impacts of communications and actions should be checked more sys-
tematically and from the lessons learned, the impact on prescription behav-
iour should be improved.
• The marketing authorisation holders are primarily responsible for the safety
of their products. More resources are necessary to check if they comply
with their legal obligations, and at the same time it should be identified how
the requirements can be made as supportive as possible (e.g. as far as
PSURs are concerned).
• General principles of quality management and continuous quality improve-
ment should be introduced, among others:
(1) setting realistic and measurable targets for key interim impacts and for
final outcomes;
(2) regularly checking if these target values have been reached;
(3) use of internal audit and peer review;
(4) identifying and deleting weaknesses (bottlenecks in procedures, un-
der-performance or under-equipment of actors, waste of resources…).
Abbreviations
ADE Adverse Drug Experience
ADR Adverse Drug Reaction
AE Adverse Event
AERS Adverse Event Reporting System
CADRIS Canadian Adverse Drug Reaction Information System
CAP Centrally Authorised Medicinal Product
CFR Code of Federal Regulations (USA)
CHMP Committee for Medicinal Products for Human Use
CIOMS Council for International Organizations of Medical Sciences
EEA European Economic Area
EMEA European Medicines Agency
ERMS European Risk Management Strategy
EU European Union
FDA Food and Drug Administration (USA)
GPMSP Good Post-Marketing Surveillance Practice
HCP Healthcare Professional
HMA Heads of Medicines Agencies
ICH International Conference on Harmonisation of Technical Requirements
ICSR Individual Case Safety Reports
MAH Marketing Authorisation Holder
MedDRA Medical Dictionary for Regulatory Affairs
MHLW Ministry of Health, Labour and Welfare (Japan)
MHPD Marketed Health Product Directorate (Canada)
MRP Mutual Recognition Procedure; Mutual Recognition authorised Product
MS EU Member State
NAP Nationally Authorised Medicinal Product
NCA National Competent Authority
NDA New Drug Application
NUIS Non Urgent Information System
PAL Pharmaceutical Affairs Law (Japan)
PASS Post-authorisation Safety Study
PhVWP Pharmacovigilance Working Party
PMDA Pharmaceutical and Medical Device Agency (Japan)
PMS Post-Marketing Surveillance
PSUR Periodic Safety Update Report
RAS Rapid Alert System
RC Regional centre for pharmacovigilance
SPC Summary of Product Characteristics
VAERS Vaccine Adverse Event Reporting System (USA)
WHO World Health Organisation
Abbreviation of country names

AT Austria
BE Belgium
CY Cyprus
CZ Czech Republic
DE-BFARM Germany
DE-PEI Germany-PEI
DK Denmark
EE Estonia
EEA-28 EMEA as responsible for 28 EEA countries
EI Ireland
ES Spain
EU-25 EU-25
FI Finland
FR France
GR Greece
HU Hungary
IC Iceland
IT Italy
LI Liechtenstein
LT Lithuania
LU Luxembourg
LV Latvia
MT Malta
NL Netherlands
NO Norway
PL Poland
PT Portugal
SE Sweden
SL Slovenia
SV Slovakia
UK United Kingdom
List of figures
Figure 0.1. Cooperation between national agencies and EMEA .....................................7
Figure 0.2. Compliance of MAHs in analysis of signals...................................................8
Figure 0.3. Preparation for last crisis by routine data ......................................................9
Figure 0.4. Receive support from experts routinely.......................................................11
Figure 0.5. Decisions for safety issues found in adequate time ....................................11
Figure 0.6. Influence of agencies' communications on prescription

behaviour.............................................................................................12
Figure 2.1. Delphi survey form for the evaluation of success factors (part) ..................33
Figure 2.2. Delphi survey form for the evaluation of performance indicators
(part) ....................................................................................................33
Figure 3.1. Authorisation Procedures ............................................................................38
Figure 3.2. Regulation (EC) No 726/2004 .....................................................................39
Figure 3.3. Directive 2001/83/EC ..................................................................................41
Figure 3.4. Related guidance documents......................................................................42
Figure 3.5. Key points guideline E2A ............................................................................44
Figure 3.6. Key points guideline E2B(M) .......................................................................44
Figure 3.7. Key points guideline E2C ............................................................................45
Figure 3.8. Key points guideline E2C Addendum..........................................................46
Figure 3.9. Key points guideline E2D ............................................................................47
Figure 3.10. Key points guideline E2E ..........................................................................48
Figure 3.11. Key requirements to the pharmacovigilance systems of MAHs ................51
Figure 3.12. Overview of processes in collection of data relevant to

medicines safety..................................................................................53
Figure 3.13. Expedited reporting obligations by MS Authorities....................................54
Figure 3.14. Expedited reporting by the EMEA .............................................................55
Figure 3.15. Types of notification of safety concerns and exchange of

information...........................................................................................56
Figure 3.16. Expedited reporting by MAHs....................................................................57
Figure 3.17. The responsibilities of the qualified person ...............................................59

Figure 3.18. Organisation of pharmacovigilance regarding centrally

authorised medicinal products.............................................................66
Figure 3.19. Organisation of pharmacovigilance regarding non-centrally

authorised medicinal products.............................................................67
Figure 3.20. Expedited reporting requirements in the USA ...........................................69
Figure 3.21. Reports qualifying for inclusion in a US periodic report.............................70
Figure 3.22. Expedited reporting requirements in Japan...............................................72
Figure 3.23. Expedited reporting requirements in Canada............................................74
Figure 3.24. Reports qualifying for inclusion in a Canadian summary report ................75
Figure 3.25. Political support for PhV ............................................................................80
Figure 3.26. Average agency budget in 2004................................................................81
Figure 3.27. Existence of regional centres ....................................................................83
Figure 3.28. Cooperation between national agencies and EMEA .................................87
Figure 3.29. Duplication of work ....................................................................................87
Figure 3.30. Assessment of IT-resources......................................................................88
Figure 3.31. Guidance in national version (part 1) ........................................................89
Figure 3.32. Guidance in national version (part 2) ........................................................89
Figure 3.33. Meeting of internal targets.........................................................................91
Figure 3.34. ICSRs received 2003 and 2004 (countries with numbers of
ICSRs ≥2000) ......................................................................................94
Figure 3.35. ICSRs received 2003 and 2004 (countries with numbers of
ICSRs <2000)......................................................................................94
Figure 3.36. Total number of ADR reports in the national databases ...........................95
Figure 3.37. Reporting rates for total populations over time, number of
ICSRs divided by population size ........................................................97
Figure 3.38. Reporting rates for total populations 2003 and 2004, number
of ICSRs divided by population size ....................................................97
Figure 3.39. Reporting rates for children over time, number of ICSRs 2004
divided by number of children..............................................................98
Figure 3.40. Reporting rates for children ≤19 years, number of ICSRs
2004 divided by number of children.....................................................99
Figure 3.41. Reporting rates for total populations; number of ICSRs 2004
divided by pharmaceutical sales........................................................101
Figure 3.42. Reporting rates for total populations, number of ICSRs 2004
divided by number of physicians .......................................................102
Figure 3.43. Time for data processing.........................................................................109
Figure 3.44. Compliance of MAHs in analysis of signals.............................................110
Figure 3.45. Assessment of statistical tools for signal detection .................................110
Figure 3.46. Preparation for last crisis by routine data ................................................111
Figure 3.47. Usefulness of routine data.......................................................................111
Figure 3.48. Receive support from experts routinely...................................................113
Figure 3.49. Adequate decisions found for safety issues ............................................114
Figure 3.50. Decisions for safety issues found in adequate time ................................115
Figure 3.51. Kinetics of total process from signal detection and reporting ..................115
Figure 3.52. Transparency of decision-making within the companies.........................116
Figure 3.53. Consistency of communications..............................................................119
Figure 3.54. Influence of agencies' communications on prescription

behaviour...........................................................................................121
Figure 5.1. Systematic of critical success factors........................................................126

List of tables
Table 0.1. Total national staff for PhV per capita ............................................................6
Table 0.2. Existence and use of data on the consumption of medicines.........................9
Table 2.1. Tasks and work steps...................................................................................25
Table 2.2. Datasets for the analyses .............................................................................28
Table 2.3. Site visits for interviews ................................................................................29
Table 2.4. Correlations of population-based reporting rates with external

criteria..................................................................................................31
Table 2.5. Correlations of sales-based reporting rates with external criteria.................32
Table 2.6. Criteria for the evaluation of indicators in the Delphi survey ........................34
Table 3.1. Number of approvals for NMEs per country in 2003 and 2004 ....................78
Table 3.2. Time acting for the Community as Rapporteur.............................................78
Table 3.3. Time acting for the Community as RMS.......................................................79
Table 3.4. Number of physicians ...................................................................................79
Table 3.5. Agency budget 2004.....................................................................................80
Table 3.6. PhV staff in national agencies ......................................................................82
Table 3.7. PhV staff in different process stages (multiple responses

possible) ..............................................................................................82
Table 3.8. Total national staff for PhV per capita ..........................................................83
Table 3.9. Tasks of regional PhV centres......................................................................84
Table 3.10. Number of ICSRs collected by regional PhV centres .................................84
Table 3.11. Contracts with other actors in PhV assessments .......................................85
Table 3.12. Competences available in countries...........................................................85
Table 3.13. Expert committee meetings per year..........................................................86
Table 3.14. Contributions to EEA PhV system ..............................................................86
Table 3.15. Submission of reports on ADRs [% of ICSRs]............................................93
Table 3.16. Suspected ADRs from reporter groups ......................................................96
Table 3.17. Reporting rates – different indices..............................................................96
Table 3.18. ICSRs by MA procedure...........................................................................100

Table 3.19. Reporting rates for total populations, number of ICSRs divided
by pharmaceutical sales in the respective country ............................100
Table 3.20. Reporting rates for total populations, number of ICSRs divided
by number of physicians....................................................................101
Table 3.21. PSURs by MA procedure .........................................................................103
Table 3.22. Existence and use of population-based health/disease

registries ............................................................................................104
Table 3.23. Existence and use of data on the consumption of medicines...................105
Table 3.24. Most important literature sources .............................................................107
Table 3.25. Information collected ................................................................................107
Table 3.26. Data quality...............................................................................................108
Table 3.27. Informational needs for signal detection...................................................112
Table 3.28. Routinely inform on safety issues.............................................................117
Table 3.29. Communication procedures......................................................................117
Table 3.30. Possible actions to prevent ADRs ............................................................118
Table 3.31. Regulatory actions....................................................................................119
Table 3.32. Impact of regulatory action audited ..........................................................120
Table 3.33. Existence of outcome studies...................................................................121
Table 3.34. Outcome studies.......................................................................................122
Table 3.35. Incidence of ADR-relevant diseases ........................................................122
Table 5.1. Draft list of critical success factors .............................................................124
Table 6.1. Most important performance indicators ......................................................127
Table 8.1: Indicators for Data Collection .....................................................................136
Table 8.2: Indicators for Data Management ................................................................138
Table 8.3: Indicators for signal detection – Availability of data sources ......................139
Table 8.4: Indicators for signal detection – Analysis tools and resources ...................140
Table 8.5: Indicators for safety issue assessment.......................................................141
Table 8.6: Indicators for decision making ....................................................................142
Table 8.7: Indicators for communication/action ...........................................................143
Table 8.8: Indicators for communication/action with HCPs .........................................144

Table 8.9: Indicators for information provided to HCPs...............................................145

1 Overview
The present report summarises the project "Assessment of the European Commu-
nity System of Pharmacovigilance" from December 2004 to September 2005. The
general aim of the project is to analyse the way in which the European central and
Member States' medicines agencies collaborate in the surveillance of adverse ef-
fects of pharmaceutical products among each other as well as with the marketing
authorisation holders and other stakeholders, and to make recommendations to
make the system more robust.
The work was based on a systemic perception of pharmacovigilance and combined
the analysis of different aspects of the system: processes, stakeholders, resource
availability and functional capability, gaps, strengths and weaknesses, as well as
best practice.
The project comprised the following 7 tasks.
Phase I:
Task 1: System analysis and description of status quo
Task 2: Definition of goals in respect of effectiveness and efficiency
Task 3: Identification of critical success factors
Phase II:
Task 4: Identification and definition of performance indicators
Task 5: Gap analysis to identify strengths and weaknesses
Phase III:
Task 6: Identification of best practice
Task 7: Recommendations
The service contract was signed on 08 December 2004, and based on an amend-
ment from 13 April 2005; its duration was not more than 10 ½ months. An exten-
sion of the duration by 60 days became necessary for organisational reasons and
was granted by the European Commission. Accordingly, the interim report was due
at 03 June 2005, the draft final report was due at 15 September 2005, and the final
report at 11 November 2005.
2 Methods
2.1 Organisation and implementation of the project

2.1.1 Partners and subcontractors
The project was carried out by the Fraunhofer Institute for Systems and Innovation
Research, Karlsruhe, Germany (Fraunhofer ISI, project leader) and the Coordina-
tion Centre for Clinical Studies at the University Hospital of Tübingen (KKS).
As subcontractors acted the Utrecht Institute for Pharmaceutical Sciences (UIPS),
Prof. Dr. H.G.M. Leufkens, head of the Department of Pharmacoepidemiology and
Pharmacotherapy, particularly to collaborate with the partner KKS in the prepara-
tion of the case studies and to give general advice, as well as Prof. Dr. U.M. Gass-
ner, Full Professor of Public Law at the University of Augsburg and director of the
research centre for law of medicinal products, to support the overview of the legal
framework.
2.1.2 Advisors
Instead of a formal steering committee for the project as suggested in the tender, it
was agreed with the Commission to ask the Members of the Heads of Medicines
Agencies ERMS Working Group to act as advisors. Additional experts from aca-
demia and industry as well as the subcontractor Prof. Leufkens joined this group.
The advisors were primarily asked to participate in an interim meeting to discuss
preliminary results and to support the development of critical success factors and
performance indicators in two Delphi surveys.
2.1.3 Agreements of confidentiality

To ensure that potentially sensitive data were kept confidential by all concerned
persons, and to allow the project to use confidential data from the agencies, agree-
ments of confidentiality were signed both between the agencies and the project
group as well as between the advisors and the project group.
2.1.4 Time schedule

The service contract was signed on 08 December 2004, and the work started di-
rectly after that. The organisation of the interviews proved to be much more time-
consuming than expected for several organisational reasons. Besides this, the en-
visaged data of the HMA ERMS survey were not available until the beginning of
June 2005.
The delay was not considered as an irreparable problem as some tasks (including
the Delphi survey on indicators and critical success factors) could be shifted after
the expert workshop without hampering the quality of the project's outcomes. None
of the tasks described in the tender were neglected or dropped.
To have the questionnaire survey quite late in the course of the project offered the
chance to include into the questionnaire survey the finalised set of the performance
indicators instead of a draft set as it was planned. The earlier plan would have re-
quired asking the agencies for data more than once and thus the workload for the
agencies could be reduced. Additionally, the questionnaire’s quality was improved
as an instrument for a potential future routine monitoring of the European pharma-
covigilance system.
To ensure the quality of the project's results it was important that the remaining
tasks could be performed with the time budgets planned originally. Therefore, the
Commission was asked for and accepted an extension of the duration of the con-
tract by four weeks and a slightly reduced time for the review of the final report.
Based on an amendment from 13 April 2005; its duration is not more than 10 ½
months. A second extension of the duration by 60 days was granted by the Euro-
pean Commission. Accordingly, the interim report was due at 03 June 2005; the
draft final report was submitted at 15 September 2005.
2.2 Tasks and methods

The work was divided in seven tasks (Table 2.1).
2.2.1 Task 1: System analysis and description of status quo

Task 1 was the first step in project Phase I (Description of the current system
based on submitted documentation, questionnaires and site visits). It consisted of
two subtasks. Based on an extensive analysis of scientific literature, previous re-
ports, and previous studies, as well as on official documents from internet and na-
tional competent authorities, the specific features of the European Community sys-
tem of pharmacovigilance that are relevant for its functioning were defined. One
main output of this exercise is a review of the regulatory situation in the EU, which
can be found in paragraph 3.1. The other output is the description of the processes
in pharmacovigilance carried out by the national agencies and the EMEA based on
the empirical data-collection with interviews and written survey.
Table 2.1. Tasks and work steps
Task/Milestone/Deliverable Detailed work steps
D14 Detailed work plan
MS1 Kick-off meeting with Commission
Task1 System analysis Analysis of scientific literature
Analysis of documents
Written survey of agencies
Personal interviews with agencies
Telephone interviews with stakeholders
MS2 Description of the relevant features of Working paper on regulation

the system
Task2 Goals See Task 1
MS3 List of effectiveness/efficiency goals Analysis of literature, Delphi process
Task3 Critical success factors Analysis of the literature
Personal interviews
Delphi process
MS4 List of critical success factors –
Task4 Performance indicators Analysis of the literature
Personal interviews
Delphi process
Case studies
MS5 List of validated and practicable –

performance indicators
Task5 Gap analysis Comparison of indicators with success factors
Interviews
MS6 List of differentiated strengths and –

weaknesses
4 MS: Milestone; D: Deliverable

Task/Milestone/Deliverable Detailed work steps
D2/MS7 Interim report and interim meeting –
Task6 Best practice Comparison of indicators with success factors
Interviews
MS8 List of best practice approaches –
Task7 Recommendations Expert workshop
Optimisation tree
MS9 Expert workshop finished/ –

list of recommendations
D3/MS10 Draft final report and final meeting –
D4/MS11 Final report –
Task 1 was completed by telephone interviews with additional stakeholders from

industry, science and independent organisations on specific topics.
2.2.2 Task 2: Definition of goals in respect of effectiveness and

efficiency
Task 2 used the same methodological approaches as for task 1 (analysis of scien-
tific literature, reports, studies, interviews with representatives of the competent
authorities, representatives of the industry). On this basis, the goals in respect of
effectiveness and efficiency of the European Community system of pharmacovigi-
lance were defined.
Initially it was intended to supplement a provisional list of goals that was based on
the literature by the respective results from the personal interviews, and to ask the
advisors in a Delphi-process to comment on the list. However, the interviews re-
vealed nearly no new aspects in this respect; most of the interviewees found the
actual scope of pharmacovigilance (with some modifications) in general sufficient,
many referred to the related WHO definition. Therefore, there was no need to
elaborate more on the aspect of additional goals for pharmacovigilance in the Del-
phi process.
2.2.3 Task 3: Identification of critical success factors

Critical success factors are those elements of the whole process that determine its
performance and can be modified to improve a system. For the 25 EU Member
States and EMEA the most critical success factors for an effective and efficient
functioning of the pharmacovigilance system (with respect to cost-effectiveness,
time-efficiency, quality and safety) were identified firstly on the basis of a systems
approach supported by data from the interviews and literature. First results were
presented in Brussels at the expert workshop on 15 June 2005. After this, the advi-
sors were asked in a Delphi-process to comment on the list of critical success fac-
tors.
2.2.4 Task 4: Definition of performance indicators

Task 4 was the first task in project Phase II (Assessment of the robustness of the
current system highlighting both strengths and weaknesses).
Within task 4 a set of performance indicators was developed based on literature
and interviews. The indicators were distinguished in input, process, and output in-
dicators. The advisors were asked to support the evaluation of the indicators in the
Delphi-process. This task was delayed because of the late completion of the inter-
views which formed an important input also for this step.
2.2.5 Task 5: Gap analysis to identify strengths and weaknesses

Task 5 identified strengths and weaknesses of the pharmacovigilance system on
basis of performance indicators, critical success factors and effective-
ness/efficiency goals.
Most important working step was the analysis of the interview and agency survey
data. They were partially combined with the data from the two ERMS surveys,
which assessed the old MS in 2002 and the new MS in 2004 (more details are
found in chapter 2.3).
2.2.6 Task 6: Identification of best practice

This was the first task of Phase III (Proposals to make the European Community
system of pharmacovigilance more robust).
Many interesting approaches to solve at least some of the issues that are dis-
cussed within the system were collected from the literature and even more from the
interviews with the national agencies. On the national level, some of the problems
have been resolved by measures which could partially serve as models for the
whole EU system.
2.2.7 Task 7: Recommendations

According to the original project plan, recommendations for making the European
Community system of pharmacovigilance more robust should be deducted based
on task 6 and discussed in the expert workshop. Since the expert workshop could
not be postponed for organisational reasons, draft recommendations were derived
basically from the literature review and the interviews and discussed at the expert
workshop. The draft character of the recommendations was in accordance with the
requirement that the final recommendations were as far as possible independent
from the influence of the stakeholders (e.g. advisors), but nevertheless were in-
formed by external expertise.
The preliminary conclusions and recommendations were discussed with the advi-
sors during the workshop on June 15.
2.3 Collection and analysis of data

The empirical approach was based on personal interviews with representatives of
the competent authorities and a written agency survey. In addition, two Delphi-
surveys were carried out.
Own data were as far as possible collected with the same questions that were used
in the earlier surveys of the Heads of Medicines Agencies ERMS Working group,
but nevertheless are not totally comparable with the ERMS data. In addition, the
samples differ in the countries that are covered. The ERMS-survey of 2002 cov-
ered the 15 old EU Member States with 2 datasets for Germany (BfArM and PEI)
as well as NO (n=17 datasets), the 2004 ERMS-survey comprised the 10 New EU
Member States.
Table 2.2. Datasets for the analyses

N of data-
sets
ERMS survey 2002 17
(Old EU MS, 2 datasets from DE, plus NO)
ERMS survey 2004 10
(New EU MS)
Own survey:
Interviews 27
(25 EU MS, one additional dataset from second German agency, plus EMEA)
Written survey completed by agency 30
(25 EU MS, one additional dataset from second German agency, plus EMEA,
plus 3 EEA MS: IC, LI, NO)
One common dataset for the 2 German agencies, plus one for the EU-25 2
Own survey maximum number of datasets 32
Caused by the different samples used, quite frequently different numbers of re-
spondents have to be taken into consideration; missing values for single countries
add to this and lead to variable sample sizes, but only to small differences in the
appearance of some figures (single countries missing etc.).
2.3.1 Interviews
27 site visits in all 25 EU Member States and at the EMEA have been conducted
(Table 2.3).
Table 2.3. Site visits for interviews
Date of site visit Country5 City carried out by

10.06.2005 Latvia Riga ISI
09.06.2005 Estonia Tartu ISI
09.06.2005 Lithuania Vilnius ISI
07.06.2005 Ireland Dublin ISI
02.06.2005 Poland Warszawa ISI
30.05.2005 Czech Republic Praha ISI
26.05.2005 Greece Athens ISI
24.05.2005 Cyprus Lefkosia ISI
20.05.2005 Hungary Budapest ISI
19.05.2005 Slovak Republic Bratislava ISI
18.05.2005 Malta Gzira KKS
11.05.2005 Italy Roma KKS
10.05.2005 Portugal Lisboa ISI
09.05.2005 Spain Majadahonda - Madrid ISI
04.05.2005 Sweden Uppsala KKS
03.05.2005 Finland Helsinki KKS
02.05.2005 Denmark Copenhagen KKS
02.05.2005 France Saint-Denis ISI
27.04.2005 Germany-PEI Langen KKS
26.04.2005 Germany-BfArM Bonn ISI and KKS
20.04.2005 UK-MHRA London ISI
14.04.2005 Belgium Brussels KKS
12.04.2005 Austria Wien KKS
07.04.2005 Luxembourg Luxembourg KKS
07.04.2005 Netherlands Den Haag KKS
04.04.2005 Slovenia Ljubljana KKS
16.03.2005 EMEA London ISI and KKS
The site visits were carried out by only four persons (two senior researchers from
both contractors each) to ensure sufficient consistency in the carrying-out of the
interviews.
The interviews were done on the basis of an interview guide and took about four
hours each. In most cases, the agency's head of pharmacovigilance and one or
two members of the staff, sometimes also the head of the division were present at
least for a part of the time. The main topics of the interview guide are questions
with respect to process activities (especially data collection, data management,
quality control/quality assurance, safety signal detection, safety issue assessment,
decision making process, action plans to protect public health, communication
process with stakeholders, quality assurance), the relevant stakeholders and ques-
tions with respect to the resource availabilities/functional capabilities of these
stakeholders.
The collected interview data were stored in an MS-Access database to allow easy
handling and the production of overviews on the answers to specific questions
5 In Germany two agencies are responsible for PhV on the national level, the Paul-
Ehrlich-Institute (PEI), which is responsible for blood products, biologicals and
vaccines, and the Bundesinstitut für Arzneimittel und Medizinprodukte (BfArM),
which is responsible for all other medicinal products.
across agencies, as well as to use the quantitative data for statistical analyses to-
gether with the data collected within the written questionnaire.
The textual data from the interviews were categorized and summarized. The fre-
quency of different categories of answers was counted. Parts of the interview data
(mainly closed questions and numerical data) were analysed statistically with
SPSS version 11.
2.3.2 Written agency survey

The written survey was sent out as an electronic form (MS-Excel sheet) to the
agencies on 7 July 2004. All Agencies from EU Member States, EEA countries and
EMEA participated in the survey. The last completed questionnaires were returned
on October 28 and November 8, respectively.
The data were mostly submitted in electronic form. They were reviewed for com-
pleteness, eventual problems within the data were clarified, and the data were
stored in an Excel database. From there they were imported into SPSS version 11,
which was used to analyse statistically the survey data and parts of the interview
data.
Different reporting rates were computed with the number of collected ADR reports
related to the population size in the countries, but also to the type of products,
(NAPs, MRPs, CAPs), to the value of pharmaceutical sales, and to the density of
physicians in the country.
These reporting rates were correlated with various external criteria to evaluate their
explanatory power (see two following tables). Spearman-rho correlation coefficients
were calculated to account for unsymmetrical distributions which are very likely in
studies like the present in which only a small number of cases (here: the 29 agen-
cies) are available.
The criteria are population sizes from Eurostat, as well as figures on density of
physicians, pharmaceutical sales, and the incidence of ADR-relevant diseases (a-
ge-standardized death rate; absolute numbers and per 100000) from WHO-Euro
European health for all database (HFA-DB; source: http://data.euro.who.int; those
numbers of 2002).
In our analysis the population of children was defined as ≤19 years of age, be-
cause only such data were available for all participating countries. However, the
numbers of reports for "children" were probably counted for persons <18 years.
Therefore, the computed reporting rates for children will underestimate the true
value a bit.
The following table shows that the population-based reporting rates for 2003 and
2004 correlate with none of the external criteria except the 2003 population sizes.
However, this is at least partially an effect of the small sample size of maximally 28
countries.
Table 2.4. Correlations of population-based reporting rates with external criteria

Correlation … with Physicians Pharma-sales Incidence Incidence
of … ↓ → total per total per ADR- ADR-relevant
100000 capita (in US$) relevant diseases
capita diseases (per 100000
(absolute) capita)
Reporting rate Rho 0.215 0.012 0.125 -0.147
total 2003 per
million capita p 0.293 0.963 0.560 0.494
N 26 17 24 24
Reporting rate Rho 0.066 -0.178 0.106 -0.116
total 2004 per
million capita p 0.738 0.467 0.605 0.574
N 28 19 26 26
The rows "Rho" give the correlation coefficient for the reporting rates of 2003 and
2004 with the external criteria.
None of the correlations for the population-based reporting rates with the external
criteria was significant on the 5%-level (row "p"); the same was true for the report-
ing rates for children (not presented). The rows market with an "N" contain the
number of valid cases for which the single correlations could be computed.
To compute a reporting rate by dividing the absolute number of ADRs collected in a
country by the number of inhabitants of this country means to control the reporting
for the size of the population so that countries with different population sizes are
comparable in the relative reporting rate. But, because the frequency of ADRs in a
population can plausibly not only depend on the size of the population but will also
vary with the number of physicians which can submit ADR reports, or with the vol-
ume of pharmaceutical products that are sold within a country, two other reporting
rates were computed that seem to be more adequate than the one that is only
based on the size of the population. These rates do not only control for the size of
the populations, but the first also for the number of physicians and the second for
the pharmaceutical sales; therefore they are more valid indicators for the function-
ing of the national pharmacovigilance systems than the population-based rate.
As the following table shows, the reporting rates based on pharmaceutical sales
(row "Reporting rate total 2004 per sales in US$" as well as those based on num-
bers or physicians in the countries (row " Reporting rate total 2004 per physicians
per 100,000 capita"), do not only correlate with the population size, but also with
the WHO-figures on the absolute incidence of ADR-relevant diseases which the
population-based indicator (row "Reporting rate total 2004 per million capita") does
not. The correlation disappears for the relative incidence rate (Table 2.5).
Table 2.5. Correlations of sales-based reporting rates with external criteria

Correlation … with Physicians Pharma- Incidence Incidence
of … ↓ → total per sales total ADR- ADR-relevant
100000 capita per capita (in relevant diseases
US$) diseases (per 100000
(absolute) capita)
Reporting rate Rho 0.075 -0.244 0.025 -0.110
total 2004 per
million capita p 0.720 0.362 0.911 0.617
N 25 16 23 23
Reporting rate Rho -0.009 0.288 0.756 0.503
total 2004 per
sales in US$ p 0.974 0.279 0.001** 0.047*
N 16 16 16 16
Reporting rate Rho 0.168 0.312 0.517 -0.137
total 2004 per
physicians per p 0.423 0.240 0.012* 0.534
100,000 capita N 25 16 23 23
Source: Fraunhofer ISI 2005; ** correlation significant on the 1%-level; * correlation significant on the
5%-level
The best reporting rate – measured by its correlation with the WHO incidence rates
– is the reporting rate standardized at the pharmaceutical sales in the middle row. It
has the highest absolute correlations of Rho=.76 with the absolute incidence and
Rho=.50 with the relative incidence, and both of the correlations are significant or
highly significant despite the small sample size. This reporting rate should be used
for further analyses, as it also controls for different consumption patterns in the
countries.
2.3.3 Delphi survey

2.3.3.1.1 Critical success factors
The preliminary list of critical success factors for the pharmacovigilance processes
was submitted to the expert panel with the following question: "Relevance: How
important is the factor for the performance of the European System for Pharma-
covigilance (or parts of it)?"
Each factor was assessed according to the criteria
• Quality of the work
• Compliance with requirements,
• Speed ("kinetics"),
• Work load/costs,
on a five-point-rating scale (values: ++; +; 0; -; --). The values were explained as
++: strong positive influence; 0: not relevant; --: strong negative influence, and
space was left for comments and for the indication and evaluation of additional
important factors.
The following Figure 2.1 contains a part of the evaluation form. The full form includ-
ing the aggregated values collected from the participants can be found in Annex 3.
Figure 2.1. Delphi survey form for the evaluation of success factors (part)
Evaluation Round 1: Relevance for…
… compliance with
… quality of the work … speed ("kinetics") … work load/costs
requirements
Success factor ++ + 0 - -- ++ + 0 - -- ++ + 0 - -- ++ + 0 - --
1. … for Data collection
1.1 Comprehensiveness of the data
Mandatory reporting by HCPs 1 1 1 1
Spontaneous reports from pharmacists 1 1 1 1
Access to FDA data for national agencies 1 1 1 1
Access to drug utilisation statistics 1 1 1 1
Access to database of patients' medical
records 1 1 1 1
Highest-possible number of spontaneous
reports 1 1 1 1
The evaluations by the survey participants in each cell of the table were summed
up across the participants. Some of the success factors were not assessed by all
participants. Therefore, to make the results comparable between indicators, per-
centages of answers in this cell of all answers were computed for each cell.
2.3.3.1.2 Performance indicators

The experts were asked either to complete an electronic form or to print and com-
plete it manually. The form is presented in Figure 2.2.
Figure 2.2. Delphi survey form for the evaluation of performance indicators (part)
Evaluation Round 1
Relevance Practicability Interpretation
Performance indicator 3 2 1 0 3 2 1 0 3 2 1 0
1. ...for the input
Total number of ICSRs from your country received in last year 1 1 1
Number of ICSRs from your country received in last year from MAHs 1 1 1
Number of ICSRs from your country received in last year direct from HCPs 1 1 1
Number of ICSRs from your country received in last year direct from patients 1 1 1
Number of ICSRs from your country received in last year direct from pharmacists
Number of ICSRs from your country received in last year direct from other HCPs
Number of cases received/total number of ICSRs from your country
% of serious ICSRs from your country 1 1 1
% of ICSRs from your country as concerned MS 1 1 1
% of ICSRs from your country as reference MS 1 1 1
Number of PSURs received by origin and type of product 1 1 1
Number of studies carried out on national database/ target number for database
studies 1 1 1
The indicators were rated on three dimensions on rating scales with 4 values (see
Table 2.6).
It was explained that some of the indicators would in their final version need com-
bination with other indicators or relation e.g. to the size of the country in order to
compute relative indicators or percentages.
Table 2.6. Criteria for the evaluation of indicators in the Delphi survey
Dimension/Explanation Scale
Relevance:
How important is the indicator to obtain a valid picture of the 3: very relevant … 0: not relevant
performance of the European System for Pharmacovigi-
lance?
Practicability:
How easy is it to obtain the data for this indicator? 3: very easy to measure … 0:
measurable only at very high costs
We suppose that the data would have to be collected by the
national agency or come from other sources.
Please assume the availability of data in the country/region
for which your agency is responsible in January 2006.
Interpretation:
How easy is it to interpret the results? 3: very easy to interpret … 0:
nearly not interpretable
The evaluations by the survey participants in each cell of the table were summed
up across the participants. Again, some of the indicators were not assessed by all
participants. To make the results comparable between indicators, percentages of
answers in this cell of all answers were computed for each cell.
3 System analysis and description of the status

quo
In this section an overview of the European system of pharmacovigilance is pro-
vided It is structured in the following way: Section 3.1 describes the regulatory fra-
mework for pharmacovigilance. It is based on an overview provided by Professor
Gassner from the University of Augsburg, Germany, and KKS-UKT. The implemen-
tation of the legal framework is reviewed in section 3.2, and important points from
the systems in the USA, Japan and Canada are presented in paragraph 3.3.
The empirical approach to describe the status quo in the EU (including 25 EU
Member States and EMEA) is based on personal interviews with representatives of
the competent authorities, the supplementing questionnaire survey among agen-
cies and on telephone interviews with additional stakeholders from industry, sci-
ence and independent organisations on specific topics. The results from these
work steps are presented in sections 3.4 to 3.8.
The results are presented as objectively as possible without an appraisal of the
results by the authors.
Results of the Delphi process and the agency interviews also form the basis for the
elaboration of critical success factors and performance indicators which were used
to design the surveys; these steps will be presented in chapters 5 and 6, respec-
tively. Then, the most important results from the case studies are presented (chap-
ter 7) and some examples for best practice are given in chapter 8).
Pharmacovigilance is described as consisting of six phases:
• Data collection,
• Data management,
• Signal detection,
• Safety issue assessment,
• Decision-making,
• Communication and action to protect public health.
Besides this, general aspects are described in terms of
• Framework conditions, particularly the regulatory framework
• Resources for pharmacovigilance,
• Definitions and standards,
• General quality management,
• Outcomes.
3.1 Description of the regulatory framework of

pharmacovigilance in Europe
In this section, the regulatory framework that shapes the functioning of the Euro-
pean pharmacovigilance system is presented.
3.1.1 Introduction
The European Medicines Agency defines pharmacovigilance as “the process of
monitoring, evaluating and improving the safety of medicines in use. It is carried
out by pharmaceutical companies on their products and by government agencies
on all medicinal products. Healthcare Professionals (e.g. doctors and pharmacists)
have a role too, in reporting suspected side effects of medicines to government
agencies or pharmaceutical companies (EMEA 2005).
The World Health Organisation (WHO) defines pharmacovigilance as “the science
and activities relating to the detection, assessment, understanding and prevention
of adverse effects or any other drug-related problems”.
Pharmacovigilance activities include actions to detect and assess adverse drug
reactions (ADR), evaluation of the probability of a causal relationship between the
medicinal product and the adverse drug reaction, and actions taken in order to pro-
tect public health. This means e.g. the establishment of systems for the reporting of
individual cases of adverse drug reactions to the supervising authority, discussion
of safety problems within expert committees, the order of the Authorities for under-
taking epidemiological safety studies or the change of the authorisation status of a
medicinal product.
One of the objectives of the project “Assessment of the European Community Sys-
tem of Pharmacovigilance” is to describe the current system regarding the pharma-
covigilance of marketed medicinal products for human use6. The legal framework
of this system is based on the European pharmaceutical legislation, whose regula-
tions are applicable on the Community level as well as on the level of the EU
Member States (MS). The three other EEA Members Iceland, Liechtenstein and
Norway have also joined this framework, and together with the central authorities
and EU Member States they build the European system of pharmacovigilance. The
actual application of the European legislation in the Member States varies, due to
partly needed implementation of the European laws into the national legislation and
due to the adjustment to the national conditions.
In the present report the legal framework of laws and guidance documents that has
to be applied in the European system of pharmacovigilance is presented and ex-
amined. National laws of the Member States do not lie within the scope of this re-
port.
3.1.2 Compilation and description of the relevant European laws

and associated guidance documents
3.1.2.1 Legal framework of pharmacovigilance of marketed drugs
The interplay between national and EC authorities in the area of pharmacovigi-

lance, in particular with respect to the actions that can be taken and the procedures
applicable to the processing of safety concerns, depends on the type of marketing
authorisation (Bendall 2004).
6 The pharmacovigilance for veterinary medicinal products, which is not addressed

here, is regulated in a quite similar manner.
Basically, there are two routes for marketing medicinal products throughout the EU:
a centralised procedure at European level and a decentralised system at national
level encompassing two types of authorisation procedures. A marketing authorisa-
tion for a medicinal product in more than one Member State must therefore be ap-
plied for through one of three procedures: either the “Centralised Procedure”, de-
termined by Regulation (EC) No 726/2004, or the “Mutual Recognition Procedure”
or the new “Decentralised Procedure”, regulated by Directive 2001/83/EC. Of
course, national authorisations remain available for products to be marketed in one
single Member State. Even purely national marketing authorisation procedures are,
however, subject to harmonising provisions of Directive 2001/83/EC.
Centralised Procedure
The Centralised Procedure is administered by the EMEA. It consists of a single
application which, when approved, grants marketing authorisation for all markets
within the European Union (and the EEA).The European Commission is the re-
sponsible competent authority for the products which come to the market through
the centralised procedure. This procedure is available to all new, or so-called “in-
novative” pharmaceuticals, and is obligatory for biotechnology-derived medicines
and products containing new substances, for which the therapeutic indication is the
treatment of several severe diseases.
Mutual Recognition Procedure

Under this procedure the assessment and marketing authorisation of one Member
State, the reference Member State, should be “mutually recognised” by other con-
cerned Member States. Member States who recognise the first authorisation on the
basis of the assessment report to be prepared by the reference Member State
within 90 days, will grant a marketing authorisation with an identical summary of
product characteristics. If a Member State raises objections and does not recog-
nise the original marketing authorisation the matter may be referred for arbitration
to the EMEA.
Decentralised Procedure
The new Decentralised Procedure is applicable in cases where an authorisation
does not yet exist in any of the Member States. Identical dossiers will be submitted
in all Member States where a marketing authorisation is sought. A reference Mem-
ber State, selected by the applicant, will prepare draft assessment documents
within 120 days and send them to the concerned Member States. They, in turn, will
either approve the assessment or the application will continue into arbitration pro-
cedures.
Figure 3.1. Authorisation Procedures
Authorisation
Procedures
Centralised Procedure Mutual Recognition Procedure Decentralised Procedure

(Regulation (EC) (Directive 2001/83/EC) (Directive 2001/83/EC)
No 726/2004)
The legal framework of pharmacovigilance for drugs marketed within the EU is

specified mainly in Articles 21 to 29 of Regulation (EC) No 726/2004 with respect to
centrally authorised medicinal products and in Articles 101 to 108 of Directive
2001/83/EC with respect to both decentrally and nationally authorised medicinal
products. The Community has sought over the years to ensure that the pharma-
covigilance systems for centrally authorised medicinal products and those author-
ised by other procedures become more and more consistent. Yet, there are some
disparities and inconsistencies resulting from a non-optimal compliance of both
national law and practice with the EC regulations.
The said basic legal texts are supplemented by Commission Regulation (EC) No
1085/2003 and Commission Regulation (EC) No 1084/2003, which describe the
procedures that have to be followed in the case that an existing marketing authori-
sation of medicinal products on the European market has to be changed, further by
Commission Regulation (EC) No 540/95 that regulates the procedures concerning
“suspected unexpected non-serious adverse reactions”.
Additionally detailed instructions, definitions, standards and information regarding
the precise conduct of pharmacovigilance related procedures are to be found in a
number of guidance documents, first of all in “Volume 9 of the rules governing me-
dicinal products in the European Union – Pharmacovigilance” and in the pharma-
covigilance related guidelines of the International Conference on Harmonisation.
In the following sections, the regulatory state of the art is described in more detail.
3.1.2.1.1 Applicable European laws concerning centrally authorised medicinal

products
Regulation (EC) No 726/2004

Council Regulation (EEC) No 2309/93 was replaced by Regulation (EC) No
726/2004.
Title II of Regulation (EC) No 726/2004 which includes the provisions relating to the
pharmacovigilance of human drugs will apply from 20 November 2005.
Relevant information concerning pharmacovigilance are to be found in:
• Article 13 para 4, subpara 3 and article 16 para 2, subpara 2 and 3 (MHA’s

obligation for the provision of data concerning the sales, prescriptions and
the benefit risk evaluation of a product to the EMEA)
• Articles 21 to 29 “pharmacovigilance“ (tasks and procedures of the EMEA,
responsibilities and procedures of the Marketing Authorisation Holder
(MAH) and the Member States, reference to guidance document Volume 9
and to the data network of the Authorities, cooperation with the WHO)
• Articles 19 and 20 (control and execution procedures as to MAHs’ obliga-
tion to fulfil the requirements laid down in Title IX of Directive 2001/83/EC)
• Article 57 para. 1 (c) to (f) (tasks of the EMEA)
• Article 67 para 4 (funding of activities relating to pharmacovigilance)
Figure 3.2. Regulation (EC) No 726/2004
REGULATION (EC) NO 726/2004
• applicable to centrally authorised medicinal products
• Title II of the Regulation concerning, i.a., the pharmacovigilance of human

drugs applicable with effect from 20 November 2005
• published in Volume 1 of “The rules governing medicinal products in the
European Union”:
http://www.pharmacos.eudra.org/F2/eudralex/vol-1/home.htm
Commission Regulation (EC) No 540/95

• Commission Regulation (EC) No 540/95 complements Regulation (EC) No
726/2004.
• The Regulation lays down specific requirements for reporting non-serious
unexpected adverse reactions.
• It is published in Volume 1 of “The rules governing medicinal products in the
European Union”: http://pharmacos.eudra.org/F2/eudralex/vol-1/home.htm.

• Commission Regulation (EC) No 1085/2003 describes the procedures to be
applied to change the marketing authorisation of centrally authorised me-
dicinal products and to temporarily restrict their authorisation in case of
emergency measures.
• It is published in Volume 1 of “The rules governing medicinal products in the
European Union”: http://pharmacos.eudra.org/F2/eudralex/vol-1/home.htm.
3.1.2.1.2 Applicable European laws concerning non-centrally authorised medici-

nal products
Directive 2001/83/EC
Directive 2001/83/EC constitutes the Community Code for medicinal products for
human use marketed in the EU and authorised either in one single Member State
or in more than one Member State under either the Mutual Recognition Procedure
or the Decentralised Procedure.
The Directive was amended by Directive 2004/27/EC. This Directive should, i.e.,
step up pharmacovigilance and, more generally, market surveillance and sanctions
in the event of failure to comply with the provisions. Furthermore, in the field of
pharmacovigilance, account should be taken of the facilities offered by new infor-
mation technologies to improve exchanges between Member States. The imple-
mentation into national law has to be completed no later than 30 October 2005.
The relevant information concerning pharmacovigilance and sanctions in the event
of failure to comply with the provisions are to be found in:
• Article 1 „definitions“, in particular Article 1 Nos. 11 to 16
• Article 8(3)(ia) and 8(3)(n) (obligation of the applicant of a marketing au-
thorisation concerning pharmacovigilance)
• Articles 23 paras. 1 and 3, 23a para. 3 “information obligations as regards
marketed products”
• Articles 31, 32, 36 “Community referrals” (community interests, precondi-
tions and procedure)
• Articles 101 to 108 „pharmacovigilance“ (tasks, responsibilities and proce-
dures of the Member States and the EMEA, responsibilities and procedures
of the MAH, reference to guidance document Volume 9 and to the data
network of the Authorities)
• Article 111 “pharmacovigilance inspections”
• Articles 116, 117 „supervision and sanctions“ (responsibilities of the Mem-
ber States)
• Articles 122, 123 (notification obligation of the Member States and of the
MAH in case of changes of the authorisation status and emergency meas-
ures).
• Article 127a (Commission decisions on risk management)
Figure 3.3. Directive 2001/83/EC
DIRECTIVE 2001/83/EC
• Community Code for medicinal products for human use in the EU
• applicable to nationally and non-centrally authorised medicinal products +

some provisions are also relevant to centrally authorised products
• sat last amended with respect to pharmacovigilance by Directive

2004/27/EC
• published by the European Commission in Volume 1 of “The rules govern-
ing medicinal products in the European Union”:
http://www.pharmacos.eudra.org/F2/eudralex/vol-1/home.htm

Commission Regulation (EC) No 1084/2003 describes the procedures to be ap-
plied to change the marketing authorisation of medicinal products that are author-
ised with the procedure of mutual recognition or that are subject of a referral (acc.
to Articles 32, 33 and 34 of Directive 2001/83/EC) and to temporarily restrict their
authorisation in case of emergency measures.
• The Regulation is published in Volume 1 of “The rules governing medicinal
products in the European Union”:
http://www.pharmacos.eudra.org/F2/eudralex/vol-1/home.htm.
3.1.2.1.3 New Community legislation

The new Community legislation, coming into force in November 2005, will introduce
additional tools to strengthen further the existing pharmacovigilance system in
terms of communication quality and quantity. It comprises
• the submission of risk context data and, where appropriate, the description
of the risk management system the applicant will introduce by applicants for
a marketing authorisation;
• the collection of specific pharmacovigilance data for centrally authorised
products from targeted groups of patients;
• the possibility for regulatory Authorities to take urgent provisional measures,
for instance as a result of the evaluation of pharmacovigilance data;
• a reinforcement of the benefit/risk balance concept in the scientific assess-
ment throughout the life cycle of medicinal products;
• a shorter Periodic Safety Update Report (PSUR) periodicity;
• a mandatory electronic reporting, save in exceptional circumstances, of

ADRs by the National Competent Authorities (NCAs) and the MAHs;
• a strengthening of the enforcement through the possibility for financial pen-
alties for pharmaceutical companies in case of non-adherence to the legal
obligations.
Furthermore the new Community legislation concentrates on meeting the aim of
transparency. Thus, e.g., the EudraVigilance database will be made accessible to
Healthcare Professionals and the general public (Heads of Medicines Agencies
2005a; Moseley 2004).
3.1.2.2 Related guidance documents
Figure 3.4. Related guidance documents
RELATED GUIDANCE DOCUMENTS
Volume 9 of the rules governing medicinal ICH/CHMP Guidelines

products in the EU
• European Commission – repre-
• legal basis: Article 26 of Regula- sented by the CHMP of the EMEA –
tion (EC) No 726/2004 and Article adopted 6 Guidelines of the ICH
106 para. 1 of Directive
2001/83/EC • objective: achievement of greater
harmonisation in the interpretation
• objective: giving guidance on the and application of technical guide-
collection, verification and presen- lines and requirements
tation of adverse drug reports
• scientific guidelines
• pharmacovigilance guidelines
• no legal force yet binding in practi-
• no legal force yet binding in practical terms (EMEA 2004d).
cal terms (EMEA 2004d)
• published in Volume 9 of “The
rules governing medicinal products
in the EU”:
http://pharmacos.eudra.org/F2/eud
ralex/vol-9/home.htm
3.1.2.2.1 “Volume 9 of the rules governing medicinal products in the European

Union” – Pharmacovigilance
This document is drawn up by the European Commission in consultation with the
EMEA, Member States and interested parties in accordance with Article 26 para. 1
of Regulation (EC) No 726/2004 and Article 106 para. 1 of Directive 2001/83/EC to
give “guidance on the collection, verification and presentation of adverse reaction
reports in order to facilitate the exchange of information about pharmacovigilance
(of authorised medicinal products for human use and veterinary medicinal prod-
ucts) within the Community (Eudralex Volume 9, Pharmacovigilance Guideline,
No. 1 2004). The current version of Volume 9 dates from June 2004. This version
is actually being updated to reflect the new legislation.
Volume 9 comprises 4 parts with the following content:

• Part I „Guidance and Procedures for Marketing Authorisation Holders” and
„Guidance and Procedures for Competent Authorities” describes in detail
the tasks, obligations and procedures of the two main responsible parties in
the field of pharmacovigilance and gives concrete guidance, definitions and
standards for the performance of all relevant processes.
• Part II deals with pharmacovigilance of veterinary medicinal products.
• Part III “EU Electronic Exchange of Pharmacovigilance Information” de-
scribes the technical requirements for the electronic exchange of pharma-
covigilance related information in the Community and references to the
agreed terminology.
• Part IV “Reference Legislative and Administrative Information” refers in
general to legal information and in particular to ICH Guidelines E2B(M) and
E2C (with E2C addendum) that are integrated in Volume 9.
3.1.2.2.2 ICH/CHMP guidelines

The EU, through its representation on the ICH Steering Committee and through
subsequent adoption of ICH guidelines by the Committee for Medicinal Products
for Human Use (CHMP) of the EMEA, has adopted six tripartite guidelines of the
International Conference on Harmonisation (ICH) relating to pharmacovigilance.
ICH guidelines are also incorporated into Volume 9 when this volume is updated.
ICH Guidelines serve “to achieve greater harmonisation in the interpretation and
application of technical guidelines and requirements” (International Conference on
Harmonisation of Technical Requirements for Registration of Pharmaceuticals for
Human Use (ICH) 2005) by the development and the use-application of medicinal
products. They represent the international standard regarding definitions, formats
and technical requirements.
The adoption of ICH Guidelines through the CHMP and Volume 9 has as a result
that the international standards described in the guidelines have to be adhered to
within the EU regarding pharmacovigilance related processes. These guidelines
are, however, not legally binding in a strict sense.
The guidelines explicitly refer to definitions, management and expedited reporting
of individual adverse reaction cases, including electronic formats, periodic reporting
of worldwide safety data and planning of pharmacovigilance activities (Arnold
2004).
ICH Guideline E2A (CPMP/ICH/377/95)

• Title: Clinical Safety Data Management: Definitions and Standards for Ex-
pedited Reporting
• This guideline came into operation in June 1995.
• The guideline is published under
http://www.emea.eu.int/pdfs/human/ich/037795en.pdf
Figure 3.5. Key points guideline E2A
KEY POINTS GUIDELINE E2A (Bahri and Tsintis 2005d})
• Definitions for adverse event (AE) and adverse drug reaction (ADR) in the pre-
authorisation phase
• Criteria for serious AE/ADR
• Expectedness of an AE/ADR based on clinical observations and its documentation in
the applicable product information
• Causality assessment as good case practice for AE/ADR cases from clinical trials
• Implied possible causality for spontaneously reported ADR cases
• Standards for expedited reporting from clinical trials
• Definition of minimum case report information for report submission to authorities
• Follow-up reporting
• Unblinding procedures for serious ADRs
• Reporting of emerging information on post-study ADRs
• Reporting requirement for active comparator
ICH Guideline E2B(M) CPMP/ICH/287/95

• Title: Clinical Safety Data Management: Data Elements for Transmission of
Individual Case Safety Reports
• This guideline came into operation in November 2000 with minor editorial
changes in March 2001.
http://www.emea.eu.int/pdfs/human/ich/028795en.pdf
Figure 3.6. Key points guideline E2B(M)
KEY POINTS E2B(M) GUIDELINE (Bahri and Tsintis 2005)
• Description of all data elements of ADR case reports: title and content of each data
field
• Technical specifications such as field length and field value for each of the data fields
and the related additional technical data fields
• List of abbreviations for units
• List of units for time intervals
• List of routes of administrations
ICH Guideline E2C (CPMP/ICH/288/95)

• Title: Clinical Safety Data Management: Periodic Safety Update Reports for
Marketed Drugs
• This guideline came into operation in June 1997.
http://www.emea.eu.int/pdfs/human/ich/028895en.pdf.
Figure 3.7. Key points guideline E2C
KEY POINTS E2C GUIDELINE (Bahri and Tsintis 2005)
• Inclusion of all product presentations on one PSUR

• Concept of international birth date of a product, determining the data lock points
of PSURs
• Provision to submit a set of PSURs, each covering subsequent 6 months, to facilitate
PSUR submission acc. to local frequency
• Description of all data sources to be covered in a PSUR
• Inclusion of worldwide information on marketing authorisation status and regulatory
safety-related action, ADR and exposure data
• Use of company core safety information (CCSI) as reference and concept of unlisted-
ness of an ADR (i.e. unlisted in comparison to the CCSI versus unexpected in com-
parison to local authorised product information)
• Presentation of individual case history
• Formats of ADR line-listings and summary tabulations
• Presentation of exposure data
• Overall safety evaluation and conclusion: analysis and discussion of data by MAH with
view to possible safety-related action
• Explanation on responsibilities of MAHs in contractual relationship
• Annex of medically unconfirmed ADR case reports to be submitted as requested locally
ICH Guideline E2C Addendum (CPMP/ICH/4679/02)

• Title: Addendum to ICH E2C: Clinical Safety Data Management, Periodic
Safety Update Reports for Marketed Drugs
• The E2C addendum guideline is in operation since August 2003.
Figure 3.8. Key points guideline E2C Addendum
KEY POINTS E2C ADDENDUM GUIDELINE (Bahri and Tsintis 2005)
• Clarification regarding inclusion of all product presentations in one (PSUR)

• Executive summary as new part of the PSUR
• New statement of proprietary information to be included in PSUR
• Use of reference safety information in relation to time periods covered by PSUR
• Further guidance on presentation of exposure data
• Organisation of some PSUR parts by system organ class
• Risk management programmes, if in place for the product, to be discussed in PSUR
• Separate benefit-risk analysis, if conducted recently for the product, to be discussed in
PSUR
• Recommendations for PSUR submission during transition period of harmonisation
towards international birth date; clarifications for such harmonisation
• Clarification on restart of PSUR submission frequency
• New concept of summary bridging report supporting submission a set of covering 6
mths/PSUR
• New concept of addendum report to cover the period between last PSUR and local
MAH renewal date
ICH Guideline E2D (CPMP/ICH/3945/03)

• Title: Post Approval Safety Data Management
• This guideline came into operation in May 2004.
Figure 3.9. Key points guideline E2D
KEY POINTS E2D GUIDELINE (BAHRI AND TSINTIS 2005)
• Definitions for AE and ADR in the post-authorisation phase

• Criteria for serious AE/ADR in accordance with ICH-E2A
• Expectedness of an ADR based on clinical observation and its documentation in the
authorised product information; explanations regarding class effects
• Differentiation between sources of unsolicited and solicited reports
• Explanation on stimulated (but unsolicited) reporting
• Standards for expedited reporting in post-authorisation phase
• Definition of minimum case report information for report submission to authorities with
explanations
• Follow-up reporting
• Lack of efficacy reporting needs
• Guidance on ADR narratives
• Guidance on ADR case assessment
• Management of cases of exposure during pregnancy
• Explanation on reporting responsibility of MAH despite any contractual relationship in
place
ICH Guideline E2E (CPMP/ICH/5716/03)

• Title: Pharmacovigilance Planning (PVP)
• This guideline has come into operation in June 2005.
Figure 3.10. Key points guideline E2E
KEY POINTS E2E GUIDELINE (BAHRI AND TSINTIS 2005)
• Elements for pharmacovigilance specification as summary of identified risks, the risks

potentially arising from populations and situations which have not yet been adequately
studied and potential other risks
• Format of a pharmacovigilance plan based on the specification
• Within the pharmacovigilance plan, description of routine pharmacovigilance as mini-
mum and inclusion of a safety action plan for specific issues/missing information as
needed
• Format of safety action plan, with description of rationale for action and timetable for
evaluation and reporting (‘milestones’)
• Possible synchronisation of timetable with regulatory timetable for post-authorisation
assessment, such as PSUR assessment or marketing authorisation renewal assess-
ment
• Principles for design and conduct pharmacoepidemiological studies of non-experimental
design with references to international guidelines
• Overview of methods for data collection to investigate the known or unknown risks and
references
3.1.2.3 Legal basis of pharmacovigilance in clinical trials
For all those medicinal products which are being applied in clinical trials (that in-
cludes clinical trials performed to collect safety data) the relevant regulations to
pharmacovigilance are to be found in Articles 11, 16, 17 and 18 of Directive
2001/20/EC (implementation of Good Clinical Practice in the conduct of clinical
trials on medicinal products for human use). All medicinal products used in clinical
trials come under this Directive, regardless of their authorisation procedure (cen-
tralised, not centralised) or whether they are marketed or not.
The Directive was adopted in May 2001. It shall have been implemented into na-
tional law at the latest with effect from 1 May 2004.
The Directive is published in Volume 1 of “The rules governing medicinal products
in the European Union”:
http://pharmacos.eudra.org/F2/eudralex/vol-1/home.htm.
More details regarding definitions and processes relevant to pharmacovigilance,
which are applicable in clinical trials, are to be found in two guidance documents
published by the Commission pursuant to Articles 11, 16, 17 and 18 of Directive
2001/20/EC „Detailed guidance on the collection, verification and presentation of
adverse reaction reports arising from clinical trials on medicinal products for human
use” and „Detailed guidance on the European database of Suspected Unexpected
Serious Adverse Reactions (Eudravigilance – Clinical Trial Module)”.
The documents are published under
http://pharmacos.eudra.org/F2/pharmacos/dir200120ec.htm.
3.1.3 Relevant stakeholders involved in the pharmacovigilance

process
The wide spectrum of persons and institutions, who are involved in the field of me-
dicinal products, starting with the manufacturers and the MAHs, up to the patients
and the safety monitoring Authorities, who are in charge of protecting public health,
results in a great number of stakeholders, who are actively involved in this process.
The responsible organs for pharmacovigilance – the Authorities of the Member
States, the EMEA, the European Commission and the MAH of a medicinal product,
their responsibilities and obligations are explicitly described in the European phar-
maceutical legislation and the associated guidance documents.
Furthermore Healthcare Professionals (HCPs), patients, distributors and address-
ees of information and actions relevant to pharmacovigilance are included.
The obligations of the stakeholders, which are explicitly stated in the European
laws and guidelines, are presented below:
3.1.3.1 Member States
The Member States are obliged to operate a pharmacovigilance system.

MS encourage doctors and other HCPs to report suspected adverse reactions to
the NCAs. Furthermore, they impose specific requirements on doctors and other
HCPs in respect of the reporting of suspected serious or unexpected adverse reac-
tions.
They should have at their disposal sufficient personnel and infrastructure, in order
to ensure the conduct of pharmacovigilance.
3.1.3.2 Competent authorities and institutions
3.1.3.2.1 Preliminary note

Main actors in pharmacovigilance matters regarding the protection of public health
are several Authorities and institutions in the EU.
Due to the different legislations for medicinal products on the European market
more than one authority is responsible for the regulatory affairs.
3.1.3.2.2 The National Competent Authorities

The NCAs in the Member States are responsible for nationally authorised products
including products that are decentrally authorised. For this case the responsibility
for the conduct of pharmacovigilance including the implementation of regulatory
actions rests with the NCAs of all Member States that have granted a marketing
authorisation.
The NCAs continually monitor the safety profile of the products available on their
territory and take appropriate actions where necessary and monitor the compliance
of MAHs with their obligations with respect to pharmacovigilance. The NCAs are
also responsible for the communication with the MAH. In order to avoid duplicate
effort the Member States have agreed that the reference Member State as the
Member State that was leading in the process of the decentralised marketing au-
thorisation takes a leading function on all activities of pharmacovigilance. With re-
spect to centrally authorised products the Member States are responsible to moni-
tor medicinal products within their respective territories and act as the supervisory
authorities. However, the rapporteur that had a leading function in allocating the
marketing authorisation takes also a lead in pharmacovigilance, unless otherwise
decided by the CHMP (Moseley 2004).
3.1.3.2.3 The European Commission

The European Commission has overall responsibility for the EU system of pharma-
covigilance including policy and EU law.
The Commission is the Competent Authority in the case of centrally authorised
products and is responsible for the adoption of decisions based on opinions of the
CHMP relating to these products. As regards decentrally authorised medicines the
Commission is responsible to adopt decisions based on opinions of the CHMP for
those products that are subject to the referral procedures (Moseley 2004).
Furthermore it shall draw up guidelines on the collection, verification and presenta-
tion of adverse reaction reports, including technical requirements for electronic ex-
change of pharmacovigilance information in accordance with internationally agreed
formats, and shall publish a reference to an internationally agreed medical termi-
nology.
3.1.3.2.4 The EMEA

Acc. to Article 57 para. 1 (c) of Regulation (EC) No 726/2004 one of the tasks of
the EMEA is “to coordinate the supervision… of medicinal products which have
been authorised within the Community and to provide advice on the measures
necessary to ensure the safe and effective use of these products, in particular by
evaluation, coordination of the implementation of pharmacovigilance obligations
and the monitoring of such implementation”.
The Agency secretariat coordinates the pharmacovigilance related processes (in-
cluding Member States’ pharmacovigilance activities), gives advice on necessary
safety measures and provides information about adverse reactions through a data-
base. It is also responsible for the communication with the MAHs of centrally
authorised products and for coordination of issues relating to the monitoring of the
compliance of the MAH with its pharmacovigilance obligations (Moseley 2004).
3.1.3.2.5 The CHMP

The EMEA’s scientific committee, the CHMP, is responsible for providing scientific
advice evaluating evidence and formulating opinions on emerging safety issues of
centrally authorised products and of products that are subject of a referral (Moseley
2004).
3.1.3.2.6 The Pharmacovigilance Working Party

The principal task of the CHMP’s Pharmacovigilance Working Party (PhVWP) is “to
provide advice on the safety of medicinal products authorised in the European Un-
ion (EU) and the investigation of adverse reactions to enable effective identifica-
tion, assessment and management of risk, at any phase in the product life cycle.
On the basis of such advice the PhVWP will provide, where applicable, recommen-
dations for regulatory action to its stakeholders, i.e. the CHMP/EMEA and NCAs”
(EMEA 2004c).
Acc. to this document the key responsibilities of the PhVWP are:

• evaluation of potential signals arising from spontaneous reporting, including
those identified from the EudraVigilance database, and all other sources,
including epidemiological databases, studies and published literature;
• provision of advice on confirmation and quantification of risk and on regula-
tory options;
• risk management by advising on risk management plans;
• monitoring regulatory action and the outcomes of such action;
• setting standards for procedures and methodologies to promote good vigi-
lance practice;
• promotion of communication and exchange of information between the
EMEA and NCAs;
• international cooperation.
3.1.3.3 Marketing authorisation holders
MAHs are primarily responsible for the safety of their medicinal products, from the
start of drug development and throughout the lifecycle of a product.
“The MAH has to fulfil various pharmacovigilance system requirements which are
either explicitly laid down in legislation or are detailed in supporting guidelines”
(EMEA 2001).
The responsibility for the safety of the individual medicinal products rests with the
MAH. He is obliged to establish and operate a system, which allows the conduct of
all obligations that derive from the ongoing safety monitoring of the medicinal prod-
uct.
Figure 3.11. Key requirements to the pharmacovigilance systems of MAHs
SUMMARY OF PHARMACOVIGILANCE SYSTEM REQUIREMENTS OF THE MAHS

(EMEA 2001)
• expedited reporting
• periodic safety update reporting
• responding to requests for information from Competent Authorities
• handling of urgent safety restrictions and safety variations
• continuous monitoring of the safety profile of the authorised medicinal
product
• notifying Competent Authorities and health professionals of changes to the
risk-benefit profile of products
• meeting commitments made at the time of authorisation
• internal audit of the pharmacovigilance system
3.1.3.4 Health care professionals and patients
Medical specialists, doctors, nurses, pharmacists and others (depending on the

Member States’ regulations) constitute the group of the HCPs. They have direct
contact to the patients and they possess medical knowledge. Therefore, they are
an important source for the collection of safety data, which arise from the applica-
tion of medicinal products.
Moreover, HCPs are responsible to inform the patients about safety related prob-
lems and changes of the application of a medicinal product. Thus, the practical
implementation of pharmacovigilance and the prescription of a safe treatment rest
with them.
3.1.4 Pharmacovigilance related main processes and required

infrastructure
3.1.4.1 General conditions
According to European law both the Regulatory Authorities and the MAHs have to
provide the on-going supervision of the safety of medicinal products marketed in
the Community. To grant this task both parties are required to establish appropriate
systems.
The systems have to fulfil the following tasks (Eudralex 2005):
1. Collection and management of data relevant to medicines’ safety
2. The detection of new or changing ‘signals’ of medicines safety issues
3. Assessment and decision making with regard to safety issues
4. Action (including regulatory action) to protect public health
5. Communication / transparency with stakeholders ·
6. Audit, both of the outcomes of actions taken and of the key processes in-
volved.
3.1.4.2 Collection and management of data relevant to medicines safety
3.1.4.2.1 Collection of data relevant to medicines safety

Competent Authorities, MAHs and HCPs/consumers contribute to the collection of
pharmacovigilance relevant data.
Figure 3.12. Overview of processes in collection of data relevant to medicines

safety
DATA SOURCES HCPS/ MAHS AUTHORITIES

CONSUMERS
Spontaneous ad- • Generating • Mandatory collec- • Mandatory

verse reaction re- tion collection
ports • Reporting
• Mandatory report- • Mandatory
ing reporting
PSURs – • Preparation • Assessment
Published sources – • Mandatory en- • Mandatory

and literature quiry enquiry
Systematic data – • Initiation • Initiation
collection / studies
• Performance • Performance
3.1.4.2.2 Transmission of spontaneous adverse reaction reports

Stored and evaluated spontaneous reports have to be transmitted from the MAH to
the Authorities and between the Authorities. To simplify the data exchange the
EMEA has set up a data network that can be used by MAHs and the Authorities to
send a report and by the Authorities to retrieve the information simultaneously.
According to ICH E2D and ICH-E2A the classification of a report determines its
forwarding as expedited or not expedited.
3.1.4.2.3 Involved authorities
The Member States authorities
The expedited reporting

Competent Authorities of the Member States are obliged to evaluate and transmit
reports. Format and content shall comply with E2A and E2B.
The following figure summarises the expedited reporting by the Member States’
Authorities:
Figure 3.13. Expedited reporting obligations by MS Authorities
COLLECTION TRANSMISSION TO
• Reports of suspected serious ad- • MAH

verse reactions occurred within
• EMEA
the territory of the Member State
(transmitted by Healthcare Pro- • WHO (according to guidance on “Prin-
fessionals and the MAH) ciples of Providing the WHO with
pharmacovigilance information”
(Eudralex Volume 9, Pharmacovigi-
lance Guideline, No. 7 2004)
• Entry into the EudraVigilance database
• EU reports of suspected serious • No transmission
adverse reactions transmitted by
the EMEA
• Reports of suspected unexpected • No transmission
serious adverse drug reac-
tions/occurred outside EU and
authorised in the Member State /
transmitted by the MAH
Pharmacovigilance tools and resources
In general the established pharmacovigilance systems of the Member States

should provide the collection and scientific evaluation of data relevant for the
pharmacovigilance of medicinal products.
In order to achieve this, the Member States need a sufficient number of scientifi-
cally qualified staff to ensure that the collection, evaluation and transmission of
relevant information for decision-making processes and for the implementation of
necessary actions according to scientific standards is possible in required amount
and time.
Additional work of personnel of the Member States in different expert committees
on EC level (e.g. CHMP) has to be taken into account in this calculation.
Sources of pharmacovigilance relevant data

All Member States are obliged to search relevant literature and evaluate PSURs
and reports of performed Post-authorisation Safety Studies (PASS) for medicinal
products that are authorised in their territory. The latter includes an overall-risk-
benefit analysis and the preparation of assessment reports that are to be transmit-
ted to the Authorities of those Member States where the medicinal product is
authorised. In the case of centrally authorised medicinal products the reports have
to be transmitted additionally to the EMEA.
For centrally authorised medicinal products the assessment of the PSURs / reports
of PASS is performed by the rapporteur (or the chosen substitute). For decentrally
authorised drugs the reference Member State (or the chosen substitute) analyses
the reports for all concerned Member States.
The EMEA

EMEA is obliged to collect and transmit reports. Format and content shall comply
with E2A and E2B.
The following table summarises the expedited reporting by the EMEA:
Figure 3.14. Expedited reporting by the EMEA
• All reports of serious adverse drug • All reports of serious adverse drug re-
reaction / occurred within the EU / actions of centrally authorised prod-
transmitted by NCAs ucts to the NCAs
• All reports of suspected unex- • Entry into the EudraVigilance database

pected serious adverse drug reac-
tions / occurred outside EU and
authorised in the EU / transmitted
by NCAs and MAHs
EudraVigilance
EMEA has established a data network in cooperation with the Member States and
the European Commission for safe and fast electronic exchange of data between
the Authorities with the following levels of information: transmission of simple mes-
sages and free text documents (e.g. assessment reports or routine contacts), ex-
change of aggregate information as described for the Rapid Alert System (RAS)
and the Non Urgent Information System (NUIS) (see below 4.2.3.3.), exchange of
cumulative information, exchange of single case data via EudraVigilance. This
data-processing network and management system was launched in December
2001. It has been developed according to internationally agreed standards.
EudraVigilance is regarded as one of the main pillars of the European Risk Man-
agement Strategy.
The Authorities have to dispose of sufficient and appropriate electronic databases
and have to ensure that for electronic transmission of ICSRs and PSURs the fol-
lowing guidelines and specification can be fulfilled: ICH Guidelines E2A / E2B /
E2C / M1/ M2. Deadline for establishment and functioning of the system of elec-
tronic transmission of ICSRs of centrally authorised medicinal products is 20 No-
vember 2005.
Communication between the authorities

In the framework of the European laws the communication between the Authorities
of the Member States with the EMEA and the European Commission is described
in detail in the first place by Volume 9 (Eudralex Volume 9, Pharmacovigilance
Guideline, No. 2 2005).
Besides the transmission of adverse reaction reports the information exchange
between Authorities and the discussion of detected safety issues that could lead to
a change in the risk-benefit-balance of the product is central to this communication.
In order to be able to process information immediately in urgent cases, the Authori-
ties maintain the Rapid Alert System (RAS) and the Non Urgent Information Sys-
tem (NUIS). The basis of the system is EudraNet, a secure intranet established by
the EMEA, through which the data can be transmitted electronically.
Figure 3.15. Types of notification of safety concerns and exchange of information
TYPES OF NOTIFICATION
OF SAFETY CONCERNS AND
EXCHANGE OF INFORMATION
RAS NUIS
“The purpose of the RAS is to alert, with the appropriate degree of urgency, other
Member States, EFTA countries concerned, the Agency and the European Com-
mission about pharmacovigilance data related to medicinal products which indicate
that action could be needed urgently to protect public health. It is essential that the
communication of such problems occurs at an early stage, normally before a deci-
sion is taken in a Member State”.
The RAS should be used when a Member State is concerned about a change in
the balance between risks and benefits of a medicinal product that could require
major changes with respect to the validity or the content of the marketing authorisa-
tion such as:
• the urgent variation, suspension or withdrawal of the marketing authorisa-
tion, the recall of the medicinal product from the market;
• changes in the SPC such as
o the introduction of new contraindications,
o the introduction of new warnings,
o the reduction of the recommended dose,
o the restriction in the indications,
o the restriction in the availability of the medicinal product;
• the need to inform health care professionals or patients about an identified
risk without delay.
The NUIS is a procedure established to support the collection and exchange of

pharmacovigilance information between the Competent Authorities of Member
States, the EC and the EMEA, which does not fulfil the criteria for a Rapid Alert.
The NUIS refers, i.e., to
• pharmacovigilance data which do not require immediate or urgent action
and/or where additional information is required from other Member States to
support the evaluation of a potential concern,
• the provision of pharmacovigilance information not requiring a response.
3.1.4.2.4 The marketing authorisation holder

MAHs are obliged to collect, evaluate and transmit adverse reaction reports (acc.
to Article 104 para. 1-5 of Directive 2001/83/EC and Article 24 para. 1-2 of Regula-
tion (EC) No 726/2004). Format and content shall comply with E2A and E2B.
The following table summarises their expedited reporting:
Figure 3.16. Expedited reporting by MAHs
All reports of suspected serious adverse • To the Competent Authorities in

reactions/occurred in the EU/ spontane- the Member States in whose
ously reported by Healthcare Profes- territory the incident occurred
sionals
• Additionally to the reference
Reports of suspected serious adverse Member State in the case of
drug reactions / occurred in the EU / of decentrally authorised products
which the MAH can reasonably be ex- and products that have been
pected to have knowledge subject of a referral
Reports of suspected serious adverse • No transmission
reactions transferred from Member
States Authorities
All reports of suspected unexpected se- • All Member States where the
rious adverse reactions / occurred out- medicinal product is authorised
side EU / spontaneously reported by
Healthcare Professionals • EMEA
Sources of relevant pharmacovigilance data: The PSUR

The MAH is legally obliged to provide information on adverse effects of a medicinal
product in the form of a Periodic Safety Update Report (PSUR) immediately on
demand or in legally defined intervals to the Authorities that have granted a market-
ing authorisation (according to Article 104 para. 6 of Directive 2001/83/EC, Article
24 para. 3 of Regulation (EC) No 726/2004).
Format and content of this report shall comply with ICH Guideline E2C, E2C ad-
dendum.
The PSUR represents the world-wide safety experience of a medicinal product. It
contains all relevant new safety information from appropriate sources, data to pa-
tient exposure, the summary of the market authorisation status in the different
countries and any significant variations of the marketing authorisation due to safety
issues.
The PSUR, regulated in Directive 2001/83/EC, was modified by Directive
2004/27/EC that contains the following amendments:
• The reports of all adverse reactions shall be submitted to the Competent
Authorities in the form of a PSUR, immediately upon request or at least
every six months after authorisation and until the placing on the market.
• PSURs shall also be submitted immediately upon request or at least every
six months during the first two years following the initial placing on the mar-
ket and once a year for the following two years. Thereafter, the reports shall
be submitted at three-yearly intervals, or immediately upon request.
• Finally they shall include a scientific evaluation of the risk-benefit balance of
the medicinal product.
Other sources of relevant pharmacovigilance data

• Relevant medical literature due to the obligation to screen it weekly
(Rosenberger and Schaefer 2003);
• Data sources specified in the framework of Article 8 para. 3 lit. ia of Direc-
tive 2001/83/EC requested pharmacovigilance system or risk management
system;
• Pharmacovigilance data generated in specific investigations requested by
the EMEA according to Article 26 para. 3 of Regulation (EC) No
726/2004.
Required staff
Required by Directive 2001/83/EC the MAH has to name permanently and continu-
ously a qualified person experienced in pharmacovigilance himself or that is ad-
vised by medical experts that execute the activities of the MAH being relevant for
the pharmacovigilance by respecting the given time limits. The qualified person
shall reside in the Community.
Figure 3.17. The responsibilities of the qualified person
THE RESPONSIBILITIES OF THE QUALIFIED PERSON
• establishment and maintenance of a system which ensures that informa-

tion about all suspected adverse reactions which are reported to the per-
sonnel of the company, and to medical representatives, is collected and
collated in order to be accessible at least at one point within the Commu-
nity;
• preparation for the Competent Authorities of the reports of all suspected
adverse reactions occurring either in the Community or in a third country
• ensuring that any request from the Competent Authorities for the provi-
sion of additional information is answered fully and promptly;
• provision of any other information to the Competent Authorities relevant
to the evaluation of the benefits and risks afforded by a medicinal prod-
uct, including appropriate information on PASS.
IT-infrastructure
The MAH has to guarantee that for the electronic transmission of ICSRs and
PSURs the following guidelines and specification will be respected: ICH Guidelines
E2A / E2B / E2C / M1 / M2.
Deadline for establishment and functioning of the system of electronic transmission
of ICSRs of centrally authorised medicinal products is 20 November, 2005.
3.1.4.2.5 The healthcare professionals
Primary reporting of spontaneous adverse reaction reports

Healthcare Professionals send spontaneously adverse drug reaction reports to the
MAH or to the Authorities. Patients are encouraged to report suspected ADRs to
their healthcare professional. The Member States are required to take measures to
support this procedure and in particular the reporting to the Competent Authorities
(Article 101 of Directive 2001/83/EC).
In particular, Authorities should communicate the importance of these reports to-
wards the HCPs, implement a user friendly communication system with acknowl-
edgment messages and feed back and inform regularly or in case of emergency on
safety issues.
Communication of authorities with healthcare professionals

Safety issues of a medicinal product that up to now were not known and not de-
scribed in the SPC or changes in the marketing authorisations have to be commu-
nicated to the HCPs or directly to the public. For this the Authorities have the fol-
lowing possibilities: they can change the product information, forward information
on adverse reactions in official bulletins / newsletters or they can initiate a so-called
Dear-Doctor-Letter (that is mostly prepared by the MAH). In some exceptional
cases co-ordinated press releases may be necessary of safety issues, information

may also be reported in the media.
For drugs that are subject to the European pharmaceutical legislation it is neces-
sary that the information to HCPs or the public is carefully coordinated in time and
content, since there will be always several Member States affected (Eudralex Vol-
ume 9, Pharmacovigilance Guideline, No. 3 2005).
3.1.4.3 The detection of new or changing ‘signals’ of medicines safety

issues
All safety relevant data and in particular all spontaneous reports are to be
searched at regular intervals in order to detect signals, i.e. up to now unknown rela-
tionships between a medicinal product and an adverse drug reaction. This can be
done by checking in a qualitative way the reports in a case by case analysis of the
reports by trained personal or analysing automatically quantitative effects of reports
that are stored in an electronic data base, evaluation of PASS and clinical trials
and screening of published sources and literature (Waller 2004).
3.1.4.3.1 Identification of possible signals

• The following parties are responsible for signal detection (Eudralex Volume
9, Pharmacovigilance Guideline, No. 4 2004; Arlett 2001): MAH: Signals
arising of its own products
• NCAs: Signals arising from information in their territory
• Reference Member State: Signals arising from information about decen-
trally authorised products under its observation
• Rapporteur: Signals arising from information about centrally authorised
products under its observation
• Agency secretariat: Signals arising from information about centrally author-
ised products in agreement with the rapporteur
3.1.4.3.2 Communication about detected signals

Competent Authorities and the MAH should inform each other about identified sig-
nals, which may impact the risk-benefit-profile of a medicinal product (Eudralex
Volume 9, Pharmacovigilance Guideline, No. 5 2004).
3.1.4.4 Assessment and decision making with regard to safety issues
The MAH is responsible for the evaluation of safety issues and subsequent deci-
sions concerning his own products.
Due to the legal obligation to monitor and control the authorisation of medicinal
products the Competent Authority, which has a leading function in the processing
of the pharmacovigilance relevant activities, i.e. the reference Member State and
the rapporteur, unless otherwise decided, is responsible for the assessment of
safety issues arising on signals, PSURs and otherwise reports and for the prepara-
tion of assessment reports (Eudralex Volume 9, Pharmacovigilance Guideline, No.
6 2004; EMEA 2004b; Moseley 2004).
3.1.4.4.1 Decentrally authorised products

Responsible for the assessment of safety issues is the reference Member State,
which also decides about the need of additional scientific advice, given by the
PhVWP. The PhVWP makes non-binding recommendations about regulatory ac-
tions. In agreement with the concerned Member State the reference Member State
prepares an assessment report with recommendation of regulatory actions. The
Competent Authorities are responsible for the implementation of the recommended
regulatory action in their MS.
In the following situations the CHMP must be involved in the assessment of safety
issues by the reference Member State:
• in the conduction of referrals according to Articles 31, 36, 37 of Directive
2001/83/EC;
• consideration to suspend, revoke or vary the marketing authorisation (Arti-
cle 107 para. 2 subpara. 2 of Directive 2001/83/EC);
• suspension of the marketing authorisation in the case of an urgent action to
protect public health (Article 107 para. 2 subpara. 2 of Directive
2001/83/EC);
• optional on request of a MS during variation procedure (Article 107 para. 2
subpara. 3 of Directive 2001/83/EC).
The CHMP adopts opinions with recommendations of regulatory actions, the Euro-
pean Commission formulates decisions on these. The competent authorities of the
MS are responsible for the implementation of the actions following the decision.
3.1.4.4.2 Centrally authorised products

The rapporteur is responsible for the assessment (including the preparation of re-
ports) of safety issues concerning centrally authorised products. He also decides
about the need of additional scientific advice by the PhVWP. The PhVWP gives
non-binding recommendation for regulatory actions. The rapporteur refers assess-
ment reports and recommendations to the CHMP. The CHMP adopts opinions with
regulatory actions in the case of safety issues that require changes of the market-
ing authorisation concerning centrally authorised products (Article 5 para. 2 of
Regulation (EC) No 726/2004). The European Commission formulates the decision
which is then binding on the Member States.
3.1.4.5 Action (including regulatory action) to protect public health
In general action can be caused compulsorily by NCAs or voluntarily by MAHs in

accordance with the legislation.
3.1.4.5.1 Regulatory actions by NCAs

The following safety issues can trigger major regulatory actions of the NCAs (Arti-
cles 116, 117 of Directive 2001/83/EC)
1. Product is harmful under the normal conditions of use;
2. Lack of therapeutic efficacy;
3. Risk benefit balance is not positive under the normal conditions of use;
4. Qualitative and quantitative composition is not as declared;

5. Particulars supporting application are incorrect or have not been amended in
accordance with Article 23 of Directive 2001/83/EC (vary due to new informa-
tion).
The following actions (sanctions) are possible (Articles 116, 117 of Directive
2001/83/EC):
• Suspension, revocation, withdrawal or variation of the marketing authorisa-
tion;
• Prohibition of the supply of the products or withdrawal from market.
For decentrally authorised products in the case that the CHMP is involved (either
on request of a NCA or according to the legal obligation (see 4.4.1) and for cen-
trally authorised products the required measures are determined by the European
Commission based on the opinion of the CHMP (Article 107 para. 2 subparas. 3
and 4 of Directive 2001/83/EC, Article 20 para. 2 of Regulation (EC) No 726/2004).
Procedures to vary the marketing authorisation are laid down in Commission Regu-
lation (EC) No 1084/2003 concerning products granted by a NCA and in Regula-
tion (EC) No 1085/2003 concerning products falling within the scope of Regulation
(EC) No 726/2004.
3.1.4.5.2 Urgent action to protect public health

Directive 2001/83/EC and Regulation (EC) No 726/2004 allow provisional meas-
ures of urgent regulatory actions to protect public health (Article 107 para. 2 of Di-
rective 2001/20/EC and Article 20 para. 4 of Regulation (EC) No 726/2004) with
rules of communication and specification of the procedure (referral).
Further actions of the authorities can be
• to conduct pharmacovigilance inspections at the MAH or a substitute (see
below);
• to impose penalties (Article 104 para. 9 subpara. 3 of Directive 2001/83/EC,
Articles 24 para. 5, 84 para. 1 of Regulation (EC) No 726/2004);
• to change the prescriptions status of a product in case of new facts (acc. to
Article 74 of Directive 2001/83/EC).
3.1.4.5.3 Action by the MAH

In case of safety issues the MAH may apply for modification of an existing authori-
sation (acc. to Regulations (EC) No 1084/2003 and 1085/2003) during the next
routine variation, non urgently or within the scope of an urgent safety restriction. In
addition, the MAH may withdraw the product.
3.1.4.6 Communication / transparency with stakeholders
3.1.4.6.1 Communication obligations

• The MAH is obliged to timely inform the Competent Authorities in case of new
information about the product, which result in changes of the authorisation
documents, changes of pharmacological and toxicological documents and
changes of SPC (Article 16 para. 2 of Regulation (EC) No 726/2004).
• The MS are obliged to inform the public in case of urgent actions taken to pro-
tect public health (Article 20 para. 5 of Regulation (EC) No 726/2004).
• The EMEA is authorised to request at any time data by the MAH, which docu-
ment that the benefit-risk-ratio remains positive (Article 16 para. 2 of Regulation
(EC) No 726/2004).
• The EMEA is obliged to disseminate pharmacovigilance information (Article 57
para. 1 (f) of Regulation (EC) No 726/2004).
3.1.4.6.2 Transparency of communication

The following measures aim to ensure a transparent communication between the
involved stakeholders:
Obligation of the MAH:
• The MAH should inform in a timely manner the Competent Authority in case
of publication of information regarding pharmacovigilance (Article 104 para.
9 of Directive 2001/93/EC, Article 24 para. 5 of Regulation (EC) No
726/2004).
• The MAH is obliged to provide the Competent Authorities with information
regarding his products, especially when these data can result in a change
of the benefit-risk-assessment (Article 23 of Directive 2001/83/EC, Article
16 para. 2 of Regulation (EC) No726/2004).
• MAH has to provide upon request of the Authorities data relating to the vol-
ume of sales and the volume of prescriptions (Article 23 a of Directive
2001/83/EC).
Access of the public to pharmacovigilance relevant data:
• EudraVigilance database has to be made accessible to the public (Article
102 of Directive 2001/83/EC).
• Pharmacovigilance relevant opinions of the CHMP have to be made acces-
sible to the public (Article 22 of Regulation (EC) No 726/2004).
• Decisions about granting or revocation of a marketing authorisation have to
be made accessible to the public (Article 125 of Directive 2001/83/EC).
• Internal procedures, agenda and minutes of the Competent Authorities of
the Member States have to be made accessible to the public (Article 126 of
Directive 2001/83/EC
• Annually a list with withdrawn medicinal products is published by the Euro-
pean Commission (Article 123 para. 4 of Directive 2001/83/EC).
3.1.4.7 Audit, both of the outcomes of actions taken and of the key proc-
esses involved
3.1.4.7.1 MAH – Pharmacovigilance inspections by the authorities

To ensure that MAH comply with pharmacovigilance regulatory obligations Member
States’ Authorities or the Commission can conduct or initiate pharmacovigilance
inspections at random and systematic as well as targeted to MAHs suspected of
being non-compliant; various options for actions (as a result of the inspection) can
follow and will be judged on a case-by-case basis reaching from education to
prosecution (EMEA 2001). The legal basis is laid down in Article 111 of Directive
2001/83/EC and Article 19 of Regulation (EC) No 726/2004.
The legislation appoints as main subjects of the inspections of the pharmacovigi-
lance system the qualified person and reporting (Article 111 of Directive
2001/83/EC).
3.1.5 Conclusions
The basis of the project „Assessment of the European Community System of
Pharmacovigilance” is the description of the rules according to which pharma-
covigilance of medicinal products is currently performed in the EU.
The legal framework of pharmacovigilance in the EU is essentially formed by
1. Regulation (EC) No 726/2004, appropriate for centrally authorised medici-
nal products,
2. Directive 2001/83/EC, appropriate for medicinal products that are author-
ised in more than one Member State through the “Mutual Recognition Pro-
cedure” or the “Decentralised Procedure”
3. the national pharmaceutical legislation of the Member States.
Additional laws are Regulation (EC) No 1084/2003 (change of the marketing au-
thorisation of products authorised with the procedure of mutual recognition), Regu-
lation (EC) No 1085/03 (change of the marketing authorisation of centrally author-
ised products), and Regulation (EC) No 540/95 (non serious adverse drug reac-
tions of centrally authorised products). The legal framework is completed by the
guidance document Volume 9 that is associated to the European law and refer-
ences itself to internationally accepted standards of the ICH Guidelines (ICH-E2A,
E2B, E2C, E2D, E2E, M1, M2).
These laws regulate the essential processes of pharmacovigilance such as data
collection and data management, safety signal detection, safety issue assessment,
decision making, action taken to protect public health and communication with
stakeholders.
The analysis above has shown that the current European Pharmacovigilance Sys-
tem has achieved an advanced state of development. This is especially true after
the implementation of the recent reform. From November 2005 onwards, Authori-
ties are given additional tools for monitoring the safety of medicines, as well as
greater scope for urgent regulatory action once the benefit/risk balance of a me-
dicinal product becomes unfavourable. The new provisions also include increased
transparency on safety issues and facilitate communication, with the provision of

timely and targeted information to Healthcare Professionals and the public7.
This reform is partly inspired by a worldwide discussion of several expert groups
who have identified the need to strengthen pharmacovigilance systems especially
in the aftermath of the Cerivastatin redrawal. There are current initiatives by the
Council for International Organizations of Medical Sciences (CIOMS) and the Inter-
national Conference on Harmonisation (ICH). The main emphasis of both concepts
is to continuously monitor the benefit-risk-profile of a medicinal product as it goes
through its life cycle. Their objective is to change the philosophy towards an earlier
and more proactive approach that will start before a medicine reaches the market.
This leads to a broader concept of pharmacovigilance towards risk management. In
parallel with these activities the regulatory agencies in the US and the EU are con-
centrating their efforts in formulating risk management strategies (Tsintis 2004).
More recently, the CIOMS Working Group stressed the need not only to incorpo-
rate newer approaches for managing of safety information from clinical trials, but
also to adapt the methods and tools used in post-approval pharmacovigilance to
the early and late stages of pre-approval development of medicinal products
(CIOMS 2005).
As regards the concept of a European Risk Management Strategy (ERMS) two
comprehensive key documents were published in spring 2005 as a result of a col-
laboration between the Heads of the National Medicines Agencies and the EMEA
(Heads of Medicines Agencies 2005b). When considering such a strategy, the spe-
cial interests of patients as regards pharmacovigilance must not be neglected at
any rate. This, at least, is one of the core messages of the EMEA/CPMP Working
Group with Patients Organisations (EMEA 2004a).
Basis for a further optimisation of the Community system of pharmacovigilance is,
however, not only the strengthening of the existing legislation with respect to the
implementation of an ERMS but also the full implementation of all legal rules and
guidelines in all Member States. Resulting from the complex legal structures of the
EU, local deviations from the rules and guidelines in the practical implementation of
pharmacovigilance in the Member States have to be assumed due to national con-
ditions and due to the necessary implementation of European Directives into na-
tional law.
By analysing potential deviations of the given rules and guidelines, it is to be
checked, if the taking into account of local peculiarities, apart from the full imple-
mentation of all legal rules and guidelines in the Member States, might further
strengthen the Community system of pharmacovigilance.
Beyond this the system may be further optimised, if all involved European stake-
holders use the existing instruments for coordination and cooperation and in par-
ticular openly and promptly communicate in consideration of the legal obligations.
7 The written agency survey revealed that 60% of the agencies believe that the
new legislation will generally improve the system, 2 agencies even believe in a
strong improvement.
3.2 Implementation of the European regulatory framework

into practice
The functioning of the European system of pharmacovigilance for CAPs is illus-
trated in the following figure.
Figure 3.18. Organisation of pharmacovigilance regarding centrally authorised

medicinal products
Member States
-overall responsibility for
PhV on their territories
and the appropriate
info safety legislation info safety
data -operate a PhV system data
NCAs
info actions communication
(with one leading safety data
NCA as the rapporteur)
EC
-overall responsibility for
info safety EU-system of PhV,
regulatory
data including policy and laws
actions
-competent authority
HCPs
-reporting of ADRs EMEA
communication
-responsibility for
safety data
coordinating PhV
resources and work of the
safety MS
relevant
information MAHs
-overall PhV responsibility for
their own products
info safety -establish/ maintain a
data voluntary actions in
coordination with
the authorities
Source: KKS-UKT/Fraunhofer ISI 2005
Figure 3.19 illustrates the main actors and relationships in pharmacovigilance for
non-centrally authorised human drugs.
The agency survey revealed that the new requirements that are binding from No-
vember 2005 on are not yet implemented in 14 agencies. Most of the agencies
plan to implement the new legislation in time; two anticipate delays until 01-JAN-
2006 and 01-JUN-2006, respectively.
60% of the agencies believe that the new legislation will generally improve the sys-
tem, 2 agencies even believe in a strong improvement.
Figure 3.19. Organisation of pharmacovigilance regarding non-centrally authorised

medicinal products
Member States
info safety
-overall responsibility
data
for PhV on their
territories and the
In case of arbitration
appropriate legislation
between the MS
-operate a PhV system
info safety data
NCAs CHMP/PhVWP
info (with one leading
actions NCA of the RMS)
-competent authorities
EC
binding -overall responsibility for
info safety decisions EU-system of PhV,
data including policy and laws
communi-
cation
safety
data EMEA
HCPs
-reporting of ADRs -responsibility for
regulatory coordinating PhV
action resources and work of the
safety MS
relevant
information MAHs
-overall PhV responsibility for
their own products
info safety -establish/ maintain a PhV
data voluntary actions in
system
coordination with
the authorities
Source: KKS-UKT/Fraunhofer ISI 2005
3.3 Systems of pharmacovigilance in other countries

3.3.1 Pharmacovigilance systems in the USA, Japan, and Canada
3.3.1.1 USA
3.3.1.1.1 Legal framework

The Keauver-Harris Amendments (‘1962 Amendments’) to the Federal Food,
Drugs and Cosmetics Acts of 1938 provides the legal basis for drug regulation in-
cluding the regulations for pharmacovigilance. The amendments do not mandate
post-marketing surveillance, but they empower the Food and Drug Administration
(FDA) to approve a New Drug Application (NDA) under the condition to conduct
post-marketing clinical studies to demonstrate further a product’s safety (Arnold
2004).
Safety reporting requirements that oblige a manufacturer to report suspected Ad-
verse Drug Reactions (ADR) to the FDA for all products sold or developed in the
USA are specified in Title 21 of the Code of Federal Regulations (CFR) (Arnold
2004).
The USA, represented by the FDA, are a participant of the International Confer-
ence on Harmonisation of Technical Requirements (ICH), so the regulatory agency
has adopted almost invariably all ICH-Guidelines which are developed since 1991
(Abraham J. 2004).
3.3.1.1.2 Actors
Main actors of pharmacovigilance in the USA are the government drug regulatory
agency, the manufacturer/industry, healthcare professionals (HCPs) and consum-
ers.
The FDA is the US drug regulatory authority. Two departments of the FDA, the
Center for Drug Evaluation and Research (CDER) and the Center for Biological
Evaluations and Research (CBER) are responsible for assuring the safety and effi-
cacy of all drugs developed or marketed in the USA.
The FDA Office of Drug Safety is responsible for post-marketing ADR reporting of
non-biological products and operates the Adverse Event Reporting System (AERS)
database for post-marketing pharmacovigilance (Arnold 2004).
The FDA has installed a strong advisory committee system that complements the
Agency’s scientific expertise. The committees give advice, the FDA is not bounded
to follow it. The advisory committees give credibility to the FDA decision-making
processes by having public discussion of controversial topics by experts, agency’s
staff, industry and consumers (Sherman L.A. 2004).
US industry is obliged to establish and maintain records, to report to the FDA all
serious, unexpected Adverse Drug Experience (ADE) associated with the use of
their products, and to develop written procedures for the surveillance, receipt,
evaluation and reporting of post-marketing ADE (Title 21 of the Code of Federal
Regulation 310.305 - Records and reports concerning adverse drug experiences
on marketed prescription drugs for human use without approved new drug applica-
tions. Revised as of April 1, 2004 2004, Title 21 of the Code of Federal Regulation
314.80 – Post-marketing reporting of adverse drug experiences 2004).
3.3.1.1.3 Post-marketing surveillance activities
Spontaneous reporting system

Database: The FDA operates the Adverse Event Reporting System (AERS) com-
puterized database since 1969 and the Vaccine Adverse Event Reporting System
(VAERS). In 2004 the AERS database had stored 3 million reports (Goetsch R.
2005).
Used definitions: Definitions are laid down in the 21 CFR and are in general con-
sistent with the corresponding ICH definition. The term ADE is used within the USA
rather than Adverse Event (AE) or ADR. An ADE is defined as any AE associated
with the use of a drug whether or not to be product related. It includes spontaneous
reports (Arnold 2004).
Reporting: HCPs and consumers report ADE voluntarily to the FDA (10% of all
reports to the FDA; Goetsch R. 2005). They can report electronically using the
online access of the MedWatch program (MedWatch 2003, MedWatch 2005b).
Industry is legally obliged to report ADEs (MedWatch 2005a), which make more
than 90% of all reports (Goetsch R. 2005). They can submit ADR electronically
(MedWatch 2005a).
Public access: The AERS collects information about adverse events, medication
errors and product problems that occur after the administration of approved drug
and therapeutic biologic products. Quarterly (non cumulative) data files since
January 2004 are online available (AERS 2005a).
Figure 3.20. Expedited reporting requirements in the USA
Report Origin ADR that qualify for expedited

reporting to the regulators
• 15 day alert report • Regardless of the • Serious and unexpected

source (domestic ADE
• 15 day alert report
cases, foreign
follow-up (if the ini-
cases, scientific lit-
tial report is incom-
erature
plete)
• 15 day alert report • Study reports, solic- • Serious and unexpected
ited information ADR (reasonable possibil-
• 15 day alert report
ity of a causal relationship),
follow-up (if the ini-
if minimum criteria acc. to
tial report is incom-
ICH E2A are obtainable.
plete)
Source: Arnold 2004, Title 21 of the Code of Federal Regulation 310.305 - Records and reports con-
cerning adverse drug experiences on marketed prescription drugs for human use without approved
new drug applications. Revised as of April 1, 2004 2004, Title 21 of the Code of Federal Regulation
314.80 – Post-marketing reporting of adverse drug experiences 2004
Periodic safety reporting by industry

The periodic reports should contain the reporting form (ADE, ADR) and a line list-
ing of the spontaneous reports, narrative summaries and analysis and narrative
discussion of actions. FDA accepts periodic reports in accordance with ICH E2C
format and content, if the applicant has secured a waiver from the FDA (Arnold
2004).
Periodicity (Title 21 of the Code of Federal Regulation 314.80 – Post-marketing
reporting of adverse drug experiences 2004): Submission is required quarterly dur-
ing the first 3 years after marketing approval, thereafter annually.
Figure 3.21. Reports qualifying for inclusion in a US periodic report
Source Type of reports qualifying for inclusion in a periodic

report
• Spontaneous reports • All domestic ADEs (serious expected, non-serious)

• Foreign serious unexpected ADEs
• Domestic reports of “lack of efficacy”
• Published literature • Only serious unexpected ADEs
• Clinical studies and • Only serious unexpected ADRs

post-marketing studies
Source: Arnold 2004
Conditional approval
FDA can approve a NDA under the condition to conduct post-marketing clinical
studies to demonstrate further the product’s safety (Title 21 of the Code of Federal
Regulation: 310.303 - Continuation of long-term studies, records, and reports on
certain drugs for which new drug applications have been approved 2004).
Post-marketing study commitments

Post-marketing study commitments are studies, required of or agreed to by a spon-
sor, that are conducted after the FDA has approved a product for marketing.
Agreement with the sponsor to conduct a study can be reached either before or
after FDA has granted approval to a sponsor to market a product. The studies are
used to gain new data about the safety, efficacy or optimal use of a drug. The
sponsor is obliged to provide an annual report to the FDA on the status of the study
until it is completed or terminated. The FDA is responsible to annually report in the
Federal Register on the performance of post-marketing commitment studies (AERS
2005b).
3.3.1.2 Japan
3.3.1.2.1 Legal framework

The primary Japanese law governing drug affairs is the Pharmaceutical Affairs Law
(PAL) which is the legal basis of pharmacovigilance requirements. The PAL is sup-
plemented by the following post-marketing provisions issued by the Ministry of
Health, Labour and Welfare (MHLW; Arnold 2004):
• Standard for implementation of post-marketing surveillance (PMS) for the re-

examination application of new drugs (1993);
• Standard for the conduct of Good Post-Marketing Surveillance Practice

(GPMSP) (1993) and MHLW Ordinance No.10 (GPMSP) (1997);
• Ordinance No.29 – Enforcement of the Pharmaceutical Affairs law, Article 66-

7 (1997);
• Notification No. 1324 – Implementation of early post-marketing phase vigi-

lance (2001);
• Notification IYACUAN No. 0531001 – Electronic reporting (2002).

Japan, represented by the government drug regulatory agency, is a participant of
the ICH, so the regulatory agency has adopted almost invariably all ICH-Guidelines
which are developed since 1991 (Abraham J. 2004).
3.3.1.2.2 Actors
Main actors are the government drug regulatory agency, the industry, HCPs and
consumers.
Since April 2004 the Pharmaceutical and Medical Device Agency (PMDA) with a
safety division is regulatory responsible for the handling of affairs concerning
safety. An independent expert advisory committee exists which is targeted to re-
view reports of reactions that may warrant labelling changes (McEwen J. 2004).
To fulfil the requirements for post-marketing surveillance Japanese companies es-
tablish a PMS Management Department with sufficient qualified staff that is inde-
pendent from the sales/marketing department. They must appoint a “Responsible
Person” for PMS management and they have to prepare and to comply with rele-
vant standard operating procedures (Arnold 2004).
1. Spontaneous reporting system:

Used definitions: Japanese definitions are in general in accordance with those
specified in the ICH guidelines. MHLW excluded the “medically important” criterion
from the definition of a serious AE and slightly changed “disability” to “any disable-
ment that is a permanent dysfunction that causes a disturbance in daily life” (Arnold
2004).
Reporting: Main source of adverse drug reactions’ report is the industry that is
obliged to report occurred serious ADR. In the year 2002, 24221 ADR reports were
transmitted by the industry, and 4195 ADR reports by Healthcare professionals
(HCPs). There is no consumer reporting (McEwen J. 2004).
Electronic reporting is mandatory for industry for expedited reports since October
2003, electronic reporting is not available for HCPs (Arnold 2004, McEwen J.
2004).
Public access: Japanese ADR reports are available on the internet and can be
accessed by researchers outside the national centre (Kubota and Koide 2004).
Figure 3.22. Expedited reporting requirements in Japan
Report Origin ADR that qualify for expedited reporting
• 15 day • Domestic • Serious and unexpected

(Fatal unexpected ADR: immediate notification
to be followed by full written report
• Foreign • Serious and unexpected
• Scientific lit- • Serious
erature
• 30-day • Domestic • Serious and expected
• Severe/moderate non-serious
• ADRs with an increased frequency, lack of
efficacy, possibility of an association with
the onset of cancer
Expedited Abroad • Measures taken abroad that relate to safety

issues
Source: Arnold 2004
Periodic safety reporting by the industry

The Periodic Safety Update Report (PSUR) is to be submitted for all drugs mar-
keted in Japan prepared in full accordance to the ICH E2C guideline, including for-
eign data. The PSUR must summarise the progress of all Japanese post-marketing
studies (Arnold 2004).
The PSURs have to be submitted every 6 months for 2 years following approval of
the Japanese NDA, then annually during the defined re-examination period, and
finally every 5 years after the re-examination period is finished (Arnold 2004).
Drug Re-examination System

To overcome the known limiting factors of pre-approval clinical studies (limited
number of involved patient, a selected population) Japan has instituted in 1979 the
Drug-Re-examination System for reassessing the safety and efficacy of a drug af-
ter its first approval. During the re-examination time the initial approval is only pro-
visional and must be reviewed subsequently considering the efficacy and safety of
the product at a specified future time point that depends upon the nature of the
drug (after 4 years for supplemental NDA, 6 years for most drugs, when they are
approved for the first time in Japan, and 10 years for orphan drugs). Practically
efficacy and safety of a drug are scrutinised by comparing the data collected during
post-marketing surveillance activities with the data submitted at the time of ap-
proval of the drug. The outcome of the second review can be the cancellation or
modification of the initial approval or no action (Fujiwara and Kobayashi 2002,
Arnold 2004).
Post-marketing surveillance activities (Arnold 2004)
• Early Post-marketing Phase Vigilance
o during the first 6 months after entry of a new product to the Japa-
nese market;
o ensures that information has been provided to the prescribers and

encouragement of caution;
o improvement of appropriate use;
o reporting promptly spontaneous information on serious ADRs, to

implement consequent safety measurements and minimize risk for
the public health;
• Clinical Experience Investigation studies
o Detection of unlabelled ADRs;
o Understanding of ADR development during actual use;
o Definition of factors suspected to influence the product’s safety

and/or efficacy profile;
• Special studies and post-marketing clinical studies
o E.g. long-term use, special populations, pharmacoepidemiological

studies with mortality outcome, pharmacokinetic studies with pa-
tients with renal failure.
3.3.1.3 Canada
3.3.1.3.1 Legal Framework
Basic Law
In Canada legislative requirements for the post–marketing surveillance of health
products are covered by the Food and Drugs Act and Regulations (Health Canada
2005a).
Canada is not a participant of the ICH, but one of its non-voting observers. Some of
the ICH-guidelines are adopted (Abraham J. 2004).
3.3.1.3.2 Actors
Main actors are the government drug regulatory agency, the industry, HCPs and
consumers.
The Marketed Health Product Directorate (MHPD) of Health Canada (the Canadian
Federal Department) is the regulatory office that is responsible for post-marketing
safety, concerning all regulated marketed health products in Canada (MHPD

2005).
There are established a number of expert advisory committees and public advisory
committees to increase transparency and public involvement (Health Canada
2005d).
Spontaneous reporting system

Database: Health Canada operates the Canadian Adverse Drug Reaction Informa-
tion System (CADRIS). January 2004 CADRIS contains over 160000 suspected
ADR reports that have occurred in Canada since 1965 (Health Canada 2005c).
Used definitions: The definitions for adverse reactions harmonize with the defini-
tions of the World Health Organisation, Council for International Organizations of
Medical Sciences (CIOMS), ICH and FDA (Health Canada 2003).
Reporting: ADR reporting applies to all drug products sold in Canada.
HCPs and consumers report voluntarily adverse drug reaction, at the moment to 7
regional ADR centres and 1 national ADR centre or to the manufacturer. Electronic
reporting will be possible in fall 2005 (Health Canada 2005b).
Manufacturers are legally obliged to report adverse drug reaction to the regulatory
authority (Health Canada 2003).
Public access: The Canadian Adverse Drug Reaction Monitoring Program
(CADRMP) Online Query and Data Extract utility provides the public with informa-
tion about suspected adverse reactions of marketed products occurring in Canada
(Health Canada 2004a).
Figure 3.23. Expedited reporting requirements in Canada
Report Origin ADR that qualify for expedited report-

ing
• 15 days • Domestic cases • Serious

• Including literature • Lack of efficacy (only new drugs)
• Foreign cases • Serious and unexpected
• Including literature
• Domestic study reports • Serious
• Lack of efficacy (only new drugs)
• Foreign study reports • Serious and unexpected
Source: Health Canada 2003

Periodic reports by the industry

On an annual basis and whenever requested the manufacturer is obliged to pre-
pare a summary report about the reports received during the last twelve months.
The summary consists of three sections: a summary line listing of ADRs, a critical
analysis of the reports and recommended actions (Health Canada 2003).
Figure 3.24. Reports qualifying for inclusion in a Canadian summary report
Source Type of reports qualifying for inclusion in a periodic report
• Spontaneous re- • Domestic ADRs (serious, non-serious and unexpected,

ports and literature lack of efficacy of new drugs)
• Foreign ADRs (serious, non-serious)
• Studies • Domestic ADR (serious and unexpected, lack of effi-
cacy of new drugs)
• Foreign ADR ( serious and unexpected)
Source: Health Canada 2003
Additional post-marketing surveillance activities (Health Canada 2004b)
• Post-marketing studies conducted by the manufacturer or health care insti-

tutions
• Publications in scientific journals
• Collaboration with patient group, academic institutions, professional asso-

ciations in Canada and internationally
• Risk communication from regulatory agencies in other countries
3.3.1.4 Summary
The comparison of the pharmacovigilance systems in the USA, Japan and Canada
has pointed out that the organisation of this surveillance is basically the same in
these three countries.
Main actors involved are the national drug regulatory authorities, the manufacturers
of the drugs and the HCPs. The legal requirements to collect information about
adverse drug reactions are implemented using spontaneous reporting systems
(provided primarily with reports by manufacturers and the HCPs) and periodic re-
ports of marketed drugs (prepared by the manufacturers of the products).
The authorities in all 3 countries operate databases for the collection of spontane-
ously reported adverse drug reactions. Each of these databases is accessible for
the public to retrieve information about a certain drug. In all countries the authori-
ties can get advice from expert advisory committees. Two countries, the USA and
Japan, are participants in the ICH; Canada is a observer of the group. As a result
the basic conditions of pharmacovigilance are harmonised within the three coun-
tries and with regard to other members of the ICH.
Concerning the access of drugs to the market, however, the FDA can approve
drugs under the condition to conduct post-marketing studies in order to demon-
strate further a product’s safety, while in Japan the re-examination system gives
the possibility to reassess the safety and efficacy of a drug after its primary provi-
sional approval at a specified time point depending upon the nature of the drug.
In the USA and Canada the public is integrated in the discussion about safety is-
sues through participation in public meetings of advisory committees (USA) or the
membership in a public advisory committee (Canada). Consumers are invited to
report adverse drug reactions directly to the authorities; in the USA this can already
be done electronically. The frequency to prepare periodic safety reports differs be-
tween the three countries. In the USA, the reports have to be submitted quarterly in
the first three years following the approval; in Japan this has to be done every 6
months for 2 years following approval of a Japanese NDA and then annually during
defined re-examination period; in Canada reports are required to be prepared an-
nually or on request.
The comparison of the pharmacovigilance related activities in the USA, Japan and
Canada with the European system shows that definitions and processes accord to
some extent. On the other hand it seems worthwhile to explore in more detail the
specific conditions in the USA, Japan and Canada, e.g. the conditional approval or
publicly accessible ADR-databases, in order to be able to asses whether and how
such provisions could contribute to strengthening the European system.
3.3.2 Evaluation of foreign systems by interview partners

In the following we present the most frequent answers on the interview question
(number 11), "What could we learn from other systems in 3rd countries?" In gen-
eral, the interviewees here referred to non-European countries, most frequently the
systems in Australia and New Zealand were mentioned as having particular
strengths. If statements were made about European countries they are presented
in chapter 8.2, together with other examples of best practice from the interviews.
3.3.2.1 General factors
Legal framework conditions

One of the interviewees from agencies assessed the regulation in USA and Japan
as less requiring than in Europe.
Expertise
Australia and New Zealand were said to have particularly efficient ADR advisory
boards (with clear responsibilities, detailed preparation by the authority, exchange
with MA department, concentration on "drugs of current interest" with intensive
monitoring programmes and prescription event monitoring especially for new
classes of drugs.
The FDA has well developed expertise in pharmacoepidemiology internally. The
good information and respective courses in epidemiology and individual training
offered by the WHO-UMC were equally appreciated. The training is free of charge.
In Canada the competent authority (HealthCanada) is effective because it has all

products in one hand (medicines, dietary supplements…). Australia is said to have
good information on the market.
3.3.2.2 Data collection
The intensive monitoring programme/ drugs of current interest/Prescription event

monitoring of Australia and New Zealand are mentioned as good practice by sev-
eral interviewees. To detect signals, several interviewees find that the practice of
the UMC to use all reports, and not only reports on severe ADRs, is helpful. The
Australian system is said to offer good information on the market. Single interview-
ees found that the system in the USA is strong because of the large amount of
data, but weak because of its low data quality. Another interviewee gave an inter-
esting example from Taiwan where HCPs at a certain time got presents for report-
ing which led to over-reporting and equally reduced quality of safety data.
Australia, New Zealand, and Canada have strengths in product-related PASSs.
3.3.2.3 Safety issue assessment
Co-operation
One interviewee from an agency requested that the WHO ADR monitoring pro-
gramme used in 3rd countries should be harmonized with the European system to
improve the opportunity of co-operation.
Access to external experts

Several interviewees emphasized the good work of the ADR advisory boards in the
Australian system.
3.3.2.4 Decision-making
According to a single opinion, the system in the USA comes to relatively quick de-
cisions.
3.3.2.5 Communication and action to protect public health
In general, much good information is seen by the interviewed agencies coming

from Australia, New Zealand, Canada, and the WHO. Several interviewees praised
the good quality of information on the websites of the Canadian and US agencies.
More publications of changes e.g. in SPCs than in Europe are identified by a num-
ber of interviewees in the USA and in Australia, in the latter country the quality of
the ADR bulletin as well as the publication of informative protocols of expert meet-
ings were also mentioned by several interviewees. The USA have as an advantage
a high transparency, e.g. by stakeholder discussions with public representation. A
good example to communicate ADRs to HCPs is the electronic ePocrates-system
that informs HCPs on all aspects of drugs on an electronic formulary and uses this
platform to give safety-relevant information too.
HealthCanada offers a good website including safety data that are published for
public use in the Canadian Adverse Drug Reaction Monitoring Program (CADRMP)
database.
3.4 Other framework conditions

A mean of 5038 products per country had a life national marketing authorisation
(NAPs) in 2004, with a range from 236 to 13,678 products. 731.4 products (MRPs)
per country (range 34 to 2207 products) had an authorisation under the Mutual
Recognition procedure. The number of Centrally Authorised Products (CAPs) is
equal for all countries. The figure of 269 CAPs authorised in 2004 was given by
EMEA based on data from DG ENTR.
Table 3.1. Number of approvals for NMEs per country in 2003 and 2004
Number of approvals for NMEs per country
2003 Minimum Maximum Median
National approvals 0 42.00 1.00
MR procedure 0 68.00 5.50
MR with country as RMS 0 6.00 .00
Centralised with country as rapporteur 0 21.00 .50
2004 Minimum Maximum Median
National approvals 0 45.00 .00
MR procedure 0 75.00 3.00
MR with country as RMS 0 14.00 .00
Centralised with country as rapporteur 0 36.00 1.00
Table 3.2 compares the values of the own survey with those collected in the ERMS
Surveys of 2002 and 2004.
Table 3.2. Time acting for the Community as Rapporteur

Old MS in 2002 ERMS New MS in ERMS Own survey data for
[%] 2004 [%] 2004 [%]
Minimum 0.00 0.00 0.00
Maximum 35.00 0.00 38.00
The share of work that is done by the agencies for the Community as Rapporteur
varies significantly. The same is true for the time that is spent for the Community as
Reference Member State (RMS) for a MRP (Table 3.3).
Table 3.3. Time acting for the Community as RMS

Old MS in 2002 [%] New MS in 2004 [%] Own survey data for
2004 [%]
Minimum 20.00 0.00 0.00
Maximum 99.00 100.00 99.00
Most of the time, however, is spent in most agencies on NAPs. Exceptions are Cy-
prus, Liechtenstein, Hungary, Latvia, the Netherlands, Denmark, Sweden, Ireland,
Finland, Malta, and the UK, who spend a maximum of 30% of their work on NAPs.
The number or physicians working in a country will also influence the reporting of
ADRs.
Table 3.4. Number of physicians

Physicians outside hospitals Physicians in hospitals Physicians total
AT 17845 17443 35288
BE , , 45991
CZ 17000 16000 33000
DE-BFARM 133000 146000 279000
DE-PEI 160078 146357 306435
DK 4600 8600 13200
EE 800 , ,
EEA-28 1583000 , ,
EI , , 7500
ES , , 179033
FI 11000 7000 18000
FR 110000 64700 174700
HU 24560 14317 38877
IC 1290 660 1950
IT 310000 50000 360000
LI 55 10 65
LT 6640 6757 7258
LU 1200 , ,
MT , , 1302
NL 9849 20060 29909
NO 5000 10000 15000
PL 89000 , ,
PT 7251 20733 27984
SE 12000 18000 30000
UK8 41340 98000 139340
DE-total 160078 146000 ,
Source: Fraunhofer ISI 2005; figures given by the Agencies.
The political support for pharmacovigilance in general was assessed by the agen-
cies. The results are presented in Figure 3.25.
8 Figures are for Great Britain only, i.e. data exclude Northern Ireland.
Figure 3.25. Political support for PhV
25
20
% of agencies
15
10
0
Very good 2 3 4 Very weak
The support is assessed as very good or good (values 1 or 2) only by half of the
agencies, and as weak or very weak (values 4 or 5) by nearly one third.
The overall compliance of MAHs with the legal requirements is assessed as very
good or good by 78% of the agencies, negative or very negative assessments did
not occur at all.
3.5 Resources for pharmacovigilance

3.5.1 Budget
The financial resources differ widely between the agencies. The budget per popu-
lation indicates that these differences are not solely explicable by the size of the
countries (Table 3.5).
Table 3.5. Agency budget 2004

Minimum Maximum Mean
Agency total budget 2004 in Mio. EUR 0.13 102.67 33.68

Budget in Mio. EUR per 1000 capita 0.00 0.11 0.01

The same figures are presented in the following diagram.
Figure 3.26. Average agency budget in 2004
120 0,12
Agency total budget 2004 in Mio. EUR
Budget in Mio. EUR per 1000 capita

100 0,1
80 0,08
Agency total budget
2004 in Mio. EUR
60 0,06
Budget in Mio. EUR
per 1000 capita
40 0,04
20 0,02
0 0
Minimum Mean Maximum
The funding of the agencies is quite different. A lot of agencies are totally financed
by public funding, some are financed through a mix of public funding and fees
(mostly from MAH) and a few agencies are solely financed by fees.
3.5.2 Staff
Only a small part of the agency's resources are directly dedicated to PhV. The me-
dian proportion of PhV staff is only 5% of the total agency staff. About two-thirds of
the PhV staff is devoted to scientific tasks.
The median number of staff in the national agencies is Md=7.13 FTE over all agen-
cies (small and large countries). The relationship of PhV staff for administrative
tasks to scientific staff lies between 0.00 administrators per scientist (that means
no administrative staff at all) and 5 administrative staff per scientist, with an aver-
age of Md=0.37 administrative staff per scientist.
Table 3.6. PhV staff in national agencies
Median Minimum Maximum
Staff total Agency [FTE] 170.00 1.40 911.00

PhV staff total [FTE] 7.13 0.10 60.00
Share of PhV-staff of total staff 5% 1% 8%1)
PhV staff administrative [FTE] 2.50 0.00 41.00
PhV staff scientific [FTE] 4.50 0.10 43.70
Proportion of administrative PhV
37% 0% 500%
staff to scientific PhV staff
1) One country with an extreme value of 35% of PhV staff was dropped.
As expected there are differences between the countries in size of the staff, but
also in the proportion of administrative to scientific staff.
A more detailed analysis of the required staff in the different process stages shows
the highest value for "Regulatory action" (Table 3.7).
Table 3.7. PhV staff in different process stages (multiple responses possible)
Staff involved in processes [persons]
Median
Data coll./entry 3.00
Data management 2.00
Risk assessment 3.00
Regulatory action 4.00
Risk communication 2.00
Audit and QA 1.00
Monitor. compliance 1.25
The combined staff of national agencies and the regional centres can be set in re-
lation to the total population in one country. The highest value was achieved by a
small country that collaborates very intensively with a regional centre in a
neighbour country, therefore as maximum the second-highest value is given com-
ing from a country that acts independently. The EMEA is also left out of this analy-
sis, having the lowest value of 0.1 FTE per million European (i.e. EEA-28) citizens.
To account for the different size of the countries, the available staff for pharma-
covigilance (scientific plus administrative) was divided by the size of the popula-
tions (see following table).
Table 3.8. Total national staff for PhV per capita

PhV-staff NCA PhV-staff NCA+RC
[FTE per million capita] [FTE per million capita]
Minimum 0.2 0.2
Median 0.772 1.183
Maximum 4.6 4.6
Staff for pharmacovigilance, scientific and administrative.
3.5.3 Collaboration in national system

The national agencies of bigger countries have the possibility to collaborate with
regional centres for PhV. This is the case in 12 of the answering 29 countries
(Figure 3.27).
Figure 3.27. Existence of regional centres
70
60
50
Responses in %
40
30
20
10
0
No Yes
The staff for regional centres in these countries differs between 5 and 82 persons.
The regional centres are usually specialised in and responsible for some PhV
tasks. They are in all eleven countries with regional centres responsible for data
collection, but – as expected – none of them for decision making (Table 3.9).
Table 3.9. Tasks of regional PhV centres

N of countries
Data coll./management 11.00

Signal detection 7.00
Safety issue assessment 7.00
Decision making .00
Communication 6.00
Scientific studies 8.00
Informal advice 7.00
Inspection of MAHs .00
In some countries the regional centres are responsible for the total collection of
ICSRs, in other countries they play only a minor in this process stage (Table 3.10).
Table 3.10. Number of ICSRs collected by regional PhV centres

N of regional Total N of reports col- % of ICSRs collected
centres lected by all regional by regional centres
centres 2004
Country DE 6 1000 6.35
ES 17 7476 100.00
FR 31 20116 100.00
NL 1 5050 100.00
NO 5 1490 86.00
PL 3 92 8.86
PT 3 1104 64.26
SE 6 4124 100.00
UK 5 5054 25.21
LU 1 n.a. n.a.
The median population-based reporting rate is 160 (see Table 3.17). Above this
limit lay the 12 countries SE, NO, DK, UK, FR, NL, BE, DE, FI, SL, ES, PT, and
SV, 8 of these 12 (those printed in bold) have regional centres. This is a strong
indicator that the RCs contribute to high reporting rates.
In addition to regional centres there are some other possibilities for the national
agencies to collaborate in PhV. Most common is contracting-out tasks to academia
and to HCPs.
Table 3.11. Contracts with other actors in PhV assessments

Contracting-out to...
N of countries
Contract out to academia 15
Contract out to HCPs 13
Contract out consultants 5
Contract out RCs 6
3.5.4 Competences
In combination of national agencies with regional centres and external experts,
most of the countries have competences in the various fields of PhV (Table 3.12),
but for smallest countries it is quite a problem to get enough national expertise at
least in some fields.
Table 3.12. Competences available in countries

National Regional External Not at
agencies centres Experts all available
in country
Exp.toxicol. 17 1 18 3
Animal studies 15 , 20 2
In vitro testing 16 , 17 3
Pharmacology 21 7 20 1
Medicine 26 10 20 ,
Pharmacoepi 17 6 20 1
Epidemiology 13 4 22 1
Statistics agency 17 3 21 1
Human ADRs/veterinary products 13 2 11 4
Design PhV plans 19 1 10 5
Regulatory affairs 27 1 6 ,
Most of the countries have a special expert committee dedicated to PhV. This is
the case for 20 of the answering 28 countries (EU25 + EEA) and an improvement
compared to previous surveys. The countries with no expert committee are mostly
small countries.
The expert committees meet in median five times per year. This is a small reduc-
tion compared to the previous surveys (Table 3.13).
Table 3.13. Expert committee meetings per year

Minimum Maximum Median
PhV comm. meetings p.a. ERMS 2002 1 30 6
PhV comm. meetings p.a. ERMS 2004 5 11 7
Expert committee meetings (own survey) 0 22 5
In 11 countries the expert committee is not only responsible for PhV, but for mar-
keting authorisation and variations (e.g.) as well.
3.5.5 International collaboration

The national agencies contribute in different aspects to the EEA PhV system. 14 of
the countries have already been rapporteur for a centrally approved product (CAP).
17 of the 26 countries that answered this question in the survey have also the ca-
pability of leading EU-wide co-ordination of regulatory action and communication of
drug safety issues. But the most important routine task in internal collaboration is
the writing of assessment reports with a median of 25 reports per country and
country (Table 3.14). One agency produced 2012 assessment reports. On aver-
age, 1.63 assessment reports were written per staff FTE, with a maximum of 231
assessment reports per person. 0.23 Legal documents/guidelines were prepared
per staff FTE.
Table 3.14. Contributions to EEA PhV system

Minimum Maximum Median
Assessment reports written .00 2012.00 25.00
Answers to the CHMP .00 75.00 .00
Answers to CHMP per staff FTE .00 6.82 .00
Legal documents/guidelines prepared .00 131.00 2.00
The majority of the national agencies consider their cooperation with the EMEA
positive (Figure 3.28).
Figure 3.28. Cooperation between national agencies and EMEA
70
60
Responses in %
50
40
30
20
10
0
One goal in the internal collaboration is the reduction of duplication of work. At the
moment the agencies have a quite different opinion about how much work is done
in signal detection and safety issue assessment in duplicate within the own country
and at the same time in other MS or on EU level. But the majority thinks that the
duplication of work is rather little. A more detailed analysis of this issue reveals that
more duplication of work is experienced by the larger countries and most of the old
MS (potentially these are also the countries that have in the past contributed the
most to community tasks), whereas the smaller and new MS assess duplication of
work to be very little.
Figure 3.29. Duplication of work
45
40
35
30
Responses in %
25
20
15
10
0
Very little 2 3 4 Very much

3.5.6 Technical resources

In addition to financial and personal resources the technical resources are impor-
tant for PhV work. As expected, almost the whole staff in all agencies has a PC
and evaluates the number of PCs as sufficient. Almost all of the agencies also
have a local area network, permanent internet access and IT support. But not in all
agencies the IT resources are completely satisfying the needs, as the assessment
of the IT-resources in the following figure shows.
Figure 3.30. Assessment of IT-resources
70
60
50
Respones in %
40
30
20
10
0
Totally 2 3 4 Very
sufficient insufficient
3.5.7 Definitions and standards

Guidance documents play an important role for the assurance of standards in PhV.
Most of the agencies have own national versions for the relevant guidance and
SOPs, as can be seen in (Figure 3.31 and Figure 3.32).
Of all guidance documents, those for obligations of MAHs, data management, data
collection, feedback to reporters, and obligations of the NCA are the most common
ones. Relatively infrequent are guidances on Collaboration with other authorities
and international health institutions, signal detection and decision-making.
Figure 3.31. Guidance in national version (part 1)

30
25
Number of responses
20
15
10
tl.
tl.
tl.
tl.
tl.
l.
l.
at
na
na
na
at
na
na
tn
r.n
AH
M
AH
en
o
io
C
Q
th
.M
.M
ct
.N
em
ce
au
le
lif
ig
ig
ag
co
b.
ua
da
bl
bl
an
la
O
Q
ui
ta
l
m
Co
ce
ce
Da
ce
ta
n
n
n
ce
P
da
da
Da
da
SO
n
ui
ui
ui
da
P
G
SO
ui
G
The following diagram contains the respective figures for a second set of guidance
documents.
Figure 3.32. Guidance in national version (part 2)
25
20
Number of responses
15
10
0
tl.
tl.
tl.
t l.
t l.
t l.
l.
na
na
at
na
na
na
na
tn
s
rs
Ps
n
Ps
ng
AH
tio
en
rte
SO
C
i
ak
ec
em
.M
po
.H
-m
et
t
m
ag
re
en
d
m
n
an
m
k
sio
al
Co
Co
ac
p
gn
m
ci
lo
b
P
Si
P
De
is
ed
ve
SO
SO
is
P
Fe
De
Cr
P
SO
SO
P
P
SO
SO
SO

For the cases where no national document exists, there is usually an EU document
available. Only in exceptional cases it is claimed that there exists neither a national
nor an EU document.
3.6 General quality management

To assure the quality of their PhV system the national agencies use different pos-
sibilities. This depends of course at least in part on their in-house and external re-
sources. A systematic quality management is not implemented in the most PhV
departments. In a few countries the PhV department is part of an agency Quality-
Management. Some others have either an internal audit or an external peer review.
More common approaches to ensure the quality of the PhV system are the usage
of internal or external expertise in certain process stages of PhV. Especially em-
phasis is set on the collaboration with external experts. This happens in different
ways:
• consultation external experts, e.g. for PSUR evaluation
• support of expert committees, especially in the process of decision-making
• usage of external assessment reports and literature
There are also different approaches by using in-house competence to ensure qual-
ity:
• usage of relevant SOPs as guidelines (most common)
• analysing of earlier decisions or cases
• regular (daily, weekly) PhV department meetings
• meetings if there are new signals.
Less emphasis in regard to quality assurance is set on the electronic support of
signal detection, e.g. statistical calculation of signals.
To ensure and evaluate the quality of their actions the most mentioned answer of
the countries is the check of sales or consumption data of the relevant drug. But
this data is not for all agencies available.
Other practices of several agencies are:
• peer review by ministry/director etc.
• check of variations in SPC
• consultation with scientific experts
• consultation with international stakeholders
Only in 9 of the 27 answering agencies there is an audit procedure for the different
steps or pharmacovigilance. 21 agencies state that their audit procedures do not
adequately ensure the quality of their work.
An average of four scientific publications with at least one author from the agency
were published in the last year by each of 23 responding agencies.
It was asked how well the agency meets its internal targets for timing and other
requirements. The results are presented in Figure 3.33
Figure 3.33. Meeting of internal targets
45
40
35
30
% of agencies
25
20
15
10
0
Of 21 agencies, the compliance with their 15day obligation of reporting to EMEA or

to MAHs is nearly perfect in 62% and high in 19% according to their self-reports in
the agency survey. It is only medium or low in 19% of the agencies. Follow-up in-
formation is not transmitted within 15 days by 52% of the agencies.
The internal cooperation within the agency (within PhV unit, with pre-marketing
department, incl. IT staff) is assessed as very good or good in 93% of the agen-
cies. The cooperation of the agency with HCPs is very good or good in only 68%,
and the cooperation of the agency with the MAHs is very good or good in 93% of
the agencies.
The compliance of MAHs with expedited reporting is routinely checked in 41% of
the cases, the MAH's compliance regarding PSURs is only checked in 56%.
Action in the case of non-compliance is taken in 52%. Such actions are e.g. warn-
ing letters, inspections, and financial penalties.
3.6.1 Training of the staff

Personal training, e.g. in research methods or database-administration is not a very
frequent means of quality assurance. In 8 of the 28 agencies the whole staff at-
tended at least one training in 2004. An average of 50% of the staff received a
training measure in the last year. In the opposite there are 7 countries where less
than the quarter of a personal attended one. So there are huge differences in this
topic, which are not solely explicable by the size of the countries.
3.6.2 Education of reporters

In the average country, 5 events have taken place in the last year with participation
or support from the Agency to educate reporters/HCPs in pharmacovigilance?
3.7 Phases of pharmacovigilance

The combined results of the interview and questionnaire survey at the agencies are
presented in the following paragraphs according to the phases of pharmacovigi-
lance (Data collection, data management, signal detection, safety issue assess-
ment, decision-making, and communication and action to protect public health).
3.7.1 Data collection

Each national pharmacovigilance centre should have a system in place that allows
the receipt, management and evaluation of all pharmacovigilance data within that
Member State in a way which is compatible with the procedures undertaken in the
other Member States and the EMEA ("Volume 9", updated July 2004, p. 65).
Data to be collected are in the first line spontaneous reports on suspected adverse
reactions (ICSRs) from MAHs and from HCPs, as well as PSURs from MAHs. In
the following we call the combination of these two sources "routine data". Reports
can be made in writing (e.g. using report forms), by telephone, electronically, or by
any other approved way.
3.7.1.1 Spontaneous reporting
Reporting of suspected ADRs is mandatory for healthcare professionals in 17 of

the answering 26 countries, in general for physicians and pharmacists, in AT e.g.
also for dental surgeons and midwives, in EE also for doctors, nurses, midwives,
and dentists, in ES, IT and SL also for all other types of Health Care Professionals.
In Germany, reporting is not obligatory by law (except for ADRs following vaccina-
tion and blood products), but by a self-commitment of physician's associations. In
IT reporting is mandatory for all suspected serious and unexpected non-serious
ADRs for all drugs. In addition for vaccines and for those drug included in the "in-
tensive monitoring scheme" reporting of all ADRs is requested.
On the other hand, an independent report of the UK spontaneous reporting system
published in April 2004 considered whether there should be a legal requirement to
report ADRs. Published data do not support a better reporting rate in countries that
have mandatory schemes compared to those that do not9. Because of this and the
practical difficulties of enforcing such a law the review did not recommend the in-
troduction of mandatory reporting in the UK.
Absolute numbers of reports

As the following table shows, as of July 2005, the reports are generally submitted
on paper to most of the agencies. Exceptions are IT and LV, where 90% and 50%
of the reports come via the web-site, respectively. 10 or more percent are already
submitted electronically in BE, FI, IC, LT, LU, and PL with a exceptionally high 80%
of electronically submitted reports in SL.
9 Hughes ML, Whittlesea CMC and Luscombe DK. Review of national spontane-
ous reporting schemes. Adv Drug React Toxicol Rev 2002 21 (4): 231-241
Table 3.15. Submission of reports on ADRs [% of ICSRs] as of July 2005
On Paper Electronically Via web-site Other

AT 99 . . 1
BE 90 10 0 .
CY 90 0 0 10
CZ 98 1 0 1
DE-BFARM 100 0 0 0
DE-PEI 100 0 0 0
DK 93.80 2.60 2.60 1
EE 90 . 10 .
EI 99 . . 1
ES 95 . 2 3
FI 85 15 . .
FR 95 1 1 3
HU 95 5 0 0
IC 80 20 0 0
IT 10 0 90 0
LI . . 0 0
LT 80 20 0 0
LU 80 20 . .
LV 50 . 50 .
MT 100 0 0 0
NL 82.50 2.40 15.10 0
NO 100 0 0 0
PL 90 10 0 0
PT 100 0 0 0
SE 100 . . .
SL 20 80 0 0
SV 100 0 0
UK 94 3 3 0
For reporting, some countries have a specific form, often called "yellow card", but
also accept the standard CIOMS and other forms.
Incomplete ICSRs, i.e. ICSRs submitted with less than 4 minimal data points are
rare in the most countries with a mean of 0.72% of incomplete reports; the data
quality is acceptable for 23 of 27 respondents.
Figure 3.34 and Figure 3.35 give the absolute numbers of ICSRs that were re-
ceived by the national agencies in 2003 and 2004. The related relative indicators
("reporting rates") are more useful to assess the performance of the system than
these absolute numbers. The relative values are presented in Table 3.17 and
Figure 3.38, Figure 3.41 and Figure 3.42).
Figure 3.34. ICSRs received 2003 and 2004 (countries with numbers of ICSRs
≥2000)
22000
20000
18000
N u m b e r o f IC S R s
16000
14000
12000
10000
8000
6000
4000
2000
0
UK DE-total FR DE- ES IT NL DE-PEI SE BE DK
BFARM
ICSRs received 2003 ICSRs received 2004
Figure 3.35. ICSRs received 2003 and 2004 (countries with numbers of ICSRs
<2000)
2000
1800
Number of ICSRs
1600
1400
1200
1000
800
600
400
200
0
GR EI NO PT CZ FI AT SV PL SL HU LT LV EE CY IC MT LI
ICSRs received 2003 ICSRs received 2004
The numbers of ICSRs received generally rose between 2003 and 2004.
The total number of ADR reports that have been accumulated over the years in the
national databases heavily depends on the size of the country (Figure 3.36), but
also on the policies of transferring old reports into the new databases and when
data started to be collected.
Figure 3.36. Total number of ADR reports in the national databases
600000
500000
Number of ICSRs
400000
300000
200000
100000
0
AR l
M
D DE UK
EI
Y
U
IC
T
L
FI
PL
O
LV
LI
IT
FR
AT
PT
ES
SE
D SV
BE
EE
LT
BF ota
PE
N
M
C
D
C
H
N
E- -t
E-
ICSRs in database total
In 2004, more than half of the suspected ADRs were submitted by doctors or den-
tists, followed by ADRs from MAHs. Other groups played a minor role, in most
countries no reports were received from nurses, patients, coroners, or professional
bodies (Table 3.15).
Table 3.16. Suspected ADRs from reporter groups

% [Median]
Doctors/dentists 52.05
MAHs 37.29
Others 3.27
Pharmacists 1.80
Nurses .00
Patients .00
Coroners .00
Professional body .00
Particularly high (i.e. >70%) are the shares of ADRs from MAHs for the following
agencies: LI (100% of ADRs from MAHs), DE-BFARM (83%), BE (75%), and DE-
PEI (74%), whereas in IC 96% of ADRs come from doctors/dentists (DK: 92%; SV:
95%; IT: 87%; SE: 87%; FI: 77; and NO: 71% from doctors/dentists).
Reporting by size of the population

The reporting rates computed by number of received ICSRs divided by the popula-
tion size differed greatly. They are presented in Table 3.17 along with the figures
from the two ERMS surveys.
Table 3.17. Reporting rates – different indices

Rep.- Rep.- Reporting Reporting Rep.- Rep.- Reporting
rate rate rate total rate total rate rate rate chil-
total total (p.Mio. ) (p.Mio.) children children dren <=19
(p.Mio) (p.Mio) 2003 2004 (p.Mio) (p.Mio) (p.Mio.)
ERMS ERMS ERMS ERMS 2004
2002 2004 2002 2004
Valid 13 9 25 27 11 1 23
cases
Median 228.0000 40.0000 105.3531 159.8483 51.0000 9.5000 63.4622
Minimum 55.00 18.20 .00 23.13 4.20 9.50 .00
Maximum 458.00 145.70 402.98 459.46 288.00 9.50 406.85
Reporting rates for the total population were smaller (Md=40.00 reports per million
inhabitants) in the new MS in 2004 (column "Rep.-rate total (p.Mio) ERMS 2004")
than in the old MS in 2002 (column "Rep.-rate total (p.Mio) ERMS 2002":
Md=228.00 reports per million inhabitants). Taken together the rates computed
from the own data in 2003 and 2004 (3rd and 4th column), the rates increased from
105 to 159 per million, but could not reach the old level of the year 2002. However,
an influence of possible differing computation methods cannot be ruled out.
Figure 3.37. Reporting rates for total populations over time, number of ICSRs di-
vided by population size
250
Reporting rate (Reports per million)
200
150
100
50
0
Reporting rate Reporting rate Reporting rate Reporting rate
total ERMS total ERMS total 2003 (per total 2004 (per
2002 (per Mio 2004 (per Mio Mio. capita) Mio. capita)
capita) capita)
Figure 3.38. Reporting rates for total populations 2003 and 2004, number of ICSRs
divided by population size
500
Reporting rates per million capita
450
400
350
300
250
200
150
100
50
0
EI
IC
LI
NL
FA L
PL
-B S I
IT
T
LT
FR
HU
LV
PT
CZ
AT
SE
DK
UK
-t o E
ES
SV
EE
DE CY
NO
EI
l
RM
F
ta
M
DE B
-P
DE
Reporting rate total 2003 per Mio. capita Reporting rate total 2004 per Mio. capita

The highest reporting rates (2004 total population) can be found in Sweden, Ire-
land, Norway, Denmark, the UK, France, and the Netherlands. In most countries,
the rates increased between 2003 and 2004.
The reporting rates for children are far lower than those for the whole population,
which could be a result of lower incidence of diseases and lower consumption of
pharmaceutical products in children. In 2004, an average rate of Md=63 reports per
million children was computed, however also with a small bias because reports
were counted by the agencies for children and adolescents less than 18 years of
age, but the population sizes could only be found for the EU and EEA Member
States for persons up to 19 years (Eurostat). The respective figure for the total
population was Md=160 reports per million inhabitants (Table 3.17).
Figure 3.39. Reporting rates for children over time, number of ICSRs 2004 divided
by number of children
70
Reporting rate (per Mio. children)
60
50
40
30
20
10
0
Rep.-rate children <=18 Rep.-rate children <=18 Rep.-rate children <=19
ERMS 2002 (p.Mio ERMS 2004 (p.Mio (p.Mio children)
children) children)
The differences can be explained by the different samples that were assessed be-
tween the ERMS-2002 survey (old MS), ERMS-2004 (new MS) and the total sam-
ple of the own agency survey in the third column.
Figure 3.40. Reporting rates for children ≤19 years, number of ICSRs 2004 divided
by number of children
450
Reporting rate per million children
400
350
300
250
200
150
100
50
0
IC
M
T
L
28
PL
FR
EI
AR I
U
l
Z
LT
O
SE
ES
EE BE
SV
D LV
D DE K
Y
IT
FI
LI
ta
BF E
N
M
C
H
N
E- -P
to
A-
E-
Reporting rate children ≤19 years in 2004 per Mio. capita
Sweden, France, the Czech Republic, Italy and Ireland have extraordinarily high
reporting rates for children compared to other countries.
Reporting by type of product

The following table presents the numbers of ICSRs that the agencies received in
their different roles as responsible agency for nationally authorised products
(NAPs), as concerned MS for products licensed through the mutual recognition
procedure (MRPs) and as Reference Member State for MRPs. Most products are
authorised as NAPs (a median of 4286 products per country), the second most are
MRPs (Md=406 per country), the least are CAPs (269 in all countries), therefore it
is not astonishing that most of the received ICSRs in absolute numbers are re-
ceived for NAPs.
Table 3.18. ICSRs by MA procedure
ICSR on NAPs 234.00 1.00 8050.00
ICSR as concerned MS 10.00 0.00 8500.00
ICSRs as RMS 0.00 0.00 4100.00
→ National ICSRs per NAP 0.05 0.006 1.31
→ ICSRs as concerned MS per

0.03 0.00 2.98
MRP
→ ICSRs as RMS per MRP 0.00 0.00 1.03
On average (Median), 234 ICSRs were received by each agency on NAPs, 10 on

MRPs with the respective country as concerned MS, and 0 on MRPs with the coun-
try as RMS. The variances are substantial in these figures. Relative to the numbers
of products on the market, the reporting rate was best for NAPs (0.05 reports per
nationally authorised product), second best for ICSRs as concerned MS per MRP
(reporting rate = 0.03), and least for ICSRs as RMS per MRP (reporting rate =
0.00). The low median for ICSRs concerning MR products for which the country is
RMS is certainly due to the fact that many of the participating agencies have not
yet been a RMS and therefore had to answer "0 reports received as RMS" in the
survey.
Reporting by market size

The number of submitted reports will certainly depend on the sold volume of phar-
maceutical products (the market size), which is here measured as the countries'
pharmaceutical sales in US$ (source: WHO-EURO HFA-DB;
http://data.euro.who.int; those numbers of 2002).
Table 3.19. Reporting rates for total populations, number of ICSRs divided by phar-
maceutical sales in the respective country
Indicator Reporting rate total Reporting rate total Reporting rate children
2003 per sales in US$ 2004 per sales in US$ <=19 2004 per sales in US$
Valid n 20 18 16
Mean 15.4439 17.6737 2.2664
Median 7.6668 8.7678 1.3999
Minimum .00 .79 .00
Maximum 79.21 82.49 9.05

Figure 3.41. Reporting rates for total populations; number of ICSRs 2004 divided
by pharmaceutical sales
90
80
Reporting rate by market size
70
60
50
40
30
20
10
0
EI
NL
SL
FR
FI
IT
CZ
PT
AT
HU
UK
ES
SE
DE E
DK
SV
EI
FA l
RM
ta
B
-P
-t o
DE
-B
DE
Reporting rate total 2004 per sales in US$
Reporting by number of physicians

The following table gives the reporting rates that are standardized by the number of
physicians in the respective country. It can be seen that the average (median) val-
ues increase between 2003 and 2004, but that huge differences exist between the
countries with the minimum rates and the maximum rate. As it can be expected,
the reporting rate for children is again lower than the rate for the whole population.
Table 3.20. Reporting rates for total populations, number of ICSRs divided by
number of physicians
Indicator Reporting rate total Reporting rate total Reporting rate children
2003 per physicians per 2004 per physicians per <=19 2004 physicians per
100.000 capita 100.000 capita 100.000 capita
Valid n 27 27 25
Median 3.38 4.51 .41
Minimum .00 .01 .00
Maximum 90.53 94.28 20.83

Figure 3.42. Reporting rates for total populations, number of ICSRs 2004 divided
by number of physicians
Reporting rate by number of physicians
100
90
80
70
60
50
40
30
20
10
0
EI
IC
LI
FI
D E IT
LT
T
NL
PL
SL
DE E F R
LV
PT
CZ
AT
HU
UK
ES
SE
DK
BE
SV
EE
CY
NO
EI
F A al
RM
M
- B -t o t
-P
D
Reporting rate total 2004 per physicians per 100,000 capita
3.7.1.2 Underreporting
According to the agency interviews, underreporting is a problem in many countries,

especially in the smaller Member States, but no valid estimates are available that
are comparable between the countries.
The most important reasons for underreporting are
• lack of time of the reporters (11 responses in the 27 interviews),
• economic reasons/missing incentives (7 responses),
• lack of education (9 responses),
• being afraid that wrong treatment could be revealed (8 responses) and
• a negative attitude towards reporting (5 responses).
Less often mentioned were missing interest, administrative workload, missing legal
obligation to report, missing electronic system for reporting and technical reasons,
and cultural or traditional reasons.
Interesting aspects on reporting rates are related to the market structure: In one
country the domestic MAHs produce only generics which are perceived as safe; in
addition, the MAHs do not send sales agents to doctors to collect ADR reports in
this country, and have many OTC drugs and therefore do not work with physicians,
factors which all will reduce the reporting. Another assumption is that HCPs can be
disappointed because of missing perceived action based on their report. In addition
organisational and technical reasons exist for low reporting, the system is not yet
simple enough, and – as one interview partner in an agency assumed – MAHs
might use their influence on doctors in order to reduce reporting e.g. by giving them
simplified information on ADRs.
3.7.1.3 PSURs
Periodic safety update reports (PSURs) are the periodical reports that the MAHs
have to submit to the concerned agencies containing the ICSRs they have re-
ceived in the last period as well as other safety-relevant information referred to in
Article 104 of Council Directive 2001/83/EC.
The total number of PSURs received varies between a maximum of 2940 in one
country and no single PSUR received in three countries.
Table 3.21 gives an overview on the PSURs that are received by the agencies ac-
cording to the different authorisation procedures.
Table 3.21. PSURs by MA procedure

Natl. PSURs received 517.50 0.00 2500.00

National PSURs per million capita 42.1482 0.0000 228.5271
National PSURs per NAP 0.0707 0.0000 2.4604
PSURs received as RMS 11.50 0.00 675.00
MR PSURs as RMS per million capita 0.2032 0.0000 64.9263
MR PSURs as RMS per MRP 0.0243 0.0000 1.0000
PSURs received as rapporteur 11.00 0.00 370.00
CAP PSURs as rapporteur per million capita 0.4357 0.0000 23.1811
PSURs as rapporteur per CAP 0.0409 0.0000 1.3755
On average (Median), 518 national PSURs were received per country, i.e. 42 per
million population or .07 national PSURs per authorised NAP. PSURs for MR au-
thorised medicines are received much less frequently, but are related to the num-
ber of MR drugs for which the country is Reference Member State. The number of
PSURs is of the same magnitude, as is for PSURs per CAP for which the country
is rapporteur.
In some countries, all of the received and even foreign PSURs are assessed, but in
a few countries the percentage of assessed PSURs is below 10%.
3.7.1.4 Other sources of information
Other sources of information are analysed in addition to the ICSRs and PSURs.
20 of 25 answering agencies have access to databases: Besides Eudravigilance
and the EMEA website, the UMC Vigibase10 is mentioned most frequently (11 an-
swers), followed by Medline11 (8 answers). Some agencies have access to regis-
tries of products and drug consumption data. Other national registries (on birth,
cancer, hospital admissions, GPRD…) exist, and also a clinical trial database is
accessed.
The agencies were asked if there are other data that they are using or could use
for signal detection or safety issue assessment in their country. The answers are
summarized in Table 3.22. In its second column, the number of agencies is given
which mention that a certain source exists in their country. The next three columns
headlined by "Access" give the shares of agencies that have access to the source
never, in exceptional cases or always. The last three columns describe the use of
the source in similar categories as share of all responding agencies.
Table 3.22. Existence and use of population-based health/disease registries
Exist Access Use

N of % in % in
% % % %
agencies except. except.
never always never routinely
cases cases
Cancer 24 13.6% 72.7% 13.6% 36.4% 59.1% 4.5%
Causes of
23 12.5% 50.0% 37.5% 29.2% 58.3% 12.5%
death
Malformations
22 9.5% 71.4% 19.0% 13.6% 77.3% 9.1%
newborns
Inpatient care 19 15.0% 60.0% 25.0% 36.4% 50.0% 13.6%
Outpatient care 14 21.1% 52.6% 26.3% 36.8% 47.4% 15.8%
Intrauterine
11 31.3% 56.3% 12.5% 37.5% 56.3% 6.3%
drug exposure
Except for outpatient care and intrauterine drug exposure, such registries exist in
most of the countries. However, most agencies do only have access to these data
in exceptional cases, and they are quite infrequently used.
The following table shows corresponding analyses for sales and prescription data.
10 A database for ICSRs that is kept by the WHO at the Uppsala Monitoring Cent-
re. At the moment, 77 counties worldwide submit reports to this database.
11 A literature database in which many references from medical jounals are listed,
hosted by the US National Library of Medicine with free access to references
and abstracts.
Table 3.23. Existence and use of data on the consumption of medicines
Exist Access Use

N of % in % in
% % % %
agencies except. except.
never always never routinely
cases cases
Sales data 24 0% 21.7% 78.3% 0% 33.3% 66.7%
Prescription
19 14.3% 38.1% 47.6% 25.0% 40.0% 35.0%
non-hospital
Prescription by
18 15.8% 42.1% 42.1% 35.3% 35.3% 29.4%
region
Prescription
data only for 16 15.0% 45.0% 40.0% 22.2% 55.6% 22.2%
reimbursed
Prescription
14 31.6% 42.1% 26.3% 47.1% 23.5% 29.4%
hospital
Prescription by
14 22.2% 44.4% 33.3% 31.3% 43.8% 25.0%
age
Prescription by
14 22.2% 44.4% 33.3% 31.3% 43.8% 25.0%
sex
Sales data exist in nearly all countries, less frequent are data on prescriptions
made in hospitals. These data cannot always be analysed by age, sex, or geo-
graphical region.
Of all the data sources that are used in addition to the routine data, registries on
cancer cases, causes of death and malformations of newborns exist in nearly all
countries, to a lesser extent also databases on inpatient care. To all these sources,
at least half of the respective agencies have access at least in exceptional cases.
Data on malformations of newborns are used by nearly 90% of those who have
access to such registries at least in exceptional cases, followed by causes of death
data (71% of those who have access) and cancer registries (65%). Databases on
outpatient care exist in only 14 countries, but if existing they are the data source
that is used most frequently on a routine basis (16% of those agencies where out-
patient data exist).
Among the usage data, sales data are available in nearly all countries and all of
those have access at least in exceptional cases, nearly 80% always. These data
are routinely used by nearly 70% of the countries where such data exist, the rest
uses them at least in exceptional cases. Prescription data for outpatient care exists
in 19 countries, but is restricted in most cases to prescription medicines. If data on
sales or on outpatient prescription exists, the agencies normally have access to
these data and three fourths of them use these data at least in exceptional cases.
Prescription data from hospital care as well as the differentiation of age or gender
groups are less frequent. If existing, only one third of the agencies has unrestricted
access and thus these data are used only by 25 to 29% of the agencies where
such data exist.
Other data sources exist but are not used very frequently. The agencies mention
here:
• Databases on primary healthcare outcomes and prescriptions (GPRD, IMS
Health, PHARMO, General Drug Registry-implementation on-going)
• Disease-specific registries (Liver or kidney disorders, Diabetes, Infect. Dis-
eases, Rheumatoid arthritis and treated patients, HIV infected and treated
patients, organ transplantation, Case-control surveillance on blood dyscra-
sias, Cardiovascular Disease Register, rare paediatric diseases, Follow-up
of rheumatologic patients exposed to biological products)
• Medical Birth Registry, maternity
• Surveys/studies (Children and youth survey, The Reykjavik Study, National
Health Survey)
• Poison Centres
• Healthcare insurance register
• Health Protection Surveillance Centre
• Birth Register
• ISTAT
• IPCI data base
• Adverse events after vaccination
• Vaccination coverage, Post-vaccination reactions
In the ERMS-2002 survey, only 3 agencies, and in the ERMS-2004 survey only 4
agencies had plans to obtain some data sources at which they do not have access
at the moment. In the new survey, 15 agencies mention that they plan to broaden
their access to more data sources.
Seven agencies have the capability to link prescription registries with other regis-
tries which include health outcomes, and eight have experience in conducting
pharmacoepidemiological studies using such data.
In total over all respondents, 432 pharmacoepidemiology studies, post-
authorisation surveillance studies or phase IV trials have been carried out last year
with a sample from their country, taken all sponsors together (in the survey it was
not distinguished between public and private). The highest numbers of PM studies
were found in the UK (No accurate figure available, the number is estimated at
>100), ES (n=92), and HU (n=61).
Of all studies, 49 were initiated by an Agency, most of them in FR (12 of 16 French
studies initiated by the Agency), followed by DE (11 studies initiated by on of the
two German Agencies).
15 of 26 evaluate reporting rates calculated from spontaneous ADRs and usage
data.
In eight countries, ad-hoc pharmacoepidemiological studies were carried out in
2004 when a signal needed confirmation or quantification, in four countries studies
for early PM surveillance of new products took place in 2004.
Clinical trial adverse event (AE) reports are collected by 21 authorities and are a-
vailable to those staff responsible for pharmacovigilance of marketed products.
Information is collected by 22 agencies on ADRs with compassionate use / named
patient use of products.
Various literature sources are screened, the median is on a weekly basis. The most
frequently used sources are presented in Table 3.24.
Table 3.24. Most important literature sources
Number of agencies
Medline, Pubmed 17
Lancet 8
British Medical Journal 7
New England Journal of Medicine 7
Reactions Weekly 6
WHO database 5
Drug Safety 4
JAMA 4
Micromedex 3
The following table shows, that in most agencies information on ADRs with com-
passionate use, on AEs in clinical trials, and on phase IV efficacy trials are rou-
tinely collected and recorded. Information from other regulatory authorities and on
post-authorisation safety studies are collected less frequently, and only 10 agen-
cies routinely record data/information on preclinical studies.
Table 3.25. Information collected
N of agencies
ADRs with compassionate use 22.00
Clinical trial AE reports 21.00
Info phase IV studies 18.00
Info from other authorities 15.00
Info PASS 14.00
Info preclinical studies 10.00
3.7.2 Data management

The collected data, especially the ICSRs, have to be stored, cleared from dupli-
cates, checked for completeness, eventually transmitted to other stakeholders and
prepared for analysis. In most agencies the ICSRs are stored in electronic data-
bases in predefined format. Member States should ensure that data on suspected
serious ADRs occurring in their territory are uploaded into the EudraVigilance da-
tabase. A sufficient number of comparable ICSRs is a prerequisite for the identifi-

cation of signals by use of statistical tools especially for rare ADRs.
26 agencies use an electronic database to manage national ICSRs, four of the re-
sponding agencies do not. Very frequently agencies have developed their specific
software systems for this purpose, although most of them and all of the more re-
cent systems are E2B compliant and have a very similar functionality.
Duplicate reports are generally identified manually. 11 databases include audit trail
functionality. 16 agencies convert old reports into ICH standards, three hold EU
and Non-EU reports in their database. In all cases, all reports are accessible for the
signal detection process either electronically or – if only very few reports are re-
ceived in an agency – for manual analysis.
EudraVigilance is already in place in 15 agencies (as of July 2005). The standards
for electronic transmission are implemented in 20 agencies. Some of the agencies
use the WebTrader module of EudraVigilance.
Electronic reporting by the MAHs is in place in 11 agencies, in 2 agencies all re-
ports are already transmitted by MAHs electronically; in two more agencies the
share is more than 50%. If received electronically, the agencies can put the re-
ceived reports automatically into their databases.
Paper reports are validated, i.e. checked with the reporter especially for serious
cases and incomplete data or by cross-checks of the entered data with the case
narrative in 21 of 27 agencies, electronic data in 14. Data can be aggregated in 20
agencies and the routine data and information from all the other data sources are
readily accessible in 22 of 27 responding agencies (Table 3.26).
Table 3.26. Data quality

N of agencies
Data validation paper reports 21
Data validation electronic data 14
Data aggregated electronically 20
Other data readily accessible 22
The time needed for data processing is assessed as adequate by 14 of 25 answer-

ing agencies, but as "far too slow" by two of them (Figure 3.43).
Figure 3.43. Time for data processing
70
60
50
% of agencies
40
30
20
10
0
Adequate 2 3 4 Far too slow
3.7.3 Signal detection

Signal detection is the identification of probable adverse drug reactions on the ba-
sis of all available information. It includes an assessment of the causality of re-
ported symptoms as most probably being an ADR. In general signal detection
starts by the statistical analysis of ICSRs in the database.
In the surveyed agencies, signal detection is carried out in different ways. In some
agencies no specific procedure exists, some just mention ad-hoc-groups; others
have elaborated procedures as e.g.
– A staff member gets each single report that falls into his area of expertise,
– The head of the department analyses each single report,
– A list of signals is produced by an IT-specialist once a month and is then dis-
cussed by the internal staff of the department (pharmacists, biologists, physi-
cian),
– The clinical assessors present reports at a weekly meeting in the agency which
decides on measures, and a technical committee meets every month,
– Reports are sent to experts, they comment on what should be discussed at a
meeting,
– Reports are assessed in the RCs.
Data are analysed statistically in 18 of the 27 agencies.

The agencies were asked in the written survey, "How do the MAHs in your country
comply with their obligation to analyse safety signals?" Responding to this, the
compliance of the MAHs with their obligation to analyse signals was assessed as
moderate (Figure 3.44).
Figure 3.44. Compliance of MAHs in analysis of signals

50
45
40
35
% of agencies
30
25
20
15
10
0
The following figure presents the evaluation of the statistical tools that the agencies
have available for signal detection, assessed on a rating-scale with values from
"always adequate" to "often very inadequate" (Figure 3.45).
Figure 3.45. Assessment of statistical tools for signal detection

35
30
25
% of agencies
20
15
10
0
Adequate 2 3 4 Often very
inadequate
The evaluation is not very good, alone 6 of the 19 respondents assess the avail-
able statistical tools as often very inadequate.
Tools for the analysis of small numbers of reports have 13 of 25 agencies. A peer
review system is in operation in 21 of 27 agencies. 20 of 27 agencies have a peer
review system for the assessment of safety signals.
The agencies were asked, "How well has your routine data-collection prepared you
for the last pharmacovigilance crisis?" Upon this question, only 35% of the agen-
cies found themselves very good or well prepared for the last crisis by their routine
data-collection (i.e. (ICSRs and PSURs together; values 1 or 2 on a 5-point-rating-
scale), but 40% found it only moderate and 13% bad or very bad (Figure 3.46).
Figure 3.46. Preparation for last crisis by routine data

45
40
35
30
% of agencies
25
20
15
10
0
The routine data are assessed to be very useful for safety issue assessment only
by 2 agencies (7.7%), and useful by 35%. Only "moderately useful" (middle scale
value) or even "only marginally useful" was answered by 48% of the agencies
(Figure 3.47).
Figure 3.47. Usefulness of routine data

40
35
30
% of agencies
25
20
15
10
0
Very useful 2 3 4 Only
marginally
useful
70% of the agencies have not always had sufficient information to make decisions.
When asked which kind of data they would you have needed in addition, the follow-
ing data sources listed in Table 3.27 were mentioned.
Table 3.27. Informational needs for signal detection
Larger number of ICSRS

Information from other agencies
• Assessment reports from abroad;
• Info about reaction of other agencies (type, speed)
• SPCs from other countries
• Latest PSUR with benefit-risk evaluation
Basic research
• Mechanisms of drug effects
• Mechanisms of ADRs, pathomechanisms
• Basic research on vaccines
Better access to literature
Pharmacoepidemiological data
• Treatment-indications
• Patient groups
• Information on population
• National/regional situation (access to treatment, e.g. how easy it is to get
the drug …)
• Outcomes
• Epidemiological data with same diagnosis criteria for diseases / reaction
• Epidemiological studies, epidemiol. comparisons
• Incidences (by indications)
• Comparator data
Registries, databases
• Registries (on birth defects; poisoning centres; vaccines ;…)
• WHO database
• New registries for ADR-related diseases
• Historical data, narrative
• Information on old medicinal drugs
• Combination of different databases, e.g. with morbidity data
Studies
• Pre-marketing data
• Clinical studies
• PASSs
• Evidence-based data; evidence on higher level of evidence hierarchy
Utilisation data
• Usage data from the insurances
• Utilisation studies: many done for the pricing/reimbursement
• Prescription behaviour
• Drug use data; drug utilisation research
• Exposure data
3.7.4 Safety issue assessment

Safety issue assessment has the task to evaluate the causality of a signal as an
ADR and to appraise the severity and potential impacts on public health. This is
normally done with support from or by an external expert committee.
External experts are routinely involved in safety issue assessment by half of the
agencies; an expert committee to review safety assessments exists for 70% of the
agencies. To receive support from external experts on a routine basis is generally
easy, but for 40% of the agencies it is difficult and for one agency nearly impossible
(Figure 3.48).
Figure 3.48. Receive support from experts routinely
50
45
40
35
% of agencies
30
25
20
15
10
5
0
Very easy 2 3 4 (nearly)
impossible
To receive support from external experts in exceptional cases is even easier for the
agencies than in routine cases, but only one third of the agencies get this support
always when necessary.
15 of 29 agencies (52%) have the capabilities in their country to identify and as-
sess signals without help from other agencies, also meaning that 48% or 14 agen-
cies do not have this capabilities.
3.7.5 Decision-making
After the signal has been detected and identified as a safety issue it has to be de-
cided if and what action should be taken. These decisions are often made upon
advice from the pharmacovigilance staff or external experts by supervising bodies
within the agencies of even by institutions outside the Medicines Agencies, e.g. in
health ministries or by the European Commission.
Decisions about actions are made by groups in 90% of the agencies, in general by
agency-internal councils. The pharmacovigilance department prepares such deci-
sions with support of the expert committees.
In about half of the agencies external stakeholders (doctors, pharmacists and pa-
tients) are involved in decision-making (in general as advisors or asked for com-
ments on proposed decisions). In two agencies, the MAHs are consulted, in nine
agencies other groups.
Decisions are based on a set of options and are recorded together with the rea-
sons for them.
The agencies were asked how frequently adequate decisions are found for safety
issues without explaining the word "adequate". 63% of the agencies find that ade-
quate decisions concerning NAPS are always or often found. The respective value
for MRPs is 77% and 74% for CAPs, respectively (Figure 3.49).
Figure 3.49. Adequate decisions found for safety issues
NAPs
MRPs
CAPs
0% 20% 40% 60% 80% 100%

% of responses
Always 2 3 4 Seldom
The duration of the decision-making process was assessed fairly well for all three
types of products (Figure 3.50). 54% find that decisions are found in good time for
NAPs, 65% for MRPs and 58% for CAPs (assessment on a five-point rating scale
from 1 "always" to 5 "seldom").
Figure 3.50. Decisions for safety issues found in adequate time
NAPs
MRPs
CAPs
0% 20% 40% 60% 80% 100%

% of responses
Always 2 3 4 Seldom
The total time between the detection of a signal (first discussion within the agency)
and reporting (publishing) of decision with respect to this safety issue (i.e. the time
for the whole process of PhV) was assessed on a 5-point-rating-scale (1 'ade-
quate'; 5 'far too slow'; Figure 3.52). The best or second best values were only
chosen by 38% of the agencies, 50% assessed the velocity as "moderate", and
13% gave an even worse evaluation.
Figure 3.51. Kinetics of total process from signal detection and reporting
60
50
40
% of agencies
30
20
10
0
Adequate 2 3 4 Far too slow

The transparency of the process of decision-making on safety issues in the com-

panies located in the country to the agencies was only assessed as moderate, and
in 27% of the cases as bad or very bad (Figure 3.52).
Figure 3.52. Transparency of decision-making within the companies
50
45
40
35
% of agencies
30
25
20
15
10
0
3.7.6 Communication and action to protect public health

If the competent body comes to the decision that a serious safety issue has been
detected a number of measures can be taken to prevent further related ADRs and
protect public health. The range is from informing HCPs and other competent au-
thorities about the problem over changing the SPC by adding a new contraindica-
tion, e.g., to withdrawal of the product from the market.
Decisions are regularly published via press releases and on the internet. In addi-
tion, they are published for the concerned groups via bulletins, announcements in
journals, communications to medical associations, Q&A documents for patients,
and seldom on electronic networks. Sometimes, proactive information of the public
is only done in outstanding cases and some decisions are only publicly available
on request.
On average Md=6 bulletins with pharmacovigilance contents are published by the
agencies per year (range from 0 bulletins in 5 agencies to 18 bulletins per year with
an extreme value of 60 bulletins).
In the questionnaire we ask which stakeholder groups are routinely informed and
informed on general and of specific safety issues. The results are presented in the
following tables.
MAHs, individual doctors and pharmacists are routinely informed on general and
specific issues by most of the agencies, followed by the public/media and medical
associations. Professional journals and other HCPs are not so much in the focus,
and patient organisations and other groups seem to be underrepresented (Table

3.28).
Table 3.28. Routinely inform on safety issues

Inform on general Inform on specific
safety issues safety issues
% of agencies % of agencies
MAHs 92 100
Individual doctors 92 96
Pharmacists 83 96
Medical associations 71 88
Public/media 61 92
Other HCPs 52 63
Professional journals 52 73
Other groups 26 53
Patient organisations 18 46
The communication procedures that are in place in the agencies are described in
Table 3.29.
Table 3.29. Communication procedures

N of agencies
Procedures for crisis management 25
Systems for immediate communication 23
System for feedback to reporters 20
System to collect feedback 15
Procedures information/feedback 19
• Information/feedback: info on web-site 20
• Information/feedback: bulletins 16
• Information/feedback: letters 17
• Information/feedback: e-mail 15
n=27; Source: Fraunhofer ISI 2005
Although some of the interview partners stated that in general ADRs cannot always
be prevented, many possible actions were mentioned that can be taken to prevent
future ADRs (Table 3.30)
Table 3.30. Possible actions to prevent ADRs
Product interventions
• USR
• Variation of SPC/PIL (Contraindications; warning; information about
ADR/interactions)
• Suspension of the MA
• Withdrawal of the MA
• Suspension of delivery
• Withdrawal of specific lot
Collection of information
•Ask MAH for more information or to conduct post-marketing study
•Record-linkage and registries with good recording
•Risk management programme, Preauthorisation risk management planning,
Pharmacovigilance-planning
• Collaboration with insurance schemes
Provision of Information
• Press releases
• DDL
• Drug bulletin
• Information of other agencies
• Description of ADRs in the Formulary
• Publish SPCs on internet-site
• Contact to physicians in hospitals via chief physicians (=nominated contact points)
• Educate prescribers on annual pharmacovigilance symposium
Market interventions
• Move product from OTC back to prescription
• Marketing interventions (pack size…)
• Restrict advertising (e.g. for OTC)
• Change of availability (e.g. only in pharmacies, prescription only by specialists)
Other interventions
• Inspections including a person of the pharmacovigilance department
A number of common and singular interventions were identified in the categories

Product interventions, Collection of information, Provision of Information, Market
interventions, and other interventions.
Frequently MAHs do not have to be forced but take actions voluntarily. The stan-
dard interventions were assessed by some interview partners as too few to prevent
well-known ADRs. A suspension of the MA that is only possible for 3 months was
seen as too short and is therefore not used very often.
In one country most drugs are only available through pharmacies, which was seen
as an advantage because therefore the agency via the HCPs has better influence
on consumption than with OTC drugs.
Actions as e.g. the provision of actual information can be implemented within few
hours after an ADR has been detected. They are communicated in general to the
other agencies within Europe, to MAHs, doctors, pharmacists and the media, and
to a smaller extend also to patients and authorities outside the EU.
Actions implemented in the last year are shown in Table 3.31.
Table 3.31. Regulatory actions

Median of actions
per agency
Letters to MAHs to amend SPCs 50.00
Occasions DDL sent to HCPs 9.50
Drugs withdrawn from national market 3.00
Inspections of MAHs PhV issue .00
MA suspended on national market .00
The range of withdrawals of MAs was from 0 to 769, for suspensions from 0 to 120
and for inspections from 0 to 51 per agency.
In addition, an average of Md=300 variations of SPCs were evaluated per agency,
with a maximum of 10566 variations of SPCs.
A median of 91.5 responses to requests by HCPs were given per agency in the last
year.
Only 8 agencies out of 23 respondents think that they always have the best meas-
ures to minimize risks from ADRs at their disposal.
The consistency of the communication on safety issues across agencies is evalu-
ated as fairly good (Md=2 on a 5-point-rating-scale; Figure 3.53). The same is true
for the communication on safety issues between agencies on the one side and
MAHs and HCPs on the other side.
Figure 3.53. Consistency of communications
Across agencies
Agencies-
MAHs/HCPs
0% 20% 40% 60% 80% 100%

% of responses

The impact of communications is only followed-up on a routine basis by four of 29

agencies.
3.8 Outcomes of regulatory action

In the interviews we asked if impact of actions is audited, (9 agencies answered
"Yes"), if systems for capturing impact of regulatory action are in place (only in five
agencies), and if peer review and internal/external audit for actions taken is avail-
able (in 11 agencies; Table 3.32). This can be done with help of a registry (if avail-
able), data on use, or by discussions with supervising bodies as the competent
departments in the health ministries.
Table 3.32. Impact of regulatory action audited

N of agencies
Impact of major actions audited 9
Systems for capturing impact of regulatory action 5
Peer review, internal/external audit of action 11
A good way to monitor the effects of actions on prescription behaviour would be to

monitor prescription/consumption data by drug classes or individual products on a
monthly basis and compare the values before the action taken with the values after
the action. This is already done by many agencies.
Other answers relate to evaluation of actions taken by the committee, ministry or
the agency's board of directors, or consulting third parties. Other approaches are to
check compliance by inspections, to check variations/changes in SPC, and case
reviews after handling.
The influence of the agencies communications on the doctors' prescription behav-
iour is assessed as not very high (Figure 3.54).
Figure 3.54. Influence of agencies' communications on prescription behaviour
35
30
25
% of agencies
20
15
10
0
Very good 2 3 4 Very w eak
In the agency questionnaire we also asked for the outcomes of safety-relevant

studies (Incidence of ADR-relevant diseases; Mortality due to ADRs; Hospitalisa-
tions due to ADRs; Quality-adjusted life years (QUALYs) lost due to ADRs) al-
though we know that these figures are not directly comparable across studies and
countries.
However, 9 agencies stated that there are no studies with the incidence of ADRs
as endpoint in their country, 10 stated that there was no study on mortality, 11 on
hospitalisation and 13 on ADR-related loss of QUALYs (Table 3.33).
Table 3.33. Existence of outcome studies

N of agencies % of agencies
No Incidence Study 9 28.1
No Mortality Study 10 31.3
No Hospitalisation Study 11 34.4
No QUALY Study 13 40.6
These figures seem to underestimate the number of countries in which no such

studies exist because some additional agencies did not give information on any
outcome studies either. Because so few studies were notified the outcomes cannot
be analysed statistically. The following table contains the number of studies men-
tioned.
Table 3.34. Outcome studies

Number per Incidence studies Mortality Hospitalisation
country studies studies
Minimum 0 0 0
Maximum 5 4 6
To allow cross-linkages of the results of the present study with international out-
come figures, the following table presents the incidence of ADR-relevant diseases
as published by the WHO.
Table 3.35. Incidence of ADR-relevant diseases

Country Incidence of ADR-relevant diseases
(per 100000)
EI 1.73
MT 1.67
FR 0.53
LU 0.38
ES 0.31
BE 0.19
SL 0.18
EU-25 0.13
EE 0.09
LV 0.09
IC 0.07
LT 0.07
PL 0.06
SE 0.06
UK 0.05
AT 0.04
CZ 0.03
FI 0.03
DE 0.03
IT 0.03
NL 0.03
DK 0.02
GR 0.02
HU 0.02
SV 0.02
NO 0.01
PT 0.01
CY -
LI -
Source: WHO-Euro (http://data.euro.who.int; numbers of 2002)

4 Goals in respect of effectiveness and efficiency

Initially it was intended to supplement a provisional list of goals related to effective-
ness and efficiency of the European system of pharmacovigilance that was based
on the literature by the respective results from the personal interviews with repre-
sentatives of the competent authorities and representatives of the industry, and to
ask the advisors in a Delphi-process to comment on the list. However, the inter-
views revealed nearly no new aspects in this respect; most of the interviewees
found the actual scope of pharmacovigilance (with some modifications) in general
sufficient, many referred to the related WHO definition according to which pharma-
covigilance is
the science and activities relating to the detection, assessment, understand-
ing and prevention of adverse effects or any other drug-related problems12.
Therefore, there was no need to elaborate more on the aspect of additional goals
for pharmacovigilance in the project.
12 WHO-UMC, see http://www.who-umc.org/defs.html

5 Critical success factors

Critical success factors are those elements of the whole process that determine its
performance and can be modified to improve a system. The procedure for identify-
ing and assessing the critical success factors was as follows:
Step 1:
For the 25 EU Member States and EMEA the most critical success factors for an
effective and efficient functioning of the pharmacovigilance system (with respect to
cost-effectiveness, time-efficiency, quality and safety) were identified on the basis
of a systems approach supported by data from the interviews and literature. First
results were discussed at the expert workshop on 15 June 2005. This first step
resulted in a list of 75 draft factors which can be classified into the following catego-
ries (Table 5.1).
Table 5.1. Draft list of critical success factors
1. … for Data collection 5. … for Decision-making
1.1 Comprehensiveness of the data 5.1 Decision-making in legal bodies
1.2 Organisation of data collection 5.1 Decision-making in companies
2. … for Data management 6. … for Communication/Action
2.1 Electronic processing of data 6.1 Early communication
2.2 Processing of data 6.2 Communication to all stakeholders
3. … for Signal detection 6.3 Impact of communications/actions
3.1 Availability of necessary information 7. … for performance in general
3.2 Data analysis 7.1 Legal framework
3.3 International share of work 7.2 Staff
4. … for Safety issue assessment 7.3 General quality
4.1 Share of responsibilities
4.2 Expertise
4.3 Structures
Step 2:
After step 1, the advisors were asked in a Delphi-process to comment on the list of
critical success factors. The experts rated these draft factors as well as the sub-
categories ("1.1 Comprehensiveness of the data" etc. in the list above) under the
general question "Relevance: How important is the factor for the performance of
the European System for Pharmacovigilance (or parts of it)?" according to their
influence on the areas of
• quality of the work,
• compliance with requirements,
• speed ("kinetics"), and
• work load/costs
on five-point-rating scales.
The full results of the Delphi survey can be found in Annex 4.
Step 3:
On the basis of the assessments of all draft factors as well as in reflection of the
discussion of the factors at the expert workshop in Brussels on 15 June 2005, a
selection of the most important, practicable and critical (in the sense that they can
be modified from within the system) factors was made by the project team to
achieve a concise list of the best factors. The resulting systematic of critical suc-
cess factors is presented in Figure 5.1.
It was possible to identify a relatively concise list of one to three most important
success factors for each of the six phases of pharmacovigilance. For data collec-
tion, it is most important to have sufficient and high quality data. Requirements of
soundness, speed and the associated workload shape the data management. In
signal detection it is most important to come to a sound result. For safety issue
assessment, the co-operation with partners, especially with the other agencies, but
also the access to external experts are most important. Speed was the most rele-
vant factor for decision-making. Finally, in the area of communication and action to
protect public health, the speed with which decisions are implemented, the har-
monisation of communications and the outcomes of regulatory action are perceived
the most relevant critical success factors.
It appeared that besides factors for each phase of pharmacovigilance, a number of
general factors are important for the performance of the system. These are the
legal framework conditions which the agencies and other stakeholders have to
comply with, resources in terms of staff and technical equipment, co-operation and
collaboration (which is again related to the respective duties of information ex-
change etc. made by the legal framework), if pharmacovigilance is integrated into
the larger strategy for Public Health, and the quality management within the agen-
cies.
The systematic of success factors was used as the basis to assign the perform-
ance indicators (see chapter 6).
Figure 5.1. Systematic of critical success factors
Critical success factors:
Legal framework conditions

Sufficient number of staff
Expertise
Technical resources
General factors
Co-operation
PhV embedded in larger Public Health
strategy
Collaboration with stakeholders
General quality management
Have sufficient data

Data collection
Quality of data
Soundness (Reliability)
Data
Speed
Management
Phases of Workload
Pharmaco-
vigilance Signal detection Soundness
Co-operation
Safety issue
assessment Access to external experts
Decision-making Speed
Speed of implementation
Communication/
Action Harmonisation of communications
Outcomes of regulatory action

6 Performance indicators
The performance indicators were developed on the basis of a draft list derived from
the literature as well as from the agency interviews. This list can be found in Table
6.1 below; it was organised in a similar structure as the critical success factors in
chapter 5 and was also subjected to a Delphi survey with the advisors as evalua-
tors. They were asked to rate the draft indicators according to the questions
• "Relevance: How important is the indicator to obtain a valid picture of the per-
formance of the European System for Pharmacovigilance?"
• "Practicability: How easy is it to obtain the data for this indicator?", and
• "Interpretation: How easy is it to interpret the results?"

The indicators that were assessed the best by the experts in the three aspects
(relevance, practicability, interpretation) are listed in the following table (Table 6.1).
The full results of the Delphi survey on performance indicators can be found in An-
nex 4.
The performance indicators built the backbone of the written agency survey within
which data for as many indicators as possible were collected.
Table 6.1. Most important performance indicators

1 Data collection
• Total number of ICSRs from your country received in last year
• Number of ICSRs from your country received in last year from MAHs
• Number of ICSRs from your country received in last year direct from HCPs
• Number of ICSRs from your country received in last year direct from pharmacists
• Number of ICSRs from your country received in last year direct from other HCPs
• Number of cases received/total number of ICSRs from your country
• % of serious ICSRs from your country
• Rating-scale: Usefulness of routine data from your country for safety issue assessment
compared to other information (very useful…only marginally useful)
• Rating-scale: Access to all necessary data (very easy…very difficult)
• % of PSURs that comply with E2C
2 Data management
• Number of ICSRs processed
• Rating-scale: Internal cooperation within agency incl. IT staff (very good…very bad)
• Rating-scale: Time between data entry and transmission to EMEA or MAH (adequate …
far too slow)
3 Signal detection
• Rating-scale: Information for signal detection (always sufficient…often very incomplete)
• Data sources routinely used for signal detection (routine data, literature, registries…)
• Rating-scale: Available statistical tools for signal detection (always adequate…often very
inadequate)
• Rating-scale: Time between detection of signal and reporting (publishing) (adequate …
too slow)
• Rating-scale: Work that is done within your country and at the same time in other MS or
on EU level (very little…very much)
• Rating-scale: Use of information from other agencies (in nearly all cases…very seldom)
4 Safety issue assessment

• Number of PSURs assessed
• Rating-scale: MAHs compliance with duty to assess safety issues (very good…very bad)
• Rating-scale: Availability of external expertise in your country for routine cases (always
when necessary…very scarce)
5 Decision-making
• Rating-scale: Come to adequate decisions (for NAPs/MRPs/CAPs) (always…seldom)
• Rating-scale: Come to decisions in good time (for NAPs/MRPs/CAPs) (always…seldom)
6 Communication/Action
• Rating-scale: Time from 1st signal to action with respect to this safety issue (adequate …
too slow)
• Rating-scale: Implement decisions in good time (for NAPs/MRPs/CAPs) (al-
ways…seldom)
• Rating-scale: Reaching targets for timing of communications (very good…very bad)
• Number of information events for HCPs with participation of agency
• Number of responses to inquiries by HCPs
• Number of other answered queries
• Number of inspections of MAHs carried out where PhV was an issue (at least partially;
including inspections that were carried out by other authorities in the country)
• Rating-scale: Consistency of communication across stakeholders (incl. MAHs) (very
good…very bad)
• Number of ICSRs from your country before vs. after communication
• Total reporting rate per million inhabitants in 2004
• Reporting rate in children per million inhabitants in 2004
• Number of market withdrawals of drugs (compared to other countries)
• Number of suspensions of marketing authorisation
• Number of dear doctor letters sent
• Number of changes in SPCs made
• Number of applications for variations adopted/refused
• Incidence of ADR-relevant diseases
• Hospitalisations due to ADR
• Mortality due to ADR
• Number of quality-adjusted life years lost due to ADRs
• Potential years of life lost due to Adverse effects from medicines
• Changes in consumption data
• (Change in) Prescription data (controlled for population parameters)
7 General factors
• Number of staff in full-time-equivalents
• Number of scientists in full-time-equivalents
• Annual budget of the agency
• Number of Regional centres in your country
• Total number of staff (sum of all regional centres) for routine work
• Number of nationally authorised products in your country
• Number of MR authorised products in your country
• Number of centrally authorised products in your country
• Pharmaceutical consumption by drug classes
• Number of documents prepared (legal acts, guidelines)
• Number of scientific publications with at least one author from the agency in last year
• Rating-scale: Compliance of agency with dates/requirements (very good…bad)
• Rating-scale: Meeting general targets for timing (very good…very bad)
• Rating-scale: Compliance of MAHs with 15 days (very bad…very good)
• Rating-scale: Compliance of MAHs with legal requirements (very bad…very good)
• Number of documents sent through EudraNet (RAS, NUIS, others) by sender, concerned
MS, issue, channels
• Number of regular meetings with external experts

• Number of irregular consultations with external experts
• % of staff trained per year
• Number of training measures (internal or external) with at least one participant from the
agency
This set of performance indicators has been carefully developed to cover the rele-
vant areas of the pharmacovigilance system including the critical success factors.
However, it also has its shortcomings, the most important of which is that the out-
comes of communication and action cannot be adequately measured with the ex-
isting data in an economic and valid way. The experiences with the agency survey
showed that for the following indicators it was difficult to determine the necessary
data:
• Pharmaceutical consumption by drug classes: Data are frequently not
available.
• Rating-scale: Usefulness of routine data from your country for safety issue
assessment: It was unclear, how the agencies understood the term "useful-
ness".
• Rating-scale: Information for signal detection sufficient: A comment was
that in pharmacovigilance information "is never sufficient", while 6 agencies
answered that information had always been sufficient, probably pointing out
an unclear understanding of "sufficient information."
• Rating-scale: Work that is done within your country and at the same time in
other MS or on EU level: Different opinions seem to exist of what tasks are
necessary to be done on the national level and what competences should
or can be transferred to central structures.
• Rating-scale: MAHs compliance with duty to assess safety issues: Informa-
tion from the interviews indicate that the processes at the MAHs are not suf-
ficiently transparent for the agencies, partially because only few inspections
are made. Therefore it is questionable if the agencies can validly assess the
compliance of the MAHs with signal detection duties.
• The number of responses to inquiries by HCPs, number of other answered
queries, and number of documents prepared are often not documented.
• For the impact: The number of ICSRs from your country before vs. after
communication often not documented;
• Reporting rates are difficult to interpret because they are input factors for
the system but partially also the output of approaches to improve reporting
by education of the reporters etc.
• The number of market withdrawals is difficult to interpret because these re-
sult from different reasons including internal decisions within the MAHs and
causes other than safety concerns.
• Outcomes of action: Missing data on hospitalisations and mortality, QUA-
LYs and life years lost due to ADRs because of lacking prospective studies
with such endpoints.
Most of the indicators, however, proved to be practicable and useful to describe the
different aspects of the system and come to meaningful results.
7 Case studies
The two case studies were carried out to obtain a deeper understanding of the
processes that underlie the detection and assessment of safety signals and how
they lead to decisions made.
7.1 Statins case

This safety case provides us with a number of learning points: Firstly, problems like
rhabdomyolysis, the extent of which is hard to determine in a pre-clinical setting or
from clinical trials, rely heavily on the optimal use of post-marketing approaches,
including appropriate use of spontaneous reporting data, record linked databases
and the like. With respect of systematic pooling of and signal detection out of spon-
taneous reports the successful implementation of the EudraVigilance network is
essential. Moreover, priority should be given to develop further European pharma-
covigilance/pharmacoepidemiology data platforms. In various MS (e.g. UK, Den-
mark, Portugal, and the Netherlands) significant progress has been made, but
more action is needed in order to stay at the cutting edge. The case shows repeat-
edly the vulnerability of a medicine when inappropriate dosing (directly via a too
high dosage form (cerivastatin 0.8 mg) or indirectly via a pharmacokinetic interac-
tion due to co-prescription with fibrates) occurs. Inappropriate dosing has at least
two angles: firstly, are we introducing new medicines on the market with clinically
the most suitable dose. Two independent studies have revealed data that this is
still not the case13,14. Moreover, inappropriate dosing as a result from poor pre-
scribing and non-adherence with label directions is a major problem related to drug
prescribing and taking behaviour. So far we have little information about what pre-
scribers do with label information and label changes. Moreover, when information
is there, the results are not very encouraging15. Therefore, effective strategies for
risk communication towards prescribers should be a topic that should feature on
any agenda of risk management strategies. A typical problem related to the statins
is also the fact that dosing-dynamics in this drug class is driven by the dominant
paradigm to achieve the strongest cholesterol reduction as fast as possible. This
paradigm, supported also by myriad clinical trials, is evidently misused in drug
promotion and marketing. How this will impact future drug utilization and safety
issues could be a focus of further investigation.
13 Cross J, Lee H, Westelinck A, Nelson J, Grudzinskas C, Peck C. Postmarketing

drug dosage changes of 499 FDA-approved new molecular entities, 1980-1999.
Pharmacoepidemiol Drug Saf 2002; 11: 439-446.
14 Heerdink ER, Urquhart J, Leufkens HG. Changes in prescribed drug doses after
market introduction. Pharmacoepidemiol Drug Saf 2002; 11: 447-453.
15 Wilkinson JJ, Force RW, Cady PS. Impact of safety warnings on drug utilization:
marketplace life span of cisapride and troglitazone. Pharmacotherapy 2004;
24: 978–986.
7.2 SSRI case

In terms of contents this safety case provides us with a number of learning points:
Firstly, the nature of the adverse event can have a very high impact on the regula-
tory process: it is rare; no completed suicides were reported in 4100 children in-
cluded in trials, and associated with the indication of prescribing16. Furthermore, in
these types of adverse events it is hard to discern whether there is a class or a
single drug effect. Three major limitations that have been identified in the assess-
ment of safety in child and adolescent psychopharmacology can be applied to this
case as well17:
Considerable inconsistency in the way safety is assessed. Defining the adverse
outcome is difficult. Since completed suicides rarely occur in trials, other markers
must be used. An example of the problems in definition is the term ‘emotional labil-
ity’ that was used in clinical studies included in the initial submission to the FDA in
2002. Apparently, behind this term lay factors related to the outcome of interest,
suicidal behaviour.
Dearth of research. Studies into the topic of suicidal behaviour in children during
antidepressant use are still few. To further understand this topic, especially in rela-
tion to detecting future problems at an earlier stage, more research is required.
Improved identification of adverse events. In this case the main source of data was
clinical trial data. Observational data was only used in a supportive way; spontane-
ous reporting data did not play a significant role in the regulatory decision process.
With regards to the latter improved data-mining techniques and institutionalized
follow-up procedures may help to make better use of available research, hopefully
leading to better, and earlier, signal detection.
When managing drug safety issues, considering the impact of action taken on pa-
tients is of key importance. Abrupt discontinuation is often unwanted and requires
monitoring of patients after the announcement of the safety alert. This does not
only address warnings issued to HCPs but particularly publicly available informa-
tion on safety concerns, be it from official bodies, MAHs or unofficial sources. The
SSRI case shows that it is not sufficient to inform HCPs about the risks of abrupt
changes in the prescription of a drug but that the patients should be prevented
from stopping to take the drug after a public warning without consulting their physi-
cian. Moreover, especially when a withdrawal affects a significant part of treated
patients the impact can be very large. For example, when the CSM advised against
the use of most SSRIs in children and adolescents in the UK, it was estimated that
half of the antidepressants that were used in this population were of the group that
was considered ‘unsafe’. The risk for withdrawal reactions has also been described
in children18, and was noted in the 26 April 2005 EMEA press release on the re-
16 Vitiello B, Swedo S. Antidepressant medication in children. NEJM 2004; 350:

1489-1491.
17 Vitiello B, Riddle MA, Greenhill LL, March JS, Levine J, Schachar RJ, et al. How
can we improve the assessment of safety in child and adolescent psy-
chopharmacology? J Am Acad Child Adolesc Psychiatry 2003; 42: 634-641.
18 Diler RS, Avci A. Selective serotonin reuptake inhibitor discontinuation syndro-
me in children: six case reports. Clin Ther Res 2002; 63: 188-197.
view of antidepressants in adolescents and children19. Therefore, effective regula-

tory management of post-event (e.g. safety restrictions) drug use is warranted.
Great lack of efficacy and safety data of medicines (e.g. SSRIs) in children and
adolescents is acknowledged previously but repeatedly received many echoes in
this case. In September 2004 the European Commission adopted a proposal for a
Regulation of the Council and of the Parliament on Medicinal Products for Paediat-
ric Use (see http://pharmacos.eudra.org/F2/Paediatrics/index.htm), the overall pol-
icy objective is to improve the health of the children of Europe by increasing the
research, development and authorisation of medicines for use in children. Studies
shall be funded that lead to the development and marketing of drugs for children.
The long-term follow-up of adverse drug reactions would be an additional require-
ment for marketing authorisation.
Recently WHO has delivered a relevant review on this topic in the context of the
'Priority Medicines for Europe and the rest of the World'-project20. Detailed discus-
sion of this topic falls outside of the scope of this case study, but drug use in chil-
dren and adolescents will be increasingly prominent in the regulatory environment
in the coming years.
19 EMEA press release 25 April 2005. European Medicines Agency finalises re-
view of antidepressants in children and adolescents. Available from:
http://www.emea.eu.int/pdfs/human/ press/pr/ 12891805en.pdf (Accessed 14
September 2005).
20 Zuidgeest MGP, Willemen MJC, Van den Anker JN. Pharmaceuticals and chil-
dren - Background paper for the Priority medicines for Europe and the world
report. Available from:
http://mednet3.who.int/prioritymeds/report/background/children.doc (Accessed
14 September 2005).
8 Best practice
Many interesting approaches to solve at least some of the issues that are dis-
cussed within the system have been collected from the literature and even more
from the interviews with the national agencies. On the national level, some of the
problems have been resolved by measures which could partially serve as models
for the whole EU system.
8.1 Indicator-based identification of best-practice

8.1.1 Method
In this section we try to identify important differences in critical success factors be-
tween countries and possible explanations for them. Hence comparisons are made
between the assessment of the critical success factors and indicators for the input
and process, like resources or actions. For each phase, only the countries with the
best assessment are listed in the tables, as possible examples of "best practice".
To get a better overview only a medium-sized set of indicators for the respective
phases is presented. Therefore, only indicators with
• a high relevance according to the Delphi survey on performance indicators

and
• either some explanatory power or surprisingly little significance – according

to our deeper analysis with a more comprehensive set of indicators –
were included. The average value (Median) of all 28 countries plus EMEA (not only
for those mentioned in the tables) is given as a comparison according to which the
assessment of performance was made. The criterion for good performance was a
better-than-average assessment in the "output" variables, the "process" and "input"
variables are used to describe which factors might have contributed to this good
evaluation.
However, it should be noted that drawing conclusions based on correlations be-
tween the indicators is not fully adequate, because a) these correlations cannot be
tested statistically due to the small sample size, b) we do only have a tentative un-
derstanding of which features can be the causing factors, and c) a large number of
different factors (such as different institutional settings) has influence on the per-
formance of the complex system of pharmacovigilance not all of which are known.
Moreover, d) not all potentially important factors/outcomes could be measured or
can be indicated without impairing the clarity of the presentation. But of course
these other factors and additional information e.g. form the agency interviews are
used for the interpretation as far as possible.
8.1.2 Data Collection

In the phase of data collection key success factors are the sufficiency and the qual-
ity of the information basis. Possible indicators for these characteristics are as-
sessments of the usefulness of routine data and the preparation for a safety crisis
by routine data. The ratings found for these items vary in the whole scale range.
In Table 8.1 those countries are listed as examples of good practice, which come
off better than the median in one of the two output indicators and are as least as
good as the median in the other. E.g., looking at the second column of the table, in
the questionnaire the countries were asked how well they had been prepared by
routine data for the last crisis on a five-point-rating-scale ranging from 1="very
good" to 5="very bad". The median for all countries was the middle value 3, the
analogue is true for the assessment how useful the routine data are. Therefore in
Table 8.1 all countries are listed which assessed their respective capability with at
least "2" in one of these two parameters and at least a "3" in the other. The follow-
ing columns contain process and input variables that might contribute to the good
evaluation of the two output variables.
The following tables are structured in the same way.
While especially larger western countries are quite confident with their data basis,
there is much criticism in the self-assessments of some smaller eastern countries
(Table 8.1). Most of the confident countries have a quite high amount of data avail-
able, e.g. the number of received ICSRs and their reporting rate per sale 2004
mostly lie above the median. However there are some exceptions of well-
performing New Member States like Hungary, Estonia or Malta. However, the as-
sessment of the usefulness of the data seems to be lower for countries which have
received a very limited number of ICSRs in 2004 which would prevent the statistical
analysis of these data on a national basis. High reporting rates as for DE-BFARM,
in FR or in NL lead to a very good assessment.
In respect to resources there seems to be no clear connection to the usefulness of
the data. This is of course not surprising, but a few countries with high resources
even do not obtain a high amount of reports. Countries with a medium-size staff-
per-population rate (the whole staff is given in the table including the national plus
eventual regional centres in the country) are performing equally well as or even
better than countries with extraordinarily high staff, but also a number of agencies
with extraordinarily low staff can perform well (e.g. DE-PEI, IT, EMEA). In most of
the well-performing larger countries regional centres support the data collection.
Table 8.1: Indicators for Data Collection
Output
Process Input
Data
Data Amount Resources
Usefulness
PhV-
Reporting Reporting RC per-
Routine
Routine
ICSRs
rate total rate total
Assess
forms staff21
data re- IT- NCA+
data 2004 per 2004 per data coll./
prep. for ceived re- RC per
useful sales in million manage-
crisis 2004 sources million
US$ capita ment
capita
BE 3 2 2945 9.26 283.27 3 0 0.8
DE-
1 3 15750 38.6 190.84 2 1 1.0
BFARM
DE-PEI 1 1 3376 8.27 40.91 1 1 0.1
EMEA/
2 2 . . . 2 . 0.1
EEA-28
EE 2 3 61 . 45.15 3 . 0.7
EI 2 2 1727 6,67 428,78 3 . .
FR 2 2 20116 35.29 335.82 3 1 1.7
HU 3 2 234 0.79 23.13 2 . 0.3
IT 2 2 6350 13.12 109.69 2 . 0.2
MT 2 2 32 . 80.02 2 . 4.5
NL 1 1 5050 18.3 310.62 2 1 1.8
PT 2 3 1718 5.69 164.01 2 1 1.7
SE 2 2 4124 12.53 459.46 2 1 4.2
MEDIAN
all coun- 3 3 1491 8.27 152.93 2 1 1.2
tries
1= 1 = to-
1 = very
very tally 1 = yes
good
useful sufficient
rating
scale 5=
5 = very
5= very mar-
insuffi- 0 = no
bad ginally
cient
useful
8.1.3 Data Management

Reliability and speed are the most important goals in the phase of data manage-
ment. As the former factor can hardly be analyzed with the data from our surveys,
the latter shows huge country differences. The assessment time for PSURs varies
between one and ninety days across the countries. Surprisingly there is less varia-
tion in the subjective assessment of the time for data processing. As already shown
21 Staff for pharmacovigilance, scientific and administrative.

in chapter 3, the majority of agencies considers the time in which they do this task
as adequate. But there are some worrying exceptions in which the time is as-
sessed very negatively (not given in the table). These cannot be explained by the
workload alone (absolute numbers of ICSRs or PSURs as well as reporting rates).
But in combination with the input indicators it can be concluded, that a lack of hu-
man and IT-resources often results in an insufficient duration of the process. As the
following Table 8.2 shows, some countries with few resources also provide good
results. Listed are countries with process duration self-assessed as rather ade-
quate.
The well-performing countries are here defined as those which assess the time
they need for data processing as adequate (value 1) or with a value of 2 and at the
same time having assessment times for PSURs of maximally 10 days. These
agencies mostly have good IT resources. Having all data under one interface, the
number of persons working for this task (not shown in the table), and the involve-
ment of regional centres seem not to influence the time for data processing.
Table 8.2: Indicators for Data Management
Output Process Input

Duration of process Instruments/ Workload Resources
Duration All data RC per-
Assess ICSRs
assess accessible Assess IT- forms data
time data received
PSURs under one resources coll./ man-
processing 2004
(days) interface agement
BE 1 , 0 2945 3 0
DE-BFARM 1 20 0 15750 2 1
DE-PEI 1 36 1 3376 1 1
DK 1 3 1 1920 2 ,
EE 1 4 , 61 3 ,
EI 1 , 0 1727 3 ,
ES 1 15 0 7476 2 1
FI 1 , 0 1118 2 ,
HU 1 1 0 234 2 ,
IC 1 , 0 26 2 ,
IT 2 10 1 6350 2 ,
NL 1 35 0 5050 2 1
NO 1 30 0 1734 1 1
SE 1 3 1 4124 2 1
UK 1 40 0 20044 1 1
MEDIAN 1 30 0 1491 2 1
1 = ade- 1 = totally
rating quate 1 = yes sufficient 1 = yes
scale 5= far too 5 = very
slow 0 = no insufficient 0 = no

The main goal of this phase – to identify all relevant signals – is not adequately
measurable. However the analysis of the available data sources, human resources
and tools gives some interesting insights.
In the agency questionnaire it was asked which other data – besides ICSRs and
PSURs – are or could be used for signal detection or safety issue assessment,
including
• routine collection of data or information on post-authorisation safety studies,
• routine collection of data or information on phase IV efficacy trials,
• routine collection of data or information on preclinical studies,
• if and how often sales data are used,
• if the agencies have the capability to link prescription registries with other
registries which include health outcomes,
• if the agency initiated or carried out in 2004 ad hoc pharmacoepidemiology

studies when a signal needed confirmation or quantification, and
• if the agency initiated or carried out in 2004 pharmacoepidemiological stud-
ies for early post-marketing surveillance of new products.
In Table 8.3 countries with availability of at least seven of eight data sources are
presented as examples of good practice (the seven data sources in the table below
plus information from other authorities which is not presented in the table).
The availability of different data seems to correlate strongly with the country size
(none of the small countries met the inclusion criteria for this analysis); small coun-
tries seem to lack the access to these "additional" data.
Table 8.3: Indicators for signal detection – Availability of data sources
Availability of data sources

Info pha- Info pre- Studies Studies
Info Link pre-
se IV clinical Used sales signal early
PASS scription
studies studies data PM PM
collected registries
collected collected surveil. surveil.
DE-PEI 1 1 1 1 0 1 1
EMEA/ EEA-28 1 1 1 2 0 , ,
FI 1 1 0 2 1 1 ,
IT 1 1 1 2 1 1 1
SE 1 1 0 2 1 1 1
UK 1 1 1 2 1 , ,
MEDIAN 1 1 0 2 0 0 0
1 = yes 1 = yes 1 = yes 2 = always 1 = yes 1 = yes 1 = yes
rating scale
0 = no 0 = no 0 = no 0 = never 0 = no 0 = no 0 = no
Table 8.4 shows the minority of countries, which assess their available statistical
tools as at least adequate; in respect to statistical tools the majority considers their
equipment as (very) inadequate. All of the better-performing have tools for small
numbers of cases. Overall there is quite high correlation between these two indica-
tors (assessment of statistical tools and having tools for small numbers of cases)
which underlines the importance of statistical tools for small numbers of ICSRs.
Table 8.4: Indicators for signal detection – Analysis tools and resources
Analysis tools Resources
Assess avail. Tools for small RC performs

statist. tools numbers signal detection
EMEA/ EEA-28 2 1 ,
DK 2 1 ,
NL 2 1 1
UK 1 1 0
MEDIAN 4 0 1
1 = always
1 = yes 1 = yes
adequate
rating scale
5 = often very
0 = no 0 = no
inadequate
In the survey it was not distinguished between staff for signal detection and staff for risk assessment
Taken the availability and assessment of the analysis tools and the availability of
the "additional" data mentioned above together, there are obvious differences es-
pecially within the medium-sized countries. Particularly well performing are Finland,
Italy, and Sweden, among them Sweden with nearly the highest national staff for
pharmacovigilance, but also Italy with relatively low staff resources. It is interesting
to note that the well-performing countries in many cases have by far more staff for
signal detection (not shown in the table) than they have for data management,
probably due to particularly good IT resources that allow shifting staff from the ear-
lier stages of pharmacovigilance to the later.
8.1.5 Safety Issue Assessment

To generate the necessary knowledge for an adequate safety assessment either
in-house or external expertise is indispensable. Fortunately a lot of countries state
to have good access to both sources of expertise, especially the larger countries.
On the other side there are some small and medium-sized countries with serious
problems in this issue. Not surprisingly this is reflected on the input side by a small
number of staff for risk assessment in these countries (not shown in the table). In
addition, small countries mostly state insufficient compliance of MAHs with their
duty to assess signals. As can be seen in Table 8.5 there are also some excep-
tions of medium-sized countries with better conditions, e.g. Norway.
The table includes countries which state that they can identify/assess signals with-
out help and have easy access to external expertise on a routine basis and in ex-
ceptional cases as examples of good performance.
Table 8.5: Indicators for safety issue assessment
MAH
Country expertise Resources
compliance
Receive Identify/
Support from MAHs comply RC performs
support from assess
experts analysis of safety issue
experts signals
exceptional signals assessment
routine without help
BE 2 2 1 3 0
DE-PEI 1 1 1 3 1
DK 1 1 1 2 ,
EMEA/EEA-28 2 2 1 2 ,
FR 2 1 1 3 1
HU 2 2 1 1 ,
IT 2 1 1 2 ,
NL 2 2 1 3 1
NO 2 1 1 3 0
SE 2 1 1 2 1
UK 1 1 1 1 0
MEDIAN 2 2 1 3 1
1 =always
1 =very easy when 1 = yes 1 = very good 1 = yes
rating scale necess.
5 = (nearly) 5 = (nearly)
0 = no 5 = very bad 0 = no
impossible impossible
In the survey it was not distinguished between staff for signal detection and staff for risk assessment
8.1.6 Decision-Making
As already shown in chapter 3.7.5 the majority of the countries assess the ade-
quacy and duration of the decision-making process as rather positive.
In Table 8.6 only those countries are listed as particularly good performers, which
are satisfied in respect to decision-making for NAPs and MRPs. According to the
interviews, these countries do not have a common decision process which would
have allowed identifying advantageous commonalities, and also the equipment
with staff for this task varies a lot between agencies. Therefore it is very difficult to
conclude here what leads to this aspect of best practice.
The responses are quite homogeneous for the different types of products within
countries: in a number of countries the adequacy of decisions is evaluated nega-
tively be it for decentrally or centrally authorised medicines, in others it is merely
positive for both types. The negative evaluation might relate to difficult consultation
or even arbitration between MS agencies that is also needed for most decentrally
authorised products, to problems regarding regulatory aspects, but also to the
general difficulty to decide on the basis of weak signals or other uncertain condi-
tions.
Table 8.6: Indicators for decision making
Decision Making
Adequate Adequate Adequate

decisions decisions decisions
found NAPs found MRPs found CAPs
BE 1 1 1
CY 1 1 1
DE-BFARM 1 1 1
DE-PEI 1 1 3
HU 1 1 1
SE 1 1 1
UK 1 1 1
MEDIAN 2 2 2
1 = always 1 = always 1 = always
rating scale
5 = seldom 5 = seldom 5 = seldom
8.1.7 Communication/Action
Critical factors in the communication with stakeholders are the consistency of
communications and actions as well as the speed of implementing them.
In Table 8.7 those countries are presented, which come off better than the median
in one output indicator (assessment of the time between signal and decision, con-
sistency of the communication between agencies, or consistency of communica-
tions between agencies and MAHs/HCPs), and are at least as good as the median
in the other two.
The consistency of the communication between is criticized by a few countries with
the main argument that the publication times of safety information is not always
coordinated well.
The variation of the assessments is quite high and cannot be explained by the dif-
ferent size or the geography of the countries. The overall correlation between the
timeliness and the responsible staff is also quite low. However it should be noted
that the countries with the highest staff (in absolute and relative numbers; not
shown in the table) are very confident with the speed in this phase, but another
strategy seems to exist that concentrates communication/action on only a few
members of the staff. According to the interviews often the director of the agency is
solely responsible or communication is centralised through a press officer. In nearly
half of the countries where regional centres exist they are involved in communica-
tion or action.
Table 8.7: Indicators for communication/action
Output Process Input

Consistency/speed Actions Resources
Consistent Inspec-
Assess Consistent RC perform
communica- Follow-up tions of
time signal- comm. communica-
tion agencies - the impact MAHs for
decision agencies tion
MAHs/HCPs PhV issue
BE 2 1 2 0 0 0
DE-PEI 2 1 2 1 , 0
DK 1 2 2 1 8 ,
EE 1 2 2 0 , ,
EMEA/EEA-28 2 2 2 0 15 ,
GR 3 1 2 0 , ,
PT 2 2 2 0 16 0
UK 1 2 2 1 61 1
MEDIAN 3 2 2 0 0 1
1 = ade- 1 = very
1 = very good 1 = yes 1 = yes
rating quate good
scale 5= far too 5 = Very
5 = very bad 0 = no 0 = no
slow bad
For the communication between the agencies and HCPs a more detailed analysis
is needful. While the differences in the opinions about the cooperation with HCPs
are only small, the assessment of the influence on prescription behaviour differs
largely. As Table 8.8 shows, this does not seem to be country-size specific or to
correlate with human resources (not shown in the table). All countries with output
ratings at least on the median are presented. The influence on the prescription be-
haviour is in general only assessed as moderate.
Table 8.8: Indicators for communication/action with HCPs
Output Input
Influence/ Resources
Cooperation
Influence on
Cooperation RC perform
prescription
with HCPs communication
behaviour
DE-PEI 3 2 0
DK 2 2 ,
EE 2 2 ,
EI 2 2 .
GR 2 1 .
IC 3 2 .
NO 2 2 0
PT 3 2 0
SE 2 2 1
MEDIAN 3 2 1
1 = very 1 = very
1 = yes
rating good good
scale 5 = very 5 = very
0 = no
weak bad
Another explanation for the weak influence of communications or actions could lie
in the amount of actions or in the usage of certain communication channels. Most
countries provide general and specific information to different groups of HCPs. But
it is conspicuous that this does not apply for some countries with negative assess-
ments for the above output indicators for communication/action, which seem not to
use all available channels. In the respective Table 8.9 the same countries as above
are listed, that is they assess the influence on the prescription behaviour as well as
the co-operation with HCPs to at least moderately good (value 3).
Table 8.9: Indicators for information provided to HCPs
Communication Communication
General Infos to Stakeholders Specific Infos to Stakeholders
Gen- Gen- Spe- Spe- Spe-
General Spe- Spe-
eral eral Gene- Gene- cific cific cific
info cific cific
info info ral info ral info info info info
Profes- info info
Individ- Medical Phar- Other Individ- Medical Profess
sional Phar- Other
ual asso- macists HCPs ual asso- jour-
journals macists HCPs
doctors ciations doctors ciations nals
DE-PEI 1 1 1 1 1 1 1 , 1 0
DK , 1 1 1 1 , 1 1 1 1
EE 1 1 1 1 1 1 1 1 1 1
EI 1 , , , 1 1 1 1 1 1
GR 1 1 0 1 0 1 1 0 1 0
IC 1 0 1 1 0 1 0 1 1 0
NO 1 0 1 1 0 1 0 1 1 0
PT 1 1 1 1 1 1 1 1 1 1
SE 1 1 1 1 1 1 1 1 1 1
MEDIAN 1 1 0 1 0 1 1 1 1 1
rating 1 = yes 1 = yes 1 = yes 1 = yes 1 = yes 1 = yes 1 = yes 1 = yes 1 = yes 1 = yes
scale 0 = no 0 = no 0 = no 0 = no 0 = no 0 = no 0 = no 0 = no 0 = no 0 = no
8.1.8 General aspects

Overall some cautious conclusions can be drawn from this analysis which concern
the pharmacovigilance process as a whole:
• The correlation between the human resources and indicators for outcome is
for some phases rather low. This counts especially for relative indicators
like the combined PhV-staff of NCA+RC per capita. Also if other overlap-
ping explanations like geography are kept in mind, it is hardly possible to
identify causal connections between resources and outcomes. However,
the better performers generally have a certain minimum of staff and assess
their IT resources as more positive than the agencies do that perform less
well.
• Some countries are performing well in almost all phases and therefore are
frequently presented in the tables above as examples for best practice.
However it is not easy to determine the causes for this overall good per-
formance as it is certainly impossible to evaluate all critical success factors
together and because most output indicators are subjective ratings. In addi-
tion, external conditions have a high impact on the performance of the
agencies within the whole system of pharmacovigilance as are e.g. the
agency's budget, compliance of MAHs etc.
• Regional centres seem to be a very helpful support for the work of the na-
tional centres especially in data collection and communication. The regional
centres assist the national agencies particularly in the first phases, so that
the latter are able to concentrate more on tasks like signal detection and
safety issue assessment. The output indicators show some success signals
for this strategy. The positive appreciation of regional centres is confirmed
in the interviews with the corresponding countries.
8.2 Results from the interviews

This section identifies "best practice" from strong points in the national systems
that were identified on the basis of the interview partners' statements. Because of
the large quantity and diversity of the approaches in the NCAs the presentation
cannot be complete. Again, the selection of these results depends on the evalua-
tion by the evaluators of their relevance for the critical success factors. The results
are organised according to the phases or pharmacovigilance and respective critical
success factors.
8.2.1 General factors
8.2.1.1 Legal framework conditions
Germany has good experiences with the implementation of the law in the form of a
stepwise procedure including risk assessment, decision-making and communica-
tion. Other examples of supporting national legal frameworks are Slovenia and
Lithuania where the new harmonized legislation and the main points of Pharma-
covigilance are explicitly mentioned in the law. Spain has a legal obligation of
HCPs to report ADRs which cannot be controlled but is assessed as good to have,
although not sufficient, because it at least shows that PhV is an important issue.
In Italy for reimbursement of new drugs nearly always the conduction of a PASS or
other monitoring measure is required, prescription can be limited by issuing so-
called "AIFA-notes".
8.2.1.2 Sufficient number of staff
The Irish Medicines Board has set up a detailed plan on how much personnel are
needed for the different work steps from which other countries could learn.
8.2.1.3 Expertise
A number of agencies stress the importance of their long and good collaboration
with the WHO-UMC in submitting to and using the database of ICSRs for signal
detection as well as participating in training measures there.
The Cypriot agency has easy access to external experts through a system of gov-
ernmental HCPs.
The German PEI made positive experiences with ad-hoc expert groups at which
practitioners are easily won to participate. A standing committee of the Drugs
commission of the German physicians' association is also often a helpful partner.
In the Czech system, HCPs are traditionally used to provide data/statistics and
therefore has – compared to other smaller countries – relatively high reporting
rates.
8.2.1.4 Technical resources
Some of the agencies made high investments in new IT infrastructure, as e.g. the
UK, Ireland, Cyprus, Germany-BfArM, partially already in preparation of future re-
quirements regarding the planned data warehouse.
8.2.1.5 Co-operation
Co-operation with regional centres for pharmacovigilance is well-established e.g. in

France, where different meetings/workshops with MAH on PhV in the regional cen-
tres. A monthly meeting of AfSSAP with the RCs called "technical committee" is
held to review reports, signals, and publications.
The decentralized system has increased number and quality of reports and gener-
ally the contact to HCPs in Spain and in the UK. As in France, regular meetings on
technical questions are held 4 to 5 times p.a. with all Spanish RCs. A safety com-
mittee on human medicines exists consisting to one half of members nominated by
RCs and the other half by the national agency.
The regional organisation is also well-established in Sweden, good collaboration
exists between the agency in Luxemburg and the regional centre in Nancy/France.
Without having regional centres, the external clinicians working together with the
NCA in a pharmacovigilance committee are also used to promote reporting in their
hospitals in Slovakia.
The co-operation structures within the national agencies differ from country to
country. A close collaboration of PhV with the department that registers drugs is
assessed as helpful in Hungary. Separate department for pharmacovigilance and
marketing authorisation are stressed by the Belgian and other agencies, whereas
in Ireland pre- and postmarketing departments are dissolved to allow closer col-
laboration in shared units or working groups. It is appreciated to have the staff for
human, veterinarian medicines and devices in the same agency, and small coun-
tries can have advantages in easy formal and informal collaboration with experts
and other institutions.
In Sweden a preauthorisation-evaluation exists for the preauthorisation-planning
under inclusion of the PhV-department.
The Polish agency will support the interdepartmental exchange of information with
a software tool used by the whole agency.
8.2.1.6 Collaboration with stakeholders
The Finnish agency assesses its contacts to the HCPs as particularly good; PhV
and the agency in general have a good reputation in the media, they are trusted by
all parties. The same is true for Ireland where extensive contacts are nurtured with
the MAHs and industry associations. Comprehensive discussions with MAHs are
described by the German PEI.
Cyprus and the Czech Republic can draw on good collaboration with Medical Ser-
vices and medical societies as a whole that are approached as multipliers; actions
(e.g. contraindications) are discussed with practitioners which results in good feed-
back on factors that might otherwise have been underrepresented, e.g. costs of
different forms of application, reimbursement, distribution conditions, health insur-
ance companies. Contact to medical societies can improve the influence on clinical
guidelines and thus probably on the prescription behaviour
8.2.2 Data collection
8.2.2.1 Have sufficient data
The German PEI and the Italian agency can initiate additional post-authorisation
safety studies. In the Czech republic a database similar to UK GPRD, namely a
voluntary registry of medical records is available, as are health statistics, data from
vaccination programmes, on abuse of drugs etc.
In other interviews, the registry system of Sweden and a large amount of epidemi-
ological data in Spain were underlined.
8.2.2.2 Quality of data
In Denmark the provisions of "Volume 9" on PSURs have been translated into a
national guidance, after that the quality of PSURs improved.
Seminars on pharmacovigilance and regulatory activities for practitioners are of-
fered in several countries, e.g. in Italy where credit points for continuous medical
education can achieved this way; in Germany seminars on ADRs of vaccination are
offered.
8.2.3.1 Soundness
Sweden has specific data mining tools for signal detection.
8.2.4 Decision-making
Decisions in Poland are sometimes faster than on EU-level, and decisions in Spain
are made by a committee that is independent from the committee that decides on
MAs.
8.2.5 Communication and action to protect public health
8.2.5.1 Speed of implementation
Sweden has provisions for particularly fast action.
8.2.5.2 Harmonisation of communications
The Danish agency has recently changed its departmental structure to improve the
transparency and improve information for the public.
Poland publishes all SPCs on the internet. The internet is also used in Finland to
automatically forward PhV-information to the website of the Finnish Medical Soci-
ety, and Sweden – as other agencies – can use the agency's press office for
pharmacovigilance issues.
9 Discussion of strengths and weaknesses of the

European system of pharmacovigilance
The following strengths and weaknesses of the system summarize the results of
the empirical studies above. Additional items are selected from the respective
question in the agency interviews in order to represent the perspective of these
"internal experts" as far as possible. However, the project team was fully responsi-
ble for the appraisal of all the results and for the selection of the points that are
emphasized in the following paragraphs. The order of the strengths and weak-
nesses is again according to the phases of pharmacovigilance and respective criti-
cal success factors plus an advancing paragraph on general factors.
9.1 General factors

9.1.1 Legal framework conditions
The strengths and weaknesses of the European PhV System relating to the regula-
tory system can be summarised as follows:
Strengths of the PhV System Weaknesses of the PhV System
• The system harmonises regulation, • The system is complicated and diffi-
pharmacovigilance practice, prod- cult to understand (many responsi-
uct information, communication and ble authorities; different procedures
action across MS. and responsibilities for MRPs;
• International co-ordination leads to NAPs; CAPs)22.
stronger power of regular action. • The system is very difficult to over-
• The system for CAPs is straightfor- look despite the existence of de-
ward, rapid, rational and comes to tailed guidances. This makes it diffi-
binding decisions. cult to find out the steps to do in a
• The system will allow Pharmacovigi- particular situation especially for
lance Planning (E2E) incl. a more smaller agencies with less special-
proactive approach by agencies ized staff for regulatory affairs.
and MAHs. • Different implementation is caused
by e.g. diverging health systems in
the MS.
• Existing instruments are not fully
applied, especially in the control of
the MAHs' compliance with re-
quirements. Not all agencies have
all guidances for all phases of
pharmacovigilance in place.
• The use of assessment reports from
other countries sometimes ham-
22 See also Bendall 2004.


pered by confidentiality issues.
• The integration of the new MS is still
problematic and impairs the full
functioning of the whole system;
new MS are not yet able to fully con-
tribute to the PhVWP.
• Different opinions with respect to
which minimum of information/dis-
cussion is necessary at national
level.
• The collaboration of CHMP with
PhVWP is assessed by a number of
agency representatives as subopti-
mal (e.g. parallel/duplicate work,
uncoordinated decisions).
• Long-lasting discussions take place.
• The cooperation with academia is
weak (causes: confidentiality of
data; lack of funding).
• The weight on the system for spon-
taneous reports is too strong despite
the high relevance of studies in re-
cent crises. Some stakeholders
doubt that the new legislation includ-
ing the Clinical Trials Directive offer
sufficient means to yield the neces-
sary prospective safety studies.
The analysis has shown that the current European System of Pharmacovigilance
has achieved an advanced state of development. If implemented reasonably, from
November 2005 onwards the recent reform will give the authorities additional tools,
as well as greater scope for urgent regulatory action, increase transparency on
safety issues and facilitate communication. Moreover, it will allow a more proactive
approach to pharmacovigilance.
9.1.2 Sufficient number of staff

With respect to the number of staff available, the strengths and weaknesses of the
European PhV System are the following:

• There are agencies which have – • Low budgets are available for phar-
according to their self-assessment macovigilance in some agencies.
– sufficient staff for their pharma- This hampers the number of avail-
covigilance tasks, a number of able staff (partially because well-
agencies have made calculations educated staff cannot be won with
how much staff they need to com- the salaries that the agencies can
ply with the requirements. Other pay).
agencies can use this as an argu- • Staff of some agencies seems to lie
ment to request at least a minimum under a certain minimum of required
of staff for themselves. staff.
The number of staff varies tremendously across agencies. Sufficient staff is a key
factor for quality and velocity of the work. Sufficient resources are needed in the
MS to reach comparable staff numbers relative to their population sizes.
9.1.3 Expertise
Strengths and weaknesses of the European PhV System in terms of expertise can
be summarised as follows:
• The system encourages support • The capability for safety issue as-
from other MS and provides oppor- sessment does not exist in all agen-
tunity to learn from other agencies' cies.
experience. • According to the complex system
• Expertise is combined, a forum ex- and lack of experienced staff, some
ists for discussion of scientific and of the agencies would need more
practical issues; peer review is pro- support to be enabled to comply
vided. with the requirements.
• Expertise, assessments and other • Training within the system is partially
documents developed on EU level assessed by MS as insufficient and
can be used by the other agencies. expensive.
• The use of assessment reports,
SOPs and other documents is not
always optimal.
• For some agencies it is difficult to
hire well-educated staff because
there is too few in the country and
because they cannot pay competi-
tive salaries.
• For some agencies difficulties exist
to find external experts (e.g. phar-
macoepidemiologists).
Concerning the training in handling pharmacovigilance issues and the whole sys-
tem some MS refer to good offers of other institutions, e.g. the WHO-UMC.

The strengths and weaknesses of the European PhV System as far as technical
resources are concerned can be summarised as follows:
• The system allows the centralisation • Some agencies still have communi-
of database management and sig- cation problems which could in the
nal detection at the EMEA worst case lead to severe delays in
• Some of the MS agencies have al- the reaction on public health prob-
ready made large investments into lems.
their national databases and abili- • Despite EudraVigilance, MS still
ties to exchange data with need large investments into their
EudraVigilance. own database systems; some of the
• The technical resources are gener- MS agencies seem not to have ade-
ally assessed as good and suffi- quate resources for this.
cient with respect to the national • MS agencies develop own database
situation (e.g. having only a few solutions with little use of EMEA's,
ICSRs to process annually). other MSs' or third parties' experi-
ence, this reminds somehow to in-
venting the wheel a second time.
9.1.5 Co-operation
The strengths and weaknesses of the European PhV System regarding the co-
operation between the agencies can be summarised as follows:
• The share of work is good if MS • Lots of discussions are necessary to
comply with their roles (e.g. active represent all MSs' needs.
rapporteurs). • Being dependent on other agencies
• Information can be exchanged rap- e.g. as a concerned MS is a prob-
idly, agencies are generally notified lem as long as the agencies' work is
quickly of safety issues. of different quality.
• The system allows good access to • Different opinions exist what amount
information from other MS (esp. of work should be done at the na-
relevant for small MS). tional level, leading to different as-
• EMEA gives good backing for MSs' sessments of necessary and un-
decisions and arguments for their necessary duplication of work,
implementation. which is assessed by some of the
agencies as relatively high.
• Some agencies do a larger share of
work for the community than others.
• Communication between MS agen-
cies and EMEA is sometimes as-
sessed as difficult.
Some interview partners in the agencies criticised the communication of the MS

agencies with the und EMEA to some extent. Co-ordination is said to be missing if
the EMEA negotiates with a large company's headquarters and the MS with the
subsidiaries in their own country. In addition, as it was the case for Coxibes, EMEA
sometimes reacts too fast and then has to send updated information to the agen-
cies.
The completion of some of EMEA's projects (as E2E) is seen as unrealistic.
9.1.6 PhV embedded in larger Public Health strategy

The strengths and weaknesses of the European PhV System can be summarised
as follows:
• Some agencies have integrated • For some agencies the political sup-
pharmacovigilance into a broader port is weak, as pharmacovigilance
understanding of drug or even gen- is not perceived by the public as im-
eral consumer safety including e.g. portant issue of public health.
the protection against counterfeit • According to some interview part-
medicines. ners, the public and even HCPs do
frequently not understand that medi-
cines normally do have side effects
and instead of absolute safety the
balance of risk to benefits has to be
optimised.
9.1.7 Collaboration with stakeholders

as follows:
• There are different strategies of how • The agencies' influence on the pre-
agencies communicate with HCPs scription behaviour is weak.
that can be used as models of best • The MAHs' compliance e.g. with the
practice. submission of PSURs as well as the
• The new legislation offers stronger implementation of regulatory action
instruments to request information is often not checked and sometimes
or studies from MAHs and enforce seems suboptimal.
compliance by penalties. • The assessment of safety issues and
decision-making process in the
MAHs is sometimes unclear and
leads to unforeseeable results.
• Responsibilities of the agencies for
covering internationally active MAHs
with headquarters and subsidiaries
in different MS are unclear.
9.1.8 General quality management

as follows:
• • A systematic quality management is
not implemented in the most PhV
departments. The regulatory system
does not provide clear goals or pro-
visions in this respect.
• The agencies have not always met
their own internally set targets for
compliance with requirements.
• If implemented, nearly all agencies
state that their audit procedures do
not adequately ensure the quality of
their work.
• The impact of communications is
only followed-up on a routine basis
by four of the 29 agencies.
9.2 Data collection

9.2.1 Have sufficient data
as follows:
• The system combines the ICSRs • The safety issues can differ from
from a large population in order to country to country especially be-
increase statistical power with cause of varying consumption pat-
which signals can be detected; es- terns; these differences are often
pecially small countries with few re- not totally known because of a lack
ports benefit from this. of adequate and comparable data
• Most of the agencies where report- and therefore not always adequately
ing is mandatory for HCPs find this taken into account.
helpful to improve reporting; some • The reporting rates differ greatly be-
other agencies would welcome tween countries.
mandatory reporting in their coun- • The agencies are not very well pre-
try. pared for crises by routine data
(ICSRs and PSURs), their useful-
ness is restricted. Besides ICSR
and PSURs data especially on the

consumption of drugs, but also rele-
vant registries (vaccines, intoxica-
tion, drug misuse…), are perceived
as highly relevant, but not available
and not used sufficiently.
• PASSs and other data that can sup-
plement the routine data (ICSRs
and PSURs) have played a decisive
role in the last safety crises. How-
ever, only very few prospective
safety studies were prepared in the
last years, and some of them have
not been independent from the pro-
ducer of the drug under study. The
funding of necessary studies is often
not guaranteed.
• Except for outpatient care and intrau-
terine drug exposure, registries that
combine drug exposure and out-
comes data including ADRs exist in
most of the countries. However,
most agencies do only have access
to these data in exceptional cases,
and they are quite infrequently
used.
• Research into the safety of drugs for
children is disparately missing.
• A database on products on the mar-
ket is also missing.
• The collection and analysis of
PSURs is problematic: A small num-
ber of agencies have not even re-
ceived a single PSUR in 2004 which
is an indicator of non-compliance of
MAHs; compliance can often not be
checked; others get far more
PSURs than they can analyse.
• The necessities and requirements
regarding the collection and review
of ADRs and SUSARs from 3rd
countries are unclear and may lead
to unnecessary duplication.
• Even the collection and analysis of
PSURs for NAPs results in duplica-

tion of work, as many of these prod-
ucts are registered also in other MS.
• Too little information (ICSRs, studies
etc.) is available on herbal/homeo-
pathic products.
9.2.2 Quality of data

The strengths and weaknesses of the European PhV System regarding the quality
of the data collected can be summarised as follows:
• Strategies exist and are generally • The compliance of MAHs with expe-
applied to ensure the quality of dited reporting is routinely checked
ICSRs. in only 41% of the cases, the com-
pliance regarding PSURs in only
56%. This impairs the comprehen-
siveness and representativeness of
the data.
• PSURs do often not contain much
information, e.g. generics PSURS
do not include information on the
original product.
9.3 Data management

9.3.1 Soundness (Reliability)
Regarding the soundness of data management, the strengths and weaknesses of
the European PhV System can be summarised as follows:
• SOPs exist for data management • A lot of different IT solutions are
that are generally applied. used with a wide range of specifity
for the necessities of pharmacovigi-
lance.
9.3.2 Speed
The strengths and weaknesses of the European PhV System concerning the
speed of data handling can be summarised as follows:
• • Some duplication of work related to
the handling of the same data
(ICSRs, PSURs) exists at different

agencies, especially at the EMEA
on the one side and national agen-
cies on the other.
• In some agencies, the necessary IT
resources for the timely manage-
ment of the data are not available.
9.3.3 Workload
The workload related to data management has the following strengths and weak-
nesses:
• The system allows for a systematic • The issue of duplication of work
share of work between the involved (what is necessary, what is unnec-
stakeholders (MAHs vs. agencies essary duplication?) is assessed
as well as between different agen- heterogeneously by the agencies;
cies). some duplication seems to exist at
least with reports from 3rd countries
and with PSURs.
• Duplication of work results from two
international systems existing in par-
allel (i.e. EudraVigilance and the
WHO-UMC). Although these sys-
tems do partially have different
tasks and scopes, as well as differ-
ent regional coverage, this results in
a serious waste of resources.
9.4 Signal detection

9.4.1 Soundness
The strengths and weaknesses of the European PhV System with respect to signal
detection can be summarised as follows:
• EudraVigilance and the related pro- • The success of the combination of
cedures build the basis for the ef- expertise and resources for signal
fective systematic pooling of and detection depends on the full im-
signal detection out of spontaneous plementation of the provisions; with
reports. regard to other areas and depend-
ence on national resources and pri-
orities, this cannot be assumed as
guaranteed.

• As the last crises have showed,
other sources of information than
spontaneous reports are of out-
standing importance for signal de-
tection, which are at the moment
still underdeveloped.
• Some agencies assess their tools for
signal detection as insufficient, es-
pecially the tools for small numbers
of cases. The case studies showed
that improved-data mining tech-
niques and better European phar-
macovigilance/ pharmacoepidemi-
ology data platforms are needed for
Europe to stay at the cutting edge.
• As for data management, it does not
seem that the best use is made of
work that is mutually done by the
European system and the WHO-
UMC.
• As it can hardly be controlled, the
compliance of MAHs in their role to
do first-line signal detection is un-
clear.
9.5 Safety issue assessment

9.5.1 Co-operation
Regarding safety issue assessment, the strengths and weaknesses of the Euro-
pean PhV System can be summarised as follows:
• The system allows sharing work and • The concerned MS depend on the
using assessment reports from quality of assessment reports that
other countries. the rapporteurs of RMS agencies
• Opinions given by the CHMP are produce.
mostly assessed to have good • The contributions of the NCAs to the
quality. EEA PhV system in terms of as-
• Assessments are found by agree- sessments carried out are of high
ment and therefore few discussions variability. Some agencies admit
are necessary in the later stages. that they do not have the ability to
manage safety issues adequately
on their own.

• For some assessments, the agen-
cies find that the pharmacovigilance
expertise of the CHMP or the rap-
porteur's agency was not always
sufficient, and that external exper-
tise has not always been used ade-
quately.
Co-operation within the agency/unit as well as external division of labour (experts,

committees, other agencies) are especially important for safety issue assessment.
9.5.2 Access to external experts

The strengths and weaknesses of the European PhV System for the critical suc-
cess factor "Access to external experts" can be summarised as follows:
• The system encourages the use of • The quick and reliable access to ex-
external experts as for specific ternal experts is a key factor for the
safety issues no agency can keep speed and often the quality of the
all necessary expertise in-house. assessment. This is not assured for
all agencies. Not all MS agencies
have the access to external experts
when they would need it.
• In smaller MS it is unrealistic to find
experts for all possible safety issues
within the country.
9.6 Decision-making
9.6.1 Speed
The strengths and weaknesses of the European PhV System regarding decision-
making can be summarised as follows:
• • Decisions often need too much time
which is partially attributed to com-
plicated structures within the CHMP
and between CHMP and the Com-
mission, especially in the case of re-
ferrals.
9.7 Communication and action to protect public health

9.7.1 Speed of implementation
The strengths and weaknesses of the European PhV System concerning the
speed with which decisions are implemented into communication and action can be
summarised as follows:
• The system provides the structures • The time between the detection of a
to come to timely actions. signal and action (reporting/ publish-
ing of the decision) was too long in
some cases.
9.7.2 Harmonisation of communications

With regard to the Harmonisation of communications, the strengths and weak-
nesses of the European PhV System can be summarised as follows:
• The existing procedures communi- • In some cases MS agencies found
cation channels allow harmonised that their particular situation (e.g.
communications in the end in those regarding epidemiology) was not
cases where satisfactory agree- adequately represented in CHMP
ment has been obtained between opinions or Commission decisions.
the agencies and when sufficient Therefore the implementation of de-
time is available. cisions was sometimes difficult.
• Regulatory transparency is very im-
portant. In the SSRI case it was not
possible to uncover the exact con-
siderations leading to the EMEA
regulatory decisions (contrasting
with the US FDA).
• The agencies have only weak means
to influence the timing and content
of communications (e.g. changes for
SPCs) that the MAHs make.
• The information for patients e.g. in
patient information leaflets is not al-
ways harmonised, e.g. if information
on generic products is given on the
level of the product and not on the
level of the active ingredient con-
cerned.
9.7.3 Outcomes of regulatory action

The strengths and weaknesses of the European PhV System regarding the out-
comes of regulatory action can be summarised as follows:
• • The outcomes of regulatory action
are only assessed in exceptional
cases.
• There is very little information about
what prescribers do with label in-
formation and label changes. More-
over, when information is there, the
results are not very encouraging.
• The missing information on out-
comes is partially attributed to far
too few inspections of MAHs with a
pharmacovigilance focus.
Generally, the outcomes of regulatory action cannot easily be evaluated, because

even the agencies do normally not have such information. Actions are not evalu-
ated pro-actively, and even if changes in the morbidity and mortality caused by
ADRs were detected they could not causally be related to single regulatory acts.
10 Recommendations
According to the original project plan, recommendations for making the European
Community system of pharmacovigilance more robust were deducted based on
task 6 and discussed in the expert workshop. Draft recommendations were derived
basically from the literature review and the interviews and were discussed at the
expert workshop on June 15.
The recommendations are organised according to the main processes and respec-
tive success factors.
10.1 General factors

10.1.1 Legal framework conditions
• To make best use of the existing legal framework, all existing legal rules
should be implemented fully and this should be more strongly supervised.
• The impact of the new legislation especially as far as the improvement on
the side of prospective safety studies is concerned should be assessed
critically after a certain time; if necessary the respective instruments must
be sharpened further.
• It should be ensured that all stakeholders including the agencies are aware
of their obligations, e.g. that they have all necessary guidelines in place and
that they possess all necessary capabilities.
• To improve the clarity and simplicity of the guidelines in order to make it ea-
sier to find out the steps to do in a particular situation especially for agen-
cies with less specialized staff for regulatory affairs, a new version of a de-
cision-tree-shaped, probably HTML-based "super-guidance" (like Volume 9)
might help to easier navigate through decisions and help to comply with the
regulatory requirements.
10.1.2 Sufficient number of staff

• It should be ensured that the agencies have sufficient staff to guarantee the
compliance with the requirements concerning legal framework as well as
public health. Among these is that at least one qualified person should be
available 24h a day and all days of the year to react in safety crises.
• Taken the average of staff as a measure that is available in the agencies, a
value of 1.2 full-time-equivalents (FTEs) per Million inhabitants should be
attained for pharmacovigilance staff (scientific and administrative personnel
together) taken the national agencies and – if available – regional centres
that do a part of the legally required work together.
• To support this, PhV should be included into the university education which
would increase the pool of well-prepared potential staff members. Political
support has to be developed to increase financial resources to pay ade-
quate salaries in order to hold well-educated personnel in the agencies.
10.1.3 Expertise
• The staff has to be aware of its responsibilities. This has to be ensured by
sufficient training for new and older staff members on upcoming scientific
issues and new regulation.
• According to their responsibilities for the safe use of drugs in their home
countries, at least one senior pharmacovigilance expert should be available
in all agencies all the time including times of vacation etc.
• Expertise from other agencies should be used as far as possible. The adop-
tion of SOPs or other guidelines that were developed by other agencies
would make the other agencies' knowledge explicit and available for one's
own work. The same is true for other agencies' assessment reports which
are at the moment not systematically used.
• Central structures are necessary to supplement expertise that is missing in
one country by persons from other countries or from the EMEA.

• The severe communication problems that some agencies still have must
urgently be resolved in the concerned agencies.
• A standard should be defined including not only hardware to manage
ICSRs and to run Eudravigilance, but also all other communication technol-
ogy as (mobile) telephone systems with relay function, ensuring that e-mails
are read and answered in due time etc.
• With the development of database systems etc. the agencies should draw
as much as possible on pre-existing experience in other agencies and even
abroad.
10.1.5 Co-operation
• One senior pharmacovigilance staff in each agency should be reachable
24h a day. This would also improve the agency's co-operation with other
agencies, MAHs and other stakeholders in the case of a crisis.
• Within the agency, structures should ensure horizontal collaboration (e.g.
with the pre-marketing units and the inspections department/agency).
• More effective structures are needed for the agencies to collaborate with
other (national and EU) governmental bodies e.g. in decision-making.
• However, the definition of responsibilities and roles between the EU Mem-
ber States and EMEA (e.g. in Signal Detection) are not clear for all actors in
the agencies. This should be resolved by clear and simple guidelines.
• The division of labour should be as strong as possible. The diverging opin-
ions on what kind and amount of work is necessary at the national level
should be discussed and a solution should be found.
• A lot of time and money could be saved if competences were used that ex-
ist outside the EEA system. One approach might be "Centres for excel-
lence" for specific tasks (e.g. development of databases or drug classes).
All materials, working papers, draft communications, web-sites, SOPs,

blueprints of databases etc. should be made available and used.
• For the sake of homogeneity and fairness, all agencies should contribute to
the common tasks according to the size of their population. Less well in-
formed and equipped agencies should be enabled to catch up by practical
guidelines and direct support.
• It should be accepted that some tasks are as consuming for small as for
large countries and that agencies with fewer own spontaneous reports have
other information needs than those with sufficient national ICSRs.
10.1.6 PhV embedded in larger Public Health strategy

• One approach to educate the public, HCPs and policy-makers about the
tasks and necessities of pharmacovigilance would be to perceive PhV as
one part of a larger health & consumer safety strategy, e.g. by integrating
PhV in a system with other "vigilances".
• In public communications and education of HCPs not the absolute safety of
drugs, but risk/benefit should be emphasized.
10.1.7 Collaboration with stakeholders

• Agencies should improve their communication with HCPs. Good
experiences exist with regular contacts with professional associations etc.
• It is necessary to increase the influence on prescription behaviour. To this
aim it is necessary to influence clinical guidelines and Patient Information
Leaflets, not only SPCs, according to new evidence. Safety information
should be included into HCPs' day-to-day information (formulary…).
• Reporters should be educated and feedback for reporters optimised.
• The collaboration of agencies with MAHs in their decision-making should be
improved.
• The MAHs' compliance e.g. with the submission of PSURs as well as the
implementation of regulatory action should generally be better controlled.
The enforcement of compliance by penalties should be strengthened.
• The existing possibility to request additional surveillance studies from the
MAHs has been used only seldom in the past. This can be combined with a
conditional or otherwise restricted marketing authorisation for the drug in
question. It has to be ensured that necessary studies are carried out.
• The special interests of patients as regards pharmacovigilance must not be
neglected at any rate. Patients can contribute a unique perspective on
safety issues. Tendencies exist in some agencies to make better use of di-
rect and regular contacts to patients or patients' organisations, this should
be extended to all agencies. One approach is to allow patient reporting with
validation of the report by a HCP.
10.1.8 General quality management

• A general quality management system should be implemented including the
monitoring of agencies' compliance with requirements, regular assessment
of relevant indicators, internal audits, learning from practice (Continuous
quality improvement) and elimination of weaknesses.
• The outcomes of regulatory action in terms of prescription/use data should
be reviewed in important cases.
• The agencies should continue to mutually support themselves in compli-
ance; this will reduce their own workload in the long run.
10.2 Data collection

10.2.1 Have sufficient data
• To improve the access to ICSRs, well-tried multi-channel technologies exist
in some member states to improve spontaneous reporting should be ap-
plied in all countries. To introduce regional PhV centres in medium-size or
larger MS is one of the promising approaches. More education on pharma-
covigilance for HCPs could also increase the understanding of reporting
and thus improve the reporting rates. The introduction of a legal duty to re-
port should be considered.
• The access to necessary data should be facilitated. Besides ICSR and
PSURs data especially on the consumption of drugs, but also relevant reg-
istries (vaccines, intoxication, drug misuse…), are perceived as highly rele-
vant.
• A core set of data that complement spontaneous reporting and PSURs
should be defined, and where missing, the necessary structures should be
created. Priority should be given to develop further European pharmaco-
vigilance/pharmacoepidemiology data platforms.
• The access to premarketing information (including preclinical data as well
as the results of clinical trials) as well as to post-authorisation surveillance
studies (PASSs) and to relevant scientific literature has to be optimised.
• These data should be regularly used in safety issue assessment.
• To increase the efficiency of collection of ICSRs it should be distinguished
between new and other drugs for which particular attention is necessary on
the one hand, and "old" and well-known drugs similar to the different fre-
quencies for PSURs over the years after marketing authorisation. The UK
black triangle symbol seems to be helpful in directing the HCPs attention to
reporting of ADRs.
• The smaller countries rely heavily on international data/information to which
they should have best access. This will be realised for ICSRs via Eudravigi-
lance, but should be extended to other resources too.
• The pharmacovigilance planning tools should be used to systematically
generate studies in which ADRs are explicitly regarded as endpoints. A
scheme to identify priority areas (types of ADRs, classes of drugs) where
studies are of particular need has to be developed, and companies or public

sponsors urged to carry out these studies at an adequate level of quality.
This might be supported by an international institution.
• The lack of research into safety in children is tackled by the new Commis-
sion initiative. It should be checked in due time if it reaches its aims.
• Information on unpublished trials should regularly be included in PSURs.
Clinical trial registries and regulatory action should address this problem as
it has been recommended frequently in the scientific literature.
10.2.2 Quality of data

• The quality of ICSRs is of outstanding importance. The respective guide-
lines should be applied. Serious and unexpected paper reports/electronic
reports should be validated before adding them to database.
• At least serious/unexpected reports and fatal cases should be followed-up
to receive all available data on the case. This is also a measure to inform
the reporter that her/his report was well recognized.
• Education of reporters to support the quality of reports should be practiced
wherever possible.
• PSURs and all other data received should be routinely checked for timeli-
ness and quality; this should be supported by PhV inspections at MAHs.
10.3 Data management

10.3.1 Soundness (Reliability)
• With respect of systematic pooling of and signal detection out of spontane-
ous reports the successful implementation of the EudraVigilance network is
essential.
• To ensure reliable processes the respective SOPs should be applied. Vali-
dated IT solutions should be used.
• It is important to have an overview of available sources of information,
therefore all necessary data should be accessible under one user interface.
10.3.2 Speed
• Unnecessary duplication of work should be avoided.
• Good IT infrastructure incl. software is necessary to enable agencies to do
as many routine tasks as possible electronically.
10.3.3 Workload
• To avoid unnecessary duplication, it should be clarified what work is neces-
sary at the national level, e.g. in the analysis of PSURs.
• Future increases in requirements (e.g. data warehouse) should be kept in
mind when calculating the necessary personnel and technical resources.

10.4.1 Soundness
• It is necessary to adopt procedures for signal detection to the specific situa-
tion (in terms of risk, available data etc.). A common understanding of sig-
nal detection is needed and should be adhered to in order to improve the
agencies trust in results of the others' signal detection and safety issue as-
sessments.
• It should be recognised that spontaneous reports are in no way representa-
tive for the population; this means that even one single report might be a
signal, and information other than from spontaneous reporting has to be
used proactively. The sequential approach should be replaced by a cyclic
approach.
• With respect of systematic pooling of and signal detection out of spontane-
ous reports the successful implementation of the EudraVigilance network is
essential.
• Systematic development and exchange of methods to analyse routine and
supporting data (ICSRs, PSURs, consumption data, etc., and combinations
thereof) is necessary. Data-mining techniques and institutionalized follow-
up procedures may help to make better use of available research, hopefully
leading to better, and earlier, signal detection.
• Specific emphasis should laid on statistical tools for small numbers of re-
ports.
• It should be ensured that MAHs fulfil their obligation to adequately and
timely identify safety signals concerning their signals. Inspections are one
way to check for their ability to do so.

10.5.1 Co-operation
• The work load related to assessment should be distributed more equally
with respect to the size of the countries. The roles and responsibilities
should be refined to increase the use of existing assessments from other
agencies.
• The agencies should be given the resources to carry out assessments for
NAPs on their market by themselves, using all available sources of informa-
tion including EudraVigilance.
• No decision about a signal should rely on only a single person. Assessment
reports from other agencies should be used systematically.
10.5.2 Access to external experts

• Access to national external expert(s)/committee, in smaller countries also to
international experts, should be guaranteed for routine and in exceptional
cases.
10.6 Decision-making
10.6.1 Speed
• The immediate access to decision-makers (within agency or in a higher-
level authority) should be guaranteed. A reliable structure should ensure
this vertical collaboration and that decision-makers have all necessary info
incl. a suggestion for regulatory action.
• The speed of decision-making on the EU-level should be increased. Time-
consuming processes within the Commission after having received an opin-
ion should be identified.
• The cooperation of PhVWP and CHMP should be revised: More compe-
tences for the PhVWP as the primary expert group for pharmacovigilance
should be considered.
10.7 Communication and action to protect public health

10.7.1 Speed of implementation
• Communications should be prepared in time. The respective communica-
tion channels have to be kept prepared for potential crises including product
withdrawals.
• Already drafts of communications should be prepared and exchanged with
other agencies after an opinion has been submitted to the decision-makers.
This should include early communication with all necessary target groups.
The respective SOPs should be adhered to.
• To build professional communication strategies with targeted information for
the different groups of stakeholders, the agencies' press officers should be
involved, who should collaborate to make best use of their competences
and avoid duplication of work.
10.7.2 Harmonisation of communications

• The regulatory system should guarantee that the outcome of EU assess-
ments is implemented in a harmonized way on the national level.
• The coordination between agencies and MAHs should be improved, espe-
cially with regard of the time of publication on safety issues.
10.7.3 Outcomes of regulatory action

• Effective strategies for risk communication towards prescribers should be a
topic that should feature on any agenda of risk management strategies.
• It should be strived to influence clinical guidelines and patient information
leaflets, not only SPCs.
• Safety information should be included into HCPs' day-to-day information
(formulary…).
• The outcomes of regulatory action should be audited.

• When managing drug safety issues, considering the impact of action taken
on patients is of key importance. Abrupt discontinuation is often unwanted
and requires monitoring of patients after the announcement of the safety
alert. Effective regulatory management of post-event (e.g. safety restric-
tions) drug use is to be warranted.
• Major action should be accompanied by an evaluation of the impact of
safety warnings on clinical practice. Careful monitoring of drug utilization
will help regulators to better anticipate on developments relevant for drug
safety.
10.8 Core recommendations

From the present research, we derive the following most important conclusions to
make the European System of Pharmacovigilance more robust:
• The relative contribution of the different sources of safety information
(ICSRs, PSURs, registries, consumption data, safety studies etc.) and re-
spective resources that are devoted to these tasks should be reviewed. The
necessary statistical tools should be developed and specific requirements
of small countries should be kept in mind.
• The new legislation strengthens the potential impact of tackling safety is-
sues more pro-actively. This opportunity should be extensively used.
• The decision-making process should be reviewed; opportunities to stream-
line and fasten it should be identified.
• The impacts of communications and actions should be checked more sys-
tematically and from the lessons learned the impact on prescription behav-
iour should be improved.
• The marketing authorisation holders are primarily responsible for the safety
of their products. More resources are necessary to check if they comply
with their legal obligations, and at the same time it should be identified how
the requirements can be made as supportive as possible (e.g. as far as
PSURs are concerned).
• General principles of quality management and continuous quality improve-
ment should be introduced, among others:
(1) setting realistic and measurable targets for key interim impacts and for
final outcomes;
(2) regularly checking if these target values have been reached;
(3) use of internal audit and peer review;
(4) identifying and deleting weaknesses (bottlenecks in procedures, un-
der-performance or under-equipment of actors, waste of resources…).
Annex 1: Literature
Eudralex Volume 9, Pharmacovigilance Guideline, No. 1 (2004), p. 3
Eudralex Volume 9, Pharmacovigilance Guideline, No. 4 (2004), pp. 69-114
Eudralex Volume 9, Pharmacovigilance Guideline, No. 5 (2004), p. 69
Eudralex Volume 9, Pharmacovigilance Guideline, No. 6 (2004), pp. 87-114
Eudralex Volume 9, Pharmacovigilance Guideline, No. 7 (2004), pp. 139-

142
Title 21 of the Code of Federal Regulation 310.305 - Records and reports

concerning adverse drug experiences on marketed prescription drugs
for human use without approved new drug applications. Revised as
of April 1, 2004 (2004)
Title 21 of the Code of Federal Regulation 314.80 – Post-marketing report-

ing of adverse drug experiences (2004)
Title 21 of the Code of Federal Regulation: 310.303 - Continuation of long-

term studies, records, and reports on certain drugs for which new
drug applications have been approved (2004)
Eudralex Volume 9, Pharmacovigilance Guideline, No. 2 (2005), pp. 119-

121
Eudralex Volume 9, Pharmacovigilance Guideline, No. 3 (2005), pp. 66,

(91,103,116)
Abraham J. (2004): Pharmaceuticals, the state and the global harmonisa-

tion process, in: Australian health review, Vol. 28, No. 2, pp. 150-160
AERS (2005a): Adverse Event Reporting System. Internet document, URL:

http://www.fda.gov/cder/aers/default.htm. Accessed 11. Aug. 2005a
AERS (2005b): Post-marketing Study Commitments. Internet document,

URL: http://www.fda.gov/cder/pmc/default.htm. Accessed 12. Aug.
2005b
Arlett, P. R. H. P. (2001): Compliance in European pharmacovigilance: a

regulatory review, Pharmacoepidemiology and Drug Safety., pp. 301-
302
Arnold, B. D. C. (2004): Regulatory Aspects of Pharmacovigilance, in: Tal-

bot, J./Waller, P. Stephens' Detection of New Adverse Drug Reac-
tions (5th ed.)., pp. 375-451
Bahri, P.; Tsintis, P. (2005): Pharmacovigilance-related topics at the level of

the International Conference on Harmonisation (ICH), in: Pharma-
coepidemiology and Drug Safety, Vol. 14, No. 6, pp. 377-387
Bendall, C. (2004): Legal Aspects of Pharmacovigilance, in: Talbot,

J./Waller, P. (eds.), Stephens' Detection of New Adverse Drug Reac-
tions (5th ed.)., pp. 453-491
CIOMS (2005): Management of Safety Information from Clinical Trials, Re-

port of CIOMS Working Group VI, Geneva 2005.
EMEA (2001): Position Paper on Compliance with Pharmacovigilance

Regulatory Obligations, CPMP/PhVWP/1618/01.
EMEA (2004a): EMEA/CHMP Working Group with Patients Organisation,

Outcome of Discussions: Recommendations and Proposals for Ac-
tion, Doc.Ref:EMEA/CPMP/5819/04/Final.
EMEA (2004b): Handling by the CPMP for safety concern for Pre- and
Postauthorisation Applications Submitted in Accordance with the
Centralised Procedures, Doc.Ref.CPMP/4285/04/Final.
EMEA (2004c): Mandat, Objectives, and Rules of Procedure for the CHMP
Pharmacovigilance Working Party, Doc. Ref.
EMEA/CHMP/PhVWP/88786/04.
EMEA (2004d): Procedure for European Union Guidelines and related

documents within the pharmaceutical legislative framework,
Doc.Ref.EMEA/P/24143/2004.
EMEA (2005): Eudravigilance: What is pharmacovigilance? Internet docu-

ment, URL: http://www.eudravigilance.org/human/Q&A.asp#3. Ac-
cessed 12. July 2005
Eudralex (2005): An assessment of the European Community System of

Pharmacovigilance: Terms of Reference. Internet document, URL:
Eudralex
Fujiwara, Y.; Kobayashi, K. (2002): Oncology drug clinical development and

approval in Japan: the role of the pharmaceuticals and medical de-
vices evaluation center (PMDEC), in: Crit.Rev.Oncol.Hematol., Vol.
42, No. 2, pp. 145-155
Goetsch R. (2005): E2B(M) FDA Perspective. DIA June 15, 2004. Internet
document, URL:
http://www.fda.gov/cder/present/DIA2004/Goetsch.ppt. Accessed 10.

Aug. 2005
Heads of Medicines Agencies (2005a): Action Plan to Further Progress the

European Risk Management Strategy, Doc. Ref.
EMEA/115906/2005/Final.
Heads of Medicines Agencies (2005b): Progress on implementation of the

European risk management strategy, HMA(Human) 2nd Report on
ERMS 11/5/05. Internet document, URL: Heads of Medicines Agen-
cies
Health Canada (2003): Guidelines for reporting adverse reactions to mar-

keted drugs. Guidelines for the Canadian Pharmaceutical Industry.
Internet document, URL: http://www.hc-sc.gc.ca/dhp-
mps/medeff/report-declaration/guide/guide-ldir_indust_e.html. Ac-
cessed 10. Aug. 2005
Health Canada (2004a): Canadian Adverse Drug Reaction Monitoring Pro-

gram (CADRMP) Online Query and Data Extract. Internet document,
URL: http://www.hc-sc.gc.ca/dhp-
mps/medeff/databasdon/index_e.html. Accessed 14. Aug. 2005a
Health Canada (2004b): How adverse reaction information on health prod-

ucts is used. Internet document, URL: http://www.hc-sc.gc.ca/dhp-
mps/medeff/advers-react-neg/fs-if/ar-ei_info_e.html. Accessed 10.
Aug. 2005b
Health Canada (2005a): Adverse reaction information (What is the regula-

tory framework for ARs?). Internet document, URL: http://www.hc-
sc.gc.ca/dhp-mps/medeff/advers-react-neg/index_e.html#5. Ac-
cessed 14. Aug. 2005a
Health Canada (2005b): Adverse reaction reporting forms. Internet docu-

ment, URL: http://www.hc-sc.gc.ca/dhp-mps/medeff/report-
declaration/form/index_e.html. Accessed 14. Aug. 2005b
Health Canada (2005c): Canadian Adverse Drug Reaction Information Sys-

tem (CADRIS). Internet document, URL: http://www.hc-sc.gc.ca/dhp-
mps/medeff/advers-react-neg/fs-if/cadris-2_e.html. Accessed 10.
Aug. 2005c
Health Canada (2005d): Health Product and Food Branch. Internet docu-
ment, URL: http://www.hc-sc.gc.ca/ahc-asc/branch-dirgen/hpfb-
dgpsa/index_e.html. Accessed 14. Aug. 2005d
International Conference on Harmonisation of Technical Requirements for

Registration of Pharmaceuticals for Human Use (ICH) (2005): Wel-
come to the official web site for ICH. Internet document, URL:
http://www.ich.org/cache/compo/276-254-1.html. Accessed 27. June

2005
Kubota, K.; Koide, D. (2004): Comparison of data mining methodologies

using Japanese spontaneous reports, in: Pharmacoepidemiology and
Drug Safety, Vol. 13, No. 6
McEwen J. (2004): Risk Management from an Asian/Pacific Rim Regula-

tory Perspective, in: Drug Safety, Vol. 27, No. 8, pp. 491-497
MedWatch (2003): Reporting by Consumers. Internet document, URL:

http://www.fda.gov/medwatch/report/consumer/consumer.htm. Ac-
cessed 12. Aug. 2005
MedWatch (2005a): Mandatory Reporting by Drug/Biologic Manufacturers,

Distributors, and Packers. Internet document, URL:
http://www.fda.gov/medwatch/report/mfg.htm. Accessed 12. Aug.
2005a
MedWatch (2005b): Voluntary Reporting by Health Professionals. Internet

document, URL: http://www.fda.gov/medwatch/report/hcp.htm. Ac-
cessed 12. Aug. 2005b
MHPD (2005): Marketed Health Products Directorate. Internet document,

URL: http://www.hc-sc.gc.ca/ahc-asc/branch-dirgen/hpfb-
dgpsa/mhpd-dpsc/index_e.html. Accessed 12. Aug. 2005
Moseley, J. N. S. (2004): Risk Management. A European Regulatory Per-

spective, in: Drug Safety, Vol. 27, No. 8, pp. 499-508
Rosenberger, J.; Schaefer, M. (2003): Qualitative und kooperative Anfor-

derungen an ein Pharmakovigilanz-System eines pharmazeutischen
Unternehmens, in: Pharm.Ind., Vol. 65, No. 10, pp. 1005-1012
Sherman L.A. (2004): Looking through a window of the Food and Drug Ad-
ministration: FDA's advisory committee system, in: Preclinica - A
BioTechniques Publication, Vol. 2, pp. 1-4
Tsintis, P. L. M. E. (2004): CIOMS and ICH Initiatives in Pharmacovigilance

and Risk Management, in: Drug Safety, Vol. 27, No. 8, pp. 509-517
Waller, P. C. T. H. H. (2004): Managing Drug safety Issues with Marketed

Drugs, in: Talbot, J./Waller, P. (eds.), Stephens' Detection of New
Adverse Drug Reactions (5th ed.)., pp. 345-373
Annex 2: Questionnaire for Agency survey

19.07.2005
"Assessment of the European Community System of Pharmacovigilance"
Agency Survey 2005
Please complete and return this form by July 31, 2005
Your agency:
Please give details of the person we should contact with any question about this return
Name of contact person:
Department:
Telephon number:
E-mail address:
Important: The survey concerns the Pharmacovigilance unit of the agency including Human
OTC, generics, herbal and other drugs, but no non-drug products. If not requested otherwise, please refer only to
resources that you have in your own agency (and not in Regional Centres for Pharmacovigilance, e.g.), and
staff that is paid by your agency.
Please provide the requested information for the complete year 2004 (01 January to 31 December 2004).
Please enter the requested information only in the white fields of this form. yes
If adequate, please check boxes by an "x". x
If you wish to split the questionnaire into separate chapters please always include the cover sheet.
Please return the completed questionnaire by 31 July 2005 to [email protected].
It is easier for us to get the data electronically, but you can also print the PDF-document and return it by FAX (+49
721 6809 315).
Thank you very much for your support!
Some of the questions are market with "ERMS" and a number. This means that the figures have been asked for in
the ERMS-survey so that the Agencies in the New Member States will still have the data for 2004 in their
records. Please enter them in our questionnaire too! The Agencies in the Old Member States have participated
in the ERMS survey already in 2002 and would now be asked for newer (2004) data.
Further information: Dr. B. Buehrlen (+49 721 6809 182; [email protected])

Dr. T. Reiss (+49 721 6809 160; [email protected])
Abbreviations: CHMP Committee for Medicinal Products for Human Use

EMEA European Medicines Agency
EU European Union
EEA European Economic Area (i.e. EU-25 + Iceland, Liechtenstein, Norway)
ERMS WG Heads of Agencies European Risk Management Strategy Working Group
ICH International Conference on Harmonisation
ICSR Individual Case Safety Reports
MAH Marketing Authorisation Holder
MedDRA Medical Dictionary for Regulatory Affairs
NUIS Non Urgent Information System
PhV Pharmacovigilance
PhVWP Pharmacovigilance Working Party
PSUR Periodic Safety Update Report
RAS Rapid Alert System
SPC Summary of Product Characteristics
WHO World Health Organisation
Chapter A: Framework conditions
1) How much time (%) of the PhV unit is spent on pharmacovigilance work ERMS
where your Agency is: 7
a) Acting for the Community (as Rapporteur) %
b) Acting for the Community (as Reference Member State) %
c) Acting on a nationally licensed product %
d) Work that is not product-specific %
Su
m: 100 %
2) How many regulatory approvals for NMEs (New Molecular Entities) were
granted in your home country? 2003 2004
a) National approvals in your home country
b) Mutual recognition procedure
ba) Thereof: Mutual recognition procedure
with your country as Reference Member State
c) Centralized procedure with your country as rapporteur
3) How many physicians work in your home country? 2004

a) in general practice or as specialists outside hospitals? number
b) inside hospitals? number
2004
4) How many companies hold at least one marketing authorisation in your home
country? number
5) How many companies have a production plant in your home country? 2004
a) as parts of an international company with subsidiaries
in at least one other country? number
b) as domestic company without subsidiaries in other countries? number
6) How many products are authorised in your country? 2004

a) Number of nationally authorised products in your country number
b) Number of MR authorised products in your country number
c) Number of centrally authorised products in your country number
Chapter B: Ressources for PhV
7) What was the annual budget of your Agency in 2004

(total; not including budget for regional centres)? Mio €
8) How many staff are directly employed in your Agency (total)? FTE
Please give numbers of staff in real Full-time equivalents (FTE), e.g.: 1 Person with
a full-time contract working only for the task in question would equal 1 FTE, but if
she/he only works with 80% of her/his time fort he tasks in question this would
equal 0.8 FTE.
9) How many staff in your Agency are directly employed in pharmacovigilance?
(not including regional monitoring centres)
a) total ERMS
FTE 1
Thereof:
b) administrative ERMS
FTE 3
c) scientific FTE
Of scientific:
ca) Pharmaceutical ERMS
FTE 3
cb) Medical FTE
cc) Epidemiology FTE
cd) Other (please specify): FTE
cda) FTE
cdb) FTE
cdc) FTE
cde) FTE
10) How many staff are directly employed in each process? ERMS
2
(Some may be involved in more than one process, then please count in each of the
categories.)
a) Data collection and data entry Persons
b) Data management Persons
c) Risk assessment Persons
d) Regulatory action (relating to pharmacovigilance issues) Persons
e) Risk communication Persons
f) Audit and quality assurance Persons
g) Monitoring compliance with industry on reporting requirements Persons
11) Do you contract out any of your pharmacovigilance assessment work? (i.e. for ERMS
PSUR and safety related variations assessment) yes no 6
a) ...to external academics
b) ...to health professionals in health service
c) ...to regulatory consultants
d) ...to regional centres for PhV
e) ...to others (please specify)
yes no
12) Are regional centres in operation?
a) Number of regional centres centres
b) Please describe their responsibility shortly:
Routine work: yes no
ba) data collection and management
bb) signal detection
bc) safety issue assessment
bd) decision making
be) communication
bf) inspection of MAHs
Specific tasks: yes no
bg) scientific studies on PhV issues
bh) informal advice for the national agency
bi) other tasks (please specify):
c) Total number of staff (sum of all regional centres)

for routine work persons
d) other resources:
e) number of ADR reports collected by regional centres in 2004 reports
13) Describe the external expertise available in your country distinguishing the
following areas of expertise:
in
regi not at
onal all
available in cent external in the
(national) Agency re experts country
Experimental toxicology
Animal studies
In vitro testing
(Clinical) pharmacology
Medicine
Pharmacoepidemiology/ Drug utilisation
Epidemiology
Statistics
Human ADRs to veterinary medicinal products (only in
the case of veterinary medicinal products)
Design of pharmacovigilance plans
Regulatory affairs
yes no
14) Do you have an expert committee dedicated to pharmacovigilance? ERMS
11
a) If yes, how many times did it meet in 2004? times
yes no
15) This committee is not only responsible for PhV, but for marketing
authorisation and variations (e.g.) as well (e.g. Marketing Authorisation ERMS
Board) 11
16) What is the IT-system of the postmarketing unit like?

Number of PCs PCs
yes no
Sufficient number of PCs available for all scientific and technical staff
Local area network for the PCs available
Permanent Internet-access available
Sufficient support for maintenance of the IT systems available
totally very
sufficien insuffi-
t cient
17) How would you assess your IT-ressources (hardware, software, electronic
communication)?
Chapter C: Definitions and standards
yes no
18) Is MedDRA implemented as dictionary for coding of reports in your database?
19) Existing guidance documents

exists in national exists from EU
version
yes no yes no
Obligations of MAH
Obligations of NCA
Collaboration with other authorities and int. health
institutions
Quality management within the agency
Qualification of MAH
Standard Operation Procedure (SOP) for:
a) Data collection
b) Data management
c) Signal detection/
Safety issue assessment
d) Decision-making
e) Communication with MAHs
f) Communication with
Health care professionals
g) Crisis management
h) Feedback to reporters
i) Development and maintenance
of SOPs
20) Implementation of new requirements from Oct. 2005 on… yes no

a) … already in place?
b) … scheduled to be operable when? Month/
Year
Chapter D: Processes
Chapter D1: Data collection
21) How do reporters submit reports on ADRs? ERMS

14
a) Paper % of
ICSRs
b) Electronic % of
ICSRs
c) via web-site of the National Agency or of a Regional centre % of
ICSRs
d) Other, e.g. telephone % of
ICSRs
22) How many national ICSRs in total are ERMS

15
a) contained totally in your database of ADRs reports
aa) thereof received in 2003 reports
ab) thereof received in 2004 reports
23) How many reports of suspected ADRs did you receive in 2004 through your
national spontaneous reporting scheme?
a) Total reports
thereof:
b) received from MAHs ERMS
reports 13
c) direct from doctors/dentists reports
d) direct from pharmacists reports
e) direct from nurses reports
f) direct from patients reports
g) direct from coroners reports
h) direct from health professional body reports
i) Other (please specify):
reports
24) How many reports of suspected ADRs did you receive in 2004 through your
national spontaneous reporting scheme?
a) serious expected reports
b) serious unexpected reports
c) nonserious expected reports
d) nonserious unexpected reports
25) How many ICSRs did you receive in 2004 …

a) on nationally authorised medicines reports
b) from your country as concerned MS reports
c) from your country as reference MS reports
26) How many national ICSRs did you receive in 2004 on children? reports
27) How many PSURs did you receive in 2004 … ERMS

20
a) National PSURs Number
b) Mutual recognition when your counttry is reference MS Number
c) Centralised when your country is rapporteur Number
28) How many PSURs have you assessed in 2004? Number
29) What is the number of ICSRs from your country with incomplete data (i.e. less
than 4 minimal data points)? Number
very very
good bad
30) How well has your routine data-collection prepared you for the last
pharmacovigilance crisis?
only
margina
very l-ly
useful useful
31) How useful are routine data from your country (ICSRs and PSURs together)
for safety issue assessment compared to other information?
Chapter D2: Data management
yes no
32) Do you use an electronic database to manage national ICSRs?
33) Which database software do you use to manage national ICSRs?
yes no
34) Is EudraVigilance already implemented?
a) If not, when will it be fully operable from your Agency's side? Month/
Year
yes no
35) Have you implemented the standards required for the electronic transmission ERMS
of ICSRs? 4
yes no
36) Is electronic reporting by MAH due October 2005 in place?
if not: When will it be operable? Month/
Year
37) What is the share of reports from MAHs transmitted electronically of total
reports from MAH (average of 1st half of 2005) %
yes no
38) Is reporting of suspected ADRs by healthcare professionals mandatory? ERMS
13C
a) If so, do you apply the law to all Health care professionals?
b) Please specify (mandatory for…; exceptions….;
enforcement or not etc.)
39) How long does is take to assess PSURS (days from reception to finished
assessment) on average? days
adequat far too

e slow
40) How do you assess the time between data entry and transmission to EMEA or
MAHs?
Chapter D3: Signal detection
41) Are there other data that you are using or could use for signal detection or
similar ERMS 18-
safety issue assessment in your country?
19
(sources do not have to cover the whole population)
Exist in country Agency has access to Used
yes, but
in
a) Population-based health/disease only in
yes, no, routinel excep-
registries yes no excep- never
always never y tional
(=exposure-outcome databases): tional
cases
cases
aa) Inpatient medical care
ab) Outpatient medical care
ac) Cancer
ad) Causes of death
ae) Intra uterine drug exposure
af) Malformations in newborns

yes, but
in
b) Data on only in
yes, no, routinel excep-
consumption of yes no excep- never
always never y tional
medicines? tional
cases
cases
a) Sales data
b) Prescription data
ba) - non-hospital
bb) - hospital
bc) only for reimbursed medicines
bd) prescription data by age
be) ...by sex
bf) ...by geographic region

yes, but
in
only in
c) Other population-based data sources (please yes, no, routinel excep-
yes no excep- never
specify): always never y tional
tional
cases
cases
yes no
42) Do you have plans to obtain some of the data sources at which you do not
have access at the moment?
yes no
43) Do you have the capability to link prescription registries with other registries ERMS
which include health outcomes? 19C
yes no
44) Do you have experience in conducting pharmacoepidemiological studies ERMS
using such data? 18B
yes no
45) Do you evaluate reporting rates (calculated from spontaneous ADRs and ERMS
usage data)? 19A
46) How many pharmacoepidemiology studies, post-authorisation surveillance

studies or phase IV trials have been carried out last year with a sample from
your country, taken all sponsors together (public and private)? 2004
a) total number in 2004 Number
b) thereof initiated by the Agency Number
yes no
47) Has your Agency initiated or carried out ad hoc pharmacoepidemiology ERMS
studies in 2004 when a signal needed confirmation or quantification? 23
a) Using in-house expertise
b) Via collaboration with an academic department
c) Via the marketing authorisation holder
yes no
48) Has your Agency initiated or carried out pharmacoepidemiological studies for ERMS
early post-marketing surveillance of new products in 2004? 24
a) Using in-house expertise
b) Via collaboration with an academic department
c) Via the marketing authorisation holder
yes no
49) Are clinical trial adverse event (AE) reports collected by the authority and
available to those staff responsible for pharmacovigilance of marketed
products?
yes no
50) Is information collected on ADRs with compassionate use / named patient
use of products?
51) What published medical and scientific literature (including databases of

literature) are searched / screened and how often?
screened x times
Journal/Database/Source
per year
yes no
52) Are data / information on post-authorisation safety studies routinely collected
and recorded?
yes no
53) Are data/information on phase IV efficacy trials routinely collected and
recorded?
yes no
54) Are data/information on preclinical studies routinely collected and recorded?
yes no
55) For information from other regulatory authorities, are data / information
routinely collected and recorded?
yes no
56) Do you have all of the following data directly accessible under one user
interface: national ICSRs, national PSURs, reports from literature,
prescription or consumption data, and premarketing safety data?
often
always very
adequat inadequ
e ate
57) How do you assess the statistical tools that you have available for signal
detection?
yes no
58) Do you have adequate statistical tools for small numbers of cases that you
can run on your national data?
Chapter D4: Safety issue assessment
yes no
59) Are external experts (besides the Pharmacovigilance Working Party) routinely
involved in the assessment of safety issues?
(nearly)
very impos-
easy sible
60) How easy is it for you to receive support from external experts in routine
work?
always
when (nearly)
necessa impos-
ry sible
61) How easy is it for you to receive support from external experts in exceptional
cases?
yes no
62) Do you have the capabilities in your country to identify and assess signals
without help from other agencies?
very very
little much
63) How much work is done in signal detection and safety issue assessment
within your country and at the same time in other MS or on EU level?
2004
64) How many assessment reports were written by your Agency in 2004? Number
very very
good bad
65) How do the MAHs in your country comply with their obligation to analyse
safety signals?
Chapter D5: Decision-making
66) Adequate decisions are found for safety issues… always seldom
a) for Nationally authorised drugs
a) for Mutual Recognition authorised drugs
a) for Centrally authorised drugs
67) Decisions are found for safety issues in adequate time… always seldom
a) for Nationally authorised drugs
a) for Mutual Recognition authorised drugs
a) for Centrally authorised drugs
very very
good bad
68) How transparent to your Agency is the process of decision-making on safety
issues in the companies located in your country?
Chapter D6: Communication/Action
69) Do you routinely inform the following stakeholder groups on general issues
of drug safety? yes no
a) Individual doctors or doctors in hospitals
b) Medical associations
c) Professional journals
d) Pharmacists or pharmacists' associations
e) Other HCPs
f) Patient organisations
g) MAHs
i) The public/media
j) Other groups
70) Whom do you inform on specific safety issues? yes no

a) Individual doctors or doctors in hospitals
b) Medical associations
c) Professional journals
d) Pharmacists or pharmacists' associations
e) Other HCPs
f) Patient organisations
g) MAHs
i) The public/media
j) Other groups
yes no
71) Do you always have the best measures to minimize risks from ADRs?
2004
72) How many responses were given to enquiries by HCPs? Number
73) On how occasions were Dear-doctor-letters sent to HCPs in your country? Number
74) How many letters were sent to MAHs to amend SPCs? Number
75) How many variations of SPCs were evaluated? Number
76) How many inspections of MAHs were carried out where PhV was an issue (at
least partially; including inspections that were carried out by other authorities
in the coutry)? Number
77) How many drugs were withdrawn from your national market? Number
78) How many marketing authorisations were suspended for drugs on your
national market? Number
yes no
79) The organisation has the capability of leading EU wide co-ordination of
regulatory action and communication of drug safety issues.
adequat far too

e slow
80) How do you assess the time between the detection of a signal (first
discussion within the agency) and reporting (publishing) of decision with
respect to this safety issue?
very very
good bad
81) How consistent is the communication on safety issues across agencies?
82) How consistent is the communication on safety issues between agencies on

the one side and MAHs and HCPs on the other side?
Chapter E: Outcomes
yes no
83) Do you routinely follow-up the impact of communications?
very very
good weak
84) How strong is the influence of the Agency's communications on the doctors'
prescription behaviour?
85) What are the outcomes of safety-relevant studies using samples from your
country (if known)?
Please indicate relevant studies (also from the literature) that were carried out in
your country in the last 3 years.
a) Outcome: Incidence of ADR-relevant diseases
carried Outco Unit:
out in me (e.g. per million
Reference year (Rates) inhabitants)
aa) Study 1
ab) Study 2
ac) Study 3
ad) Study 4
ae) Study 5
af) Study 6
please
mark
ag) There were no such studies in our country in the last
3 years
b) Outcome: Mortality due to ADRs

carried Outco Unit:
ba) Study 1
bb) Study 2
bc) Study 3
bd) Study 4
be) Study 5
bf) Study 6
please
mark
bg) There were no such studies in our country in the last
3 years
c) Outcome: Hospitalisations due to ADRs

carried Outco Unit:
ca) Study 1
cb) Study 2
cc) Study 3
cd) Study 4
ce) Study 5
cf) Study 6
please
mark
cg) There were no such studies in our country in the last
3 years
d) Outcome: Quality-adjusted life years (QUALYs) lost due to ADRs?

carried Outco
out in me Unit: QUALYs
Reference year (Rates) (e.g. per patient)
da) Study 1
db) Study 2
dc) Study 3
dd) Study 4
de) Study 5
df) Study 6
please
mark
dg) There were no such studies in our country in the last
3 years
Chapter F: General aspects
2004
86) What percentage of the staff in the PhV unit has received a training in the last
year? %
87) How many training measures (internal or external) took place with at least one
participant from the agency? Number
2004
88) How many events have taken place in the last year with participation or
support from the Agency to educate reporters/HCPs in pharmacovigilance? Number
89) How many bulletins from your agency including safety issues were issued? Number
2004
90) How many answers to the CHMP were prepared by your Agency? Number
91) How many legal documents and guidelines were prepared by your Agency? Number
92) How many scientific publications with at least one author from the agency
were published in the last year? Number
very very
good bad
93) How does the Agency meet its internal targets for timing and other
requirements?
very very
good bad
94) How do you assess the internal cooperation within the agency (within PhV
unit, with pre-marketing departmernt, incl. IT staff)?
very very
good bad
95) How do you assess the cooperation of your agency with HCPs?
very very
good bad
96) How do you assess the cooperation of your agency with the MAHs in your
country?
very very
good bad
97) How do you assess the cooperation between the national agencies and the
EMEA?
very very
good weak
98) How strong is the political support for pharmacovigilance in your country in
general?
very very
good bad
99) How do you assess the overall compliance of the the MAHs in your country
with the legal requirements?
Annex 3: Results of Delphi survey on critical success

factors
19.07.2005
Analysis: Delphi sheet for critical success factors Round-1:
Values relative to number of respective answers
Relevance: How important is the factor for the performance

of the European System for Pharmacovigilance (or parts of it)?
Evaluation Round 1: Relevance for…23
… quality of the work24 … compliance with … speed ("kinetics") … work load/costs

requirements
Success factor ++ + 0 - -- ++ + 0 - -- ++ + 0 - -- ++ + 0 - --
1. … for Data collection
1.1 Comprehensiveness of the data 26% 46% 15% 0% 0% 16% 33% 39% 0% 4% 15% 23% 42% 8% 8% 8% 31% 23% 31% 4%
Mandatory reporting by HCPs 25% 33% 17% 8% 17% 23% 38% 31% 0% 8% 8% 23% 54% 8% 8% 8% 31% 23% 31% 8%
Spontaneous reports from 27% 64% 9% 0% 0% 17% 42% 42% 0% 0% 0% 33% 67% 0% 0% 0% 36% 55% 9% 0%
pharmacists
Access to FDA data for national 17% 42% 33% 8% 0% 15% 15% 62% 0% 8% 15% 15% 54% 8% 8% 8% 38% 38% 8% 8%
agencies
Access to drug utilisation statistics 79% 21% 0% 0% 0% 31% 46% 23% 0% 0% 23% 38% 31% 0% 8% 8% 31% 46% 15% 0%
Access to database of patients' 38% 62% 0% 0% 0% 15% 23% 62% 0% 0% 15% 38% 31% 8% 8% 31% 23% 15% 31% 0%
medical records
Highest-possible number of 27% 36% 18% 18% 0% 27% 27% 36% 0% 9% 18% 9% 45% 9% 18% 36% 27% 18% 9% 9%
spontaneous reports
More information from pre- 23% 62% 15% 0% 0% 0% 58% 33% 0% 8% 15% 23% 38% 23% 0% 8% 31% 23% 38% 0%
marketing evaluation
More post-authorisation 82% 18% 0% 0% 0% 23% 62% 15% 0% 0% 15% 31% 31% 23% 0% 17% 25% 8% 42% 8%
surveillance studies
23 Numbers represent the share of answers in this field of all answers given to this item in the respective dimension.
24 ++: strong positive influence; 0: not relevant; --: strong negative influence
Patient reporting 8% 54% 15% 23% 0% 0% 23% 69% 0% 8% 0% 23% 38% 31% 8% 8% 23% 23% 38% 8%
More direct input from patients or 17% 50% 33% 0% 0% 0% 23% 77% 0% 0% 8% 0% 46% 38% 8% 0% 31% 31% 38% 0%
patients' organisations
1.2 Organisation of data collection 35% 36% 15% 8% 0% 15% 36% 38% 0% 0% 8% 31% 44% 8% 0% 12% 19% 35% 23% 4%
Adaptive/stepwise approach 58% 33% 8% 0% 0% 17% 50% 33% 0% 0% 8% 42% 42% 8% 0% 8% 42% 25% 25% 0%
tailored to specific safety issue
Concentrate on best-quality data 54% 23% 15% 8% 0% 15% 31% 38% 15% 0% 8% 38% 46% 8% 0% 8% 38% 46% 8% 0%
Pro-active data collection 31% 69% 0% 0% 0% 8% 77% 15% 0% 0% 31% 23% 38% 8% 0% 15% 23% 8% 46% 8%
Have national database for ICSRs 31% 31% 15% 23% 0% 38% 23% 38% 0% 0% 15% 31% 38% 15% 0% 23% 8% 38% 23% 8%
in addition to EudraVigilance
Support in data collection by 23% 38% 31% 8% 0% 15% 15% 54% 15% 0% 8% 15% 46% 31% 0% 23% 15% 31% 23% 8%
regional centres for PhV
Offer single contact point for MAH 38% 31% 23% 8% 0% 0% 46% 54% 0% 0% 23% 31% 46% 0% 0% 8% 31% 54% 8% 0%
in agency for pre- and
postmarketing
Trust in HCPs as data source 8% 58% 17% 17% 0% 0% 42% 42% 17% 0% 0% 17% 75% 8% 0% 0% 8% 92% 0% 0%
Education of reporters 54% 46% 0% 0% 0% 46% 23% 31% 0% 0% 8% 46% 38% 8% 0% 15% 15% 31% 31% 8%
2. … for Data management
2.1 Electronic processing of data 54% 31% 8% 0% 0% 54% 38% 15% 0% 0% 69% 23% 8% 0% 0% 46% 38% 8% 8% 0%
Sufficient IT-ressources 62% 31% 8% 0% 0% 54% 46% 0% 0% 0% 69% 23% 8% 0% 0% 46% 46% 8% 0% 0%
(investments)
Sufficient support for maintenance 54% 38% 8% 0% 0% 54% 23% 23% 0% 0% 54% 38% 8% 0% 0% 54% 31% 8% 8% 0%
of IT systems
Have all information available 54% 23% 23% 0% 0% 46% 38% 15% 0% 0% 69% 15% 8% 0% 8% 46% 38% 8% 8% 0%
electronically
2.2 Processing of data 25% 42% 8% 0% 0% 15% 38% 38% 0% 0% 17% 58% 17% 0% 0% 17% 54% 31% 8% 0%
Internal cooperation within the 50% 42% 8% 0% 0% 8% 67% 25% 0% 0% 17% 83% 0% 0% 0% 18% 55% 27% 0% 0%
agency
Structure of the agency 23% 62% 15% 0% 0% 15% 23% 62% 0% 0% 15% 54% 31% 0% 0% 8% 54% 31% 8% 0%
Prioritization (among PSURs…) 25% 75% 0% 0% 0% 8% 25% 58% 8% 0% 17% 58% 25% 0% 0% 8% 58% 25% 8% 0%
Amount and quality of data in 58% 42% 0% 0% 0% 33% 42% 25% 0% 0% 25% 58% 17% 0% 0% 17% 33% 42% 0% 8%
national database
Processing of data as fast as 17% 25% 42% 8% 8% 23% 38% 38% 0% 0% 31% 62% 0% 8% 0% 25% 17% 42% 17% 0%
possible
3. … for Signal detection
3.1 Availability of necessary 31% 62% 8% 0% 0% 23% 38% 38% 0% 0% 31% 62% 8% 0% 0% 15% 54% 23% 0% 8%
information
Have all data directly accessible 31% 62% 8% 0% 0% 23% 38% 38% 0% 0% 31% 62% 8% 0% 0% 15% 54% 23% 0% 8%
under one user interface
3.2 Data analysis 54% 46% 0% 0% 0% 25% 25% 50% 0% 0% 38% 50% 8% 4% 0% 41% 36% 18% 5% 0%
New statistical methods to analyse 58% 42% 0% 0% 0% 25% 25% 50% 0% 0% 42% 42% 8% 8% 0% 45% 36% 18% 0% 0%
ICSRs
Adequate statistical tools also for 50% 50% 0% 0% 0% 25% 25% 50% 0% 0% 33% 58% 8% 0% 0% 36% 36% 18% 9% 0%
small numbers of cases
3.3 International share of work 60% 40% 0% 0% 0% 40% 20% 40% 0% 0% 60% 40% 0% 0% 0% 33% 50% 0% 0% 17%
4. … for Safety issue assessment
4.1 Share of responsibilities 38% 42% 12% 4% 0% 19% 42% 38% 0% 0% 19% 42% 27% 12% 0% 15% 54% 0% 23% 8%
Have national capabilities to 38% 31% 23% 8% 0% 23% 38% 38% 0% 0% 23% 38% 31% 8% 0% 31% 23% 15% 23% 8%
identify and assess signals without
help from other agencies
International cooperation/share of 77% 8% 15% 0% 0% 31% 8% 62% 0% 0% 38% 46% 15% 0% 0% 31% 46% 0% 15% 8%
work
Supplement MAHs' primary 31% 54% 8% 8% 0% 15% 46% 38% 0% 0% 15% 46% 23% 15% 0% 0% 62% 0% 31% 8%
obligation to analyse signals
Independence of the assessment 38% 54% 8% 0% 0% 8% 69% 23% 0% 0% 0% 31% 54% 15% 0% 0% 62% 0% 23% 15%
from the MAH
4.2 Expertise 58% 33% 8% 0% 0% 25% 31% 50% 0% 0% 25% 42% 25% 8% 0% 25% 8% 50% 17% 0%
Have expertise for assessment of 62% 31% 8% 0% 0% 31% 54% 15% 0% 0% 46% 46% 8% 0% 0% 25% 33% 25% 17% 0%
signals in-house
External review of assessments 23% 46% 31% 0% 0% 0% 31% 62% 8% 0% 8% 15% 46% 31% 0% 8% 8% 54% 31% 0%
Availability of external experts 58% 33% 8% 0% 0% 25% 25% 50% 0% 0% 25% 42% 25% 8% 0% 25% 8% 50% 17% 0%
within the country
4.3 Structures 0% 9% 64% 18% 9% 0% 9% 55% 27% 9% 0% 0% 55% 36% 9% 10% 20% 40% 20% 10%
Different requirements for NAPs, 0% 9% 64% 18% 9% 0% 9% 55% 27% 9% 0% 0% 55% 36% 9% 10% 20% 40% 20% 10%
MRPs, CAPs
5. … for Decision-making
5.1 Decision-making in legal bodies 23% 18% 25% 8% 0% 15% 18% 42% 8% 0% 8% 18% 64% 17% 0% 0% 25% 55% 9% 9%
Do only make decisions if the 9% 9% 27% 9% 45% 0% 18% 27% 18% 36% 0% 9% 73% 9% 9% 9% 18% 55% 9% 9%
responsible MAH does not
Decision-making as joint effort 23% 38% 23% 8% 8% 15% 23% 46% 8% 8% 8% 38% 38% 15% 0% 0% 42% 50% 8% 0%
between MAH and agency
Take into account costs of 18% 18% 64% 0% 0% 0% 9% 91% 0% 0% 0% 0% 64% 27% 9% 0% 9% 45% 36% 9%
decisions/actions (e.g. on 3rd
countries)
Lean decision-making with only 50% 17% 25% 8% 0% 33% 17% 42% 8% 0% 67% 25% 0% 8% 0% 9% 27% 45% 9% 9%
few steps or involved committees
Less influence of pre-marketing 0% 0% 36% 64% 0% 0% 0% 55% 45% 0% 0% 18% 64% 18% 0% 0% 33% 67% 0% 0%
units in decision-making
Find actions specific for safety 55% 36% 9% 0% 0% 36% 45% 18% 0% 0% 27% 18% 27% 27% 0% 18% 0% 64% 0% 18%
issue
Follow-up impact of decisions 50% 50% 0% 0% 0% 17% 58% 25% 0% 0% 8% 0% 75% 17% 0% 0% 25% 58% 17% 0%
5.1 Decision-making in companies 17% 58% 25% 0% 0% 17% 42% 42% 0% 0% 0% 17% 83% 0% 0% 0% 17% 75% 0% 8%
Transparent decision-making 17% 58% 25% 0% 0% 17% 42% 42% 0% 0% 0% 17% 83% 0% 0% 0% 17% 75% 0% 8%
within MAH
6. … for Communication/Action
6.1 Early communication 39% 36% 4% 13% 8% 16% 36% 32% 8% 8% 20% 28% 52% 0% 0% 0% 20% 40% 36% 4%
Communicate already before 8% 42% 8% 25% 17% 8% 25% 33% 17% 17% 17% 25% 58% 0% 0% 0% 17% 42% 33% 8%
decision is taken
Communicate not only reactively 69% 31% 0% 0% 0% 23% 46% 31% 0% 0% 23% 31% 46% 0% 0% 0% 23% 38% 38% 0%
6.2 Communication to all 54% 44% 0% 0% 0% 48% 31% 23% 0% 0% 8% 19% 46% 27% 0% 0% 23% 38% 27% 8%
stakeholders
Communicate to patients (patients' 38% 54% 8% 0% 0% 15% 38% 46% 0% 0% 15% 23% 31% 31% 0% 0% 31% 15% 46% 8%
organisations) directly
Contact professional journals 23% 46% 31% 0% 0% 0% 23% 77% 0% 0% 8% 15% 46% 31% 0% 0% 23% 23% 46% 8%
before they publish on safety
issues
Consistency of communications 75% 25% 0% 0% 0% 54% 31% 15% 0% 0% 8% 31% 46% 8% 8% 0% 23% 46% 23% 8%
across stakeholders and countries
Stronger harmonisation of 50% 50% 0% 0% 0% 46% 31% 23% 0% 0% 0% 15% 46% 38% 0% 0% 31% 31% 31% 8%
implementation of decisions
Communicate on risk-benefit- 77% 23% 0% 0% 0% 54% 23% 23% 0% 0% 0% 8% 62% 23% 8% 0% 15% 69% 8% 8%
ratios, not only on risk
Specific crisis communication 58% 42% 0% 0% 0% 50% 33% 17% 0% 0% 17% 58% 17% 8% 0% 0% 8% 75% 17% 0%
6.3 Impact of 50% 44% 8% 0% 0% 38% 35% 38% 0% 0% 8% 23% 54% 8% 8% 4% 35% 35% 23% 12%
communications/actions
Stronger influence on HCPs' 23% 69% 8% 0% 0% 15% 46% 38% 0% 0% 8% 23% 62% 8% 0% 0% 31% 38% 15% 15%
behaviour
Influence clinical guidelines and 15% 69% 8% 8% 0% 15% 46% 38% 0% 0% 0% 23% 54% 15% 8% 0% 23% 38% 23% 15%
PILs, not only SPCs
Supervise/control communication 15% 54% 31% 0% 0% 15% 38% 46% 0% 0% 0% 15% 54% 23% 8% 8% 8% 46% 23% 15%
of MAHs
Stronger control of MAHs' 46% 31% 23% 0% 0% 46% 23% 31% 0% 0% 8% 0% 69% 8% 15% 8% 15% 31% 23% 23%
compliance
Have the right tools to minimize 58% 42% 0% 0% 0% 42% 42% 17% 0% 0% 17% 58% 17% 8% 0% 8% 42% 25% 25% 0%
risk
Guidance for good communication 54% 38% 8% 0% 0% 38% 31% 31% 0% 0% 8% 62% 23% 0% 8% 8% 46% 31% 15% 0%
practice
Advice from communications 54% 46% 0% 0% 0% 38% 8% 54% 0% 0% 17% 33% 42% 8% 0% 0% 42% 50% 0% 8%
experts
Follow-up of the impact of 62% 38% 0% 0% 0% 38% 23% 38% 0% 0% 8% 0% 77% 8% 8% 0% 38% 23% 31% 8%
communications/actions
7. … for performance in general
7.1 Legal framework 42% 36% 23% 0% 0% 25% 46% 30% 0% 0% 0% 25% 54% 8% 0% 8% 23% 54% 8% 0%
Contents of legislation and 45% 36% 18% 0% 0% 40% 50% 10% 0% 0% 0% 40% 60% 0% 0% 10% 40% 40% 10% 0%
guidelines
Optimise conflicting legal 62% 38% 0% 0% 0% 45% 55% 0% 0% 0% 18% 36% 36% 9% 0% 18% 18% 64% 0% 0%
framework
European legislation becomes 36% 27% 27% 9% 0% 30% 30% 30% 10% 0% 0% 40% 50% 10% 0% 0% 20% 50% 30% 0%
nationally binding law incl.
penalties for non-compliance
Take into account specific national 0% 54% 23% 8% 15% 0% 46% 38% 8% 8% 0% 0% 54% 38% 8% 0% 23% 54% 8% 15%
requirements
Control over industry, power to 46% 31% 23% 0% 0% 23% 54% 23% 0% 0% 0% 23% 69% 0% 8% 8% 23% 69% 0% 0%
enforce requirements
Stronger political support for PhV 42% 50% 8% 0% 0% 25% 33% 42% 0% 0% 17% 8% 67% 8% 0% 8% 17% 58% 17% 0%
Solve problems with different 25% 33% 42% 0% 0% 17% 33% 50% 0% 0% 25% 25% 42% 8% 0% 33% 25% 25% 8% 8%
national languages
7.2 Staff 69% 31% 0% 0% 0% 42% 38% 15% 0% 0% 8% 54% 23% 0% 0% 15% 50% 23% 8% 0%
Number of internal staff in agency 46% 46% 8% 0% 0% 31% 54% 15% 0% 0% 23% 54% 23% 0% 0% 42% 50% 0% 8% 0%
Continuous education of staff 77% 23% 0% 0% 0% 54% 38% 8% 0% 0% 8% 69% 15% 8% 0% 15% 54% 23% 8% 0%
Include PhV into university 69% 31% 0% 0% 0% 42% 8% 50% 0% 0% 8% 25% 67% 0% 0% 0% 25% 58% 17% 0%
education
7.3 General quality 46% 46% 8% 0% 0% 38% 33% 33% 0% 0% 0% 46% 46% 8% 0% 0% 38% 23% 8% 0%
Internal quality management 54% 46% 0% 0% 0% 54% 38% 8% 0% 0% 0% 46% 46% 8% 0% 0% 54% 15% 31% 0%
programme of the agency
Avoid duplication of work 46% 38% 15% 0% 0% 38% 23% 38% 0% 0% 38% 46% 15% 0% 0% 23% 38% 23% 8% 8%
Public trust in the system 42% 50% 8% 0% 0% 33% 33% 33% 0% 0% 0% 0% 92% 8% 0% 0% 17% 75% 8% 0%
Please give and evaluate additional important indicators:
Advisor2
public understanding of risk 100% 0% 0% 0% 0% 0% 100% 0% 0% 0% 100% 0% 0% 0% 0% 0% 0% 0% 100% 0%
expertise for design of 100% 0% 0% 0% 0% 100% 0% 0% 0% 0% 100% 0% 0% 0% 0% 100% 0% 0% 0% 0%
pharmacovigilance plans
routine outcome measures 100% 0% 0% 0% 0% 100% 0% 0% 0% 0% 0% 0% 100% 0% 0% 0% 0% 100% 0% 0%
Advisor5
Jobs at the nationl agency 100% 0% 0% 0% 0% 0% 0% 100% 0% 0% 0% 100% 0% 0% 0% 0% 100% 0% 0% 0%
attractive for "the best people"
Annex 4: Results of Delphi survey on performance

indicators
19.07.2005
Analysis: Delphi sheet for performance indicators Round-1:

Values relative to number of respective answers
Relevance: How important is the indicator to obtain a valid picture of the performance of
the European System for Pharmacovigilance?
3: very relevant … 0: not relevant
Practicability: How easy is it to obtain the data for this indicator?
3: very easy to measure … 0: measurable only at very high costs
We suppose that the data would have to be collected by the national agency or come from other sources.
Please assume the availability of data in the country/region for which your agency is r
Interpretation: How easy is it to interpret the results?
3: very easy to interpret … 0: nearly not interpretable
Evaluation Round 125
Relevance Practicability Interpretation
Performance indicator 3 2 1 0 3 2 1 0 3 2 1 0
1. ...for the input

Total number of ICSRs from your country received in last year 67% 22% 11% 0% 88% 0% 13% 0% 56% 22% 22% 0%
Number of ICSRs from your country received in last year from MAHs 56% 33% 0% 11% 75% 25% 0% 0% 67% 22% 0% 11%
Number of ICSRs from your country received in last year direct from HCPs 75% 25% 0% 0% 71% 14% 14% 0% 67% 11% 22% 0%
Number of ICSRs from your country received in last year direct from 13% 25% 25% 38% 50% 0% 33% 17% 25% 0% 38% 38%
patients
Number of ICSRs from your country received in last year direct from 50% 33% 0% 17% 100% 0% 0% 0% 67% 17% 0% 17%
pharmacists
Number of ICSRs from your country received in last year direct from other 33% 50% 0% 17% 100% 0% 0% 0% 67% 17% 0% 17%
HCPs
Number of cases received/total number of ICSRs from your country 50% 38% 0% 13% 71% 0% 14% 14% 63% 13% 13% 13%
25 Numbers represent the share of answers in this field of all answers given to this item in the respective dimension.
% of serious ICSRs from your country 78% 11% 11% 0% 75% 13% 13% 0% 63% 38% 0% 0%
% of ICSRs from your country as concerned MS 50% 25% 13% 13% 29% 29% 29% 14% 25% 50% 13% 13%
% of ICSRs from your country as reference MS 50% 25% 13% 13% 29% 29% 29% 14% 25% 50% 13% 13%
Number of PSURs received by origin and type of product 44% 0% 56% 0% 63% 25% 13% 0% 33% 11% 56% 0%
Number of studies carried out on national database/ target number for 38% 38% 25% 0% 25% 25% 38% 13% 50% 25% 13% 13%
database studies
1.2 Quality of the data
% of PSURs that comply with E2C 22% 56% 22% 0% 0% 29% 57% 14% 0% 63% 38% 0%
Number of ICSRs from your country with incomplete data (i.e. less than 4 38% 38% 25% 0% 29% 43% 29% 0% 25% 38% 25% 13%
minimal data points)
Number of interventions of medical assessor because of incomplete ICSRs 25% 50% 25% 0% 14% 43% 43% 0% 29% 29% 43% 0%
from your country
Score for the quality of the spontaneous ICSRs from your country from 38% 38% 25% 0% 0% 29% 57% 14% 0% 25% 75% 0%
MAHs/HCPs/others
1.3 Ressources
Number of staff in full-time-equivalents 63% 38% 0% 0% 56% 33% 11% 0% 38% 25% 38% 0%
Number of scientists in full-time-equivalents 56% 33% 11% 0% 40% 50% 10% 0% 25% 38% 25% 13%
Number of staff per population 29% 29% 29% 14% 43% 43% 0% 14% 14% 43% 29% 14%
Annual budget of the agency 13% 75% 13% 0% 56% 22% 11% 11% 38% 13% 50% 0%
Number of Regional centres in your country 13% 50% 25% 13% 50% 25% 25% 0% 13% 38% 38% 13%
Total number of staff (sum of all regional centres) for routine work 33% 44% 22% 0% 56% 33% 11% 0% 25% 50% 25% 0%
Rating-scale: Difficulties in hiring new scientific staff (very easy...very 33% 44% 22% 0% 33% 22% 44% 0% 25% 13% 63% 0%
difficult)
1.4 Framework conditions
Number of nationally authorised products in your country 33% 11% 44% 11% 70% 20% 10% 0% 40% 10% 50% 0%
Number of MR authorised products in your country 22% 44% 22% 11% 70% 20% 10% 0% 40% 20% 40% 0%
Number of centrally authorised products in your country 33% 33% 22% 11% 89% 11% 0% 0% 40% 20% 40% 0%
Number of physicians in your country 0% 22% 22% 56% 60% 30% 10% 0% 22% 22% 33% 22%
Pharmaceutical consumption by drug classes 60% 0% 30% 10% 60% 40% 0% 0% 33% 33% 33% 0%
Pharmaceutical sales by drug classes 40% 30% 20% 10% 50% 20% 30% 0% 33% 11% 56% 0%
2. ...for the processes
2.1 Data collection

Rating-scale: Usefulness of routine data from your country for safety issue 50% 38% 13% 0% 38% 13% 50% 0% 38% 25% 38% 0%
assessment compared to other information (very useful…only marginally
useful)
Rating-scale: Access to all necessary data (very easy…very difficult) 25% 63% 13% 0% 13% 13% 63% 13% 13% 13% 75% 0%
2.2 Data management
Number of ICSRs processed 56% 44% 0% 0% 44% 44% 11% 0% 22% 56% 22% 0%
Time to assess PSURS (days from reception to finished assessment) 67% 11% 22% 0% 25% 38% 38% 0% 22% 33% 44% 0%
Rating-scale: IT-ressources: hardware, software, electronic communication 78% 22% 0% 0% 22% 33% 44% 0% 11% 44% 44% 0%
(totally sufficient…very insufficient)
Rating-scale: Internal cooperation within agency incl. IT staff (very 56% 44% 0% 0% 0% 56% 44% 0% 0% 56% 44% 0%
good…very bad)
Rating-scale: Time between data entry and transmission to EMEA or MAH 38% 50% 13% 0% 63% 38% 0% 0% 50% 50% 0% 0%
(adequate … far too slow)
Rating-scale: Information for signal detection (always sufficient…often very 89% 11% 0% 0% 0% 67% 33% 0% 11% 33% 56% 0%
incomplete)
Data sources routinely used for signal detection (routine data, literature, 56% 44% 0% 0% 11% 78% 11% 0% 33% 44% 22% 0%
registries…)
Rating-scale: Available statistical tools for signal detection (always 67% 22% 11% 0% 22% 56% 22% 0% 11% 56% 33% 0%
adequate…often very inadequate)
Rating-scale: Time between detection of signal and reporting (publishing) 78% 11% 11% 0% 0% 33% 67% 0% 11% 44% 44% 0%
(adequate … too slow)
2.3.1 International share of work in
signal detection
Rating-scale: Work that is done within your country and at the same time in 75% 13% 13% 0% 13% 38% 50% 0% 13% 38% 38% 13%
other MS or on EU level (very little…very much)
Rating-scale: Use of information from other agencies (in nearly all 50% 50% 0% 0% 0% 63% 38% 0% 0% 63% 25% 13%
cases…very seldom)
Number of PhVWP meetings at which one member of the agency has 25% 38% 38% 0% 38% 38% 25% 0% 0% 43% 29% 29%
participated
2.4.1 Share of responsibilities in safety issue assessment
Number of PSURs assessed 25% 50% 25% 0% 71% 29% 0% 0% 33% 17% 33% 17%
Number of assessment reports written per population 11% 33% 33% 22% 22% 56% 11% 11% 0% 25% 38% 38%
Rating-scale: National capabilities to identify and assess signals (fully 56% 33% 11% 0% 11% 22% 67% 0% 25% 0% 63% 13%
available…nearly not present)
Rating-scale: MAHs compliance with duty to assess safety issues (very 67% 33% 0% 0% 0% 63% 38% 0% 13% 50% 25% 13%
good…very bad)
2.4.2 Expertise for safety issue
assessment
Rating-scale: Availability of external expertise in your country for routine 44% 56% 0% 0% 22% 22% 56% 0% 11% 44% 44% 0%
cases (always when necessary…very scarce)
Rating-scale: Availability of external expertise in your country for 56% 33% 11% 0% 22% 44% 33% 0% 22% 44% 33% 0%
exceptional cases (always when necessary…very scarce)
2.5 Decision-making
Rating-scale: Come to adequate decisions (for NAPs/MRPs/CAPs) 75% 13% 13% 0% 25% 50% 13% 13% 13% 38% 38% 13%
(always…seldom)
Rating-scale: Come to decisions in good time (for NAPs/MRPs/CAPs) 50% 38% 13% 0% 25% 50% 13% 13% 13% 50% 25% 13%
(always…seldom)
2.6 Communication/Action
2.6.1 Timeliness of Communication/Action
Mean time from 1st ICSR to action with respect to this safety issue 30% 40% 20% 10% 0% 40% 40% 20% 10% 30% 60% 0%
Rating-scale: Time from 1st signal to action with respect to this safety issue 50% 38% 13% 0% 13% 25% 50% 13% 13% 38% 50% 0%
Rating-scale: Implement decisions in good time (for NAPs/MROs/CAPs) 56% 33% 11% 0% 0% 33% 44% 22% 11% 56% 22% 11%
(always…seldom)
Rating-scale: Reaching targets for timing of communications (very 40% 50% 10% 0% 10% 30% 50% 10% 20% 20% 60% 0%
good…very bad)
2.6.2 Comprehensiveness of
Communication/Action
Number of information events for HCPs with participation of agency 40% 40% 20% 0% 60% 40% 0% 0% 10% 40% 50% 0%
Number of responses to inquiries by HCPs 40% 50% 10% 0% 70% 10% 20% 0% 0% 70% 30% 0%
Number of letters to MAHs to amend SPCs 30% 40% 30% 0% 70% 30% 0% 0% 20% 30% 50% 0%
Number of variations evaluated 22% 67% 11% 0% 56% 33% 11% 0% 11% 33% 56% 0%
Number of answers to CHMP 33% 44% 22% 0% 44% 56% 0% 0% 22% 33% 33% 11%
Number of bulletins issued 30% 30% 40% 0% 70% 30% 0% 0% 20% 20% 50% 10%
Number of other answered queries 20% 20% 50% 10% 60% 0% 40% 0% 20% 10% 50% 20%
Number of inspections of MAHs carried out where PhV was an issue (at 60% 10% 30% 0% 80% 10% 10% 0% 25% 38% 38% 0%
least partially; including inspections that were carried out by other
authorities in the coutry)
Rating-scale: Reached all relevant stakeholders (always…seldom) 33% 44% 22% 0% 22% 33% 44% 0% 0% 50% 50% 0%
Rating-scale: Consistency of communication across stakeholders (incl. 30% 50% 20% 0% 20% 30% 50% 0% 0% 44% 56% 0%
MAHs) (very good…very bad)
Rating-scale: Consistency of communication across NCAs and EMEA (very 44% 44% 11% 0% 10% 40% 40% 10% 0% 44% 44% 11%
good…very bad)
2.7 General factors
2.7.1 Amount of work done
% of work acting for the Community (as Rapporteur) 44% 22% 22% 11% 33% 33% 33% 0% 11% 22% 56% 11%
% of work acting for the Community (as Reference Member State) 44% 22% 22% 11% 33% 33% 33% 0% 11% 33% 44% 11%
% of work acting on a nationally licensed product 44% 11% 33% 11% 33% 33% 11% 22% 11% 33% 33% 22%
Number of documents prepared (legal acts, quidelines) 40% 10% 40% 10% 60% 20% 10% 10% 20% 40% 20% 20%
Number of scientific publications with at least one author from the agency 20% 40% 40% 0% 70% 30% 0% 0% 10% 40% 40% 10%
in last year
2.7.2 Realised timing of work
Rating-scale: Compliance of agency with dates/requirements (very 70% 20% 10% 0% 50% 40% 10% 0% 30% 60% 10% 0%
good…bad)
Rating-scale: Meeting general targets for timing (very good…very bad) 60% 40% 0% 0% 40% 40% 20% 0% 30% 60% 10% 0%
Rating-scale: Compliance of MAHs with 15 days (very bad…very good) 60% 20% 20% 0% 50% 30% 20% 0% 20% 70% 10% 0%
2.7.3 Cooperation
Rating-scale: Compliance of MAHs with legal requirements (very bad…very 50% 20% 30% 0% 30% 40% 30% 0% 10% 70% 20% 0%
good)
Rating-scale: Cooperation with MAHs (very bad…very good) 20% 30% 50% 0% 10% 40% 30% 20% 0% 70% 10% 20%
Rating-scale: Cooperation with HCPs (very bad…very good) 30% 20% 50% 0% 20% 30% 20% 30% 10% 50% 10% 30%
Rating-scale: Collaboration between NCAs and EMEA (very bad…very 40% 40% 20% 0% 20% 40% 30% 10% 10% 60% 20% 10%
good)
Number of documents sent through EudraNet (RAS, NUIS, others) by 10% 30% 50% 10% 60% 0% 40% 0% 20% 30% 30% 20%
sender, concerned MS, issue, channels
Number of regular meetings with external experts 10% 50% 40% 0% 50% 50% 0% 0% 10% 30% 40% 20%
Number of irregular consultations with external experts 10% 40% 50% 0% 50% 40% 10% 0% 10% 30% 40% 20%
2.7.4 Quality management
% of staff trained per year 40% 60% 0% 0% 50% 50% 0% 0% 30% 50% 20% 0%
Number of training measures (internal or external) with at least one 30% 40% 30% 0% 50% 40% 10% 0% 40% 20% 40% 0%
participant from the agency
3. … for the impacts/outcomes
3.1 Impact of communications/actions
Number of ICSRs from your country before vs. after communication 50% 40% 10% 0% 50% 40% 10% 0% 20% 50% 30% 0%
Total reporting rate per million inhabitants in 2004 50% 40% 10% 0% 56% 44% 0% 0% 20% 80% 0% 0%
Reporting rate in children per million inhabitants in 2004 50% 40% 10% 0% 56% 33% 11% 0% 20% 80% 0% 0%
Satisfaction of health care givers, patient groups, prescribers with the work 50% 20% 30% 0% 10% 30% 50% 10% 10% 70% 20% 0%
of the agency (survey in these groups)
Satisfaction of prescribers with information (survey in these groups) 50% 40% 10% 0% 10% 30% 50% 10% 10% 50% 40% 0%
Number of market withdrawals of drugs (compared to other countries) 20% 30% 50% 0% 80% 10% 10% 0% 20% 30% 50% 0%
Number of suspensions of marketing authorisation 20% 40% 40% 0% 70% 20% 10% 0% 20% 40% 40% 0%
Number of dear doctor letters sent 10% 60% 30% 0% 80% 10% 10% 0% 0% 30% 70% 0%
Number of changes in SPCs made 20% 50% 30% 0% 60% 10% 30% 0% 10% 40% 50% 0%
Number of applications for variations adopted/refused 10% 50% 40% 0% 50% 30% 20% 0% 0% 60% 40% 0%
Number of variations not validated 0% 25% 75% 0% 63% 25% 13% 0% 0% 38% 50% 13%
Statistics on use of agency's web services 22% 44% 22% 11% 78% 22% 0% 0% 22% 22% 33% 22%
3.2 Outcomes
Incidence of ADR-relevant diseases 70% 20% 10% 0% 10% 30% 40% 20% 20% 50% 30% 0%
Hospitalisations due to ADR 70% 20% 10% 0% 10% 40% 50% 0% 10% 70% 20% 0%
Mortality due to ADR 70% 20% 10% 0% 10% 20% 70% 0% 10% 70% 20% 0%
Number of quality-adjusted life years lost due to ADRs 50% 40% 10% 0% 0% 0% 80% 20% 10% 50% 40% 0%
Potential years of life lost due to Adverse effects from medicines 30% 60% 10% 0% 0% 10% 50% 40% 0% 50% 40% 10%
Changes in consumption data 50% 20% 30% 0% 20% 20% 60% 0% 10% 40% 50% 0%
(Change in) Prescription data (controlled for population parameters) 40% 40% 20% 0% 30% 0% 70% 0% 10% 30% 60% 0%

Assessment of The European Community System of Pharmacovigilance

Uploaded by

Copyright:

Available Formats

Assessment of The European Community System of Pharmacovigilance

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Assessment of The European Community System of Pharmacovigilance

Uploaded by

Copyright:

Available Formats

Assessment of the European

Final Report – Final version

Reference: Service Contract No: FIF.20040739

List of tables ................................................................................................... 19

3 System analysis and description of the status quo.............................. 35

3.4 Other framework conditions ...................................................... 78

4 Goals in respect of effectiveness and efficiency ................................ 123

5 Critical success factors......................................................................... 124

6 Performance indicators ......................................................................... 127

7 Case studies ........................................................................................... 131

8 Best practice........................................................................................... 134

9 Discussion of strengths and weaknesses of the European

10 Recommendations ................................................................................. 163

10.3 Data management................................................................... 167

Annex 1: Literature....................................................................................... 172

Annex 2: Questionnaire for Agency survey ............................................... 176

Annex 3: Results of Delphi survey on critical success factors................ 186

Annex 4: Results of Delphi survey on performance indicators ............... 193

Overview and aim of the study

Table 0.1. Total national staff for PhV per capita

PhV staff NCA1 PhV staff NCA+RC2

Staff for pharmacovigilance, scientific and administrative.

Source: Fraunhofer ISI 2005

1 National Competent Authority

The system draws strongly on the combination of expertise: expertise, assess-

Figure 0.1. Cooperation between national agencies and EMEA

Source: Fraunhofer ISI 2005

Figure 0.2. Compliance of MAHs in analysis of signals

Source: Fraunhofer ISI 2005

Systematic quality management is not implemented in most PhV departments.

Figure 0.3. Preparation for last crisis by routine data

Source: Fraunhofer ISI 2005

Table 0.2. Existence and use of data on the consumption of medicines

Source: Fraunhofer ISI 2005

Safety issue assessment

Figure 0.4. Receive support from experts routinely

Source: Fraunhofer ISI 2005

Figure 0.5. Decisions for safety issues found in adequate time

0% 20% 40% 60% 80% 100%

Source: Fraunhofer ISI 2005

Communication and action

Figure 0.6. Influence of agencies' communications on prescription behaviour

Source: Fraunhofer ISI 2005

Abbreviation of country names

Figure 0.2. Compliance of MAHs in analysis of signals...................................................8

Figure 0.3. Preparation for last crisis by routine data ......................................................9

Figure 0.4. Receive support from experts routinely.......................................................11

Figure 0.6. Influence of agencies' communications on prescription

Figure 3.1. Authorisation Procedures ............................................................................38

Figure 3.2. Regulation (EC) No 726/2004 .....................................................................39

Figure 3.3. Directive 2001/83/EC ..................................................................................41

Figure 3.4. Related guidance documents......................................................................42

Figure 3.5. Key points guideline E2A ............................................................................44

Figure 3.6. Key points guideline E2B(M) .......................................................................44

Figure 3.7. Key points guideline E2C ............................................................................45

Figure 3.8. Key points guideline E2C Addendum..........................................................46

Figure 3.9. Key points guideline E2D ............................................................................47

Figure 3.10. Key points guideline E2E ..........................................................................48

Figure 3.11. Key requirements to the pharmacovigilance systems of MAHs ................51