See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/266140114

Six Sigma applied to reduce patients' waiting time in a cancer pharmacy

Article  in  International Journal of Six Sigma and Competitive Advantage · January 2014

DOI: 10.1504/IJSSCA.2014.064256

CITATIONS READS

15 6,091

4 authors, including:

Mazen Arafah Mahmoud Barghash

The American University of the Middle East University of Jordan
31 PUBLICATIONS   491 CITATIONS    38 PUBLICATIONS   221 CITATIONS   

SEE PROFILE SEE PROFILE

Eman Sallam
Indiana University Bloomington
1 PUBLICATION   15 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Critical Success Factors for Enterprise Resource Planning Systems from the Stakeholders’ Perspective: The Case of Jordan View project

Using Six Sigma DMAIC Methodology and Discrete Event Simulation to Reduce Patient Discharge Time in King Hussein Cancer Center View project

All content following this page was uploaded by Mazen Arafah on 11 February 2015.

The user has requested enhancement of the downloaded file.

Int. J. Six Sigma and Competitive Advantage, Vol. 8, No. 2, 2014 105

Six Sigma applied to reduce patients’ waiting time

in a cancer pharmacy

Mazen Arafeh*, Mahmoud A. Barghash

and Eman Sallam
Faculty of Engineering & Technology,
The Department of Industrial Engineering,
The University of Jordan,
Amman 11942, Jordan
E-mail: [email protected]
E-mail: [email protected]
E-mail: [email protected]
*Corresponding author

Alaa AlSamhouri
The Department of ERP System,
Head Office,
Anabtawi Sweets Company,
550 Amman 11941, Jordan
E-mail: [email protected]

Abstract: Six Sigma process improvement methodology has been applied to

reduce patients’ waiting time in an outpatient pharmacy located in a cancer
treatment hospital. Data concerning patients’ satisfaction has been collected
and analysed. Discrete event simulation (DES) model and design of
experiments are utilised as a decision support tool to optimise staffing
requirements. Throughout the different project phases, various improvement
opportunities have been proposed to reduce patients waiting time. Sensitivity
analysis was also performed to test the robustness of the processes against
possible changes in the availability of staff in the pharmacy. As a result of
implementing Six Sigma methodology, patients’ waiting time are reduced by
50%.

Keywords: patient satisfaction; Six Sigma; healthcare; DES; discrete event

simulation; design of experiment; sensitivity analysis; cancer treatment
hospital; outpatient pharmacy.

Reference to this paper should be made as follows: Arafeh, M.,

Barghash, M.A., Sallam, E. and AlSamhouri, A. (2014) ‘Six Sigma applied to
reduce patients’ waiting time in a cancer pharmacy’, Int. J. Six Sigma and
Competitive Advantage, Vol. 8, No. 2, pp.105–124.

Biographical notes: Mazen Arafeh is currently an Assistant Professor in the

Industrial Engineering Department at The University of Jordan. He received his
PhD in Industrial and Manufacturing Engineering from Cleveland State
University in 2001. He is a certified six sigma black belt (CSSBB) from ASQ.
He is also a certified EFQM Excellence assessor. His current research includes
applied Operations Research and Statistics, Quality in general, and specifically

Copyright © 2014 Inderscience Enterprises Ltd.

106 M. Arafeh et al.

in healthcare. He conducted several training workshops in Lean Six sigma

and other quality related topics. He is currently a member in ASQ, EFQM, and
JEA.

Mahmoud A. Barghash is currently an Associate Professor in the Industrial

Engineering Department at The University of Jordan. He received his
PhD in Manufacturing Engineering from Bradford University in 2000,
received his MS in Manufacturing Systems and Management from Bradford
University in 1995, and received his BS in Industrial Engineering from
The University of Jordan in 1992. His current research interests include
simulation, manufacturing processes, artificial intelligence, fuzzy logic and
genetic algorithms. He is currently a member in SME and JEA.

Eman Sallam graduated with BS in Industrial Engineering from The University

of Jordan in 2010. She is a member in JEA and JSQ (Jordanian Society for
Quality).

Alaa AlSamhouri is currently working as an ERP administrator at Anabtawi

Sweets Company. She is also pursuing her Masters degree in Quality
Management at The University of Jordan. She graduated with BS in Industrial
Engineering from the University of Jordan in 2011. She is a member at JEA
and JSQ.

1 Introduction

For many years, the healthcare industry focused on quality improvement initiatives, such
as continuous quality improvement (CQI) and total quality management (TQM).
However, the momentum and popularity of TQM was lost due to the lack of data-driven
analysis, and many managers and physicians became disillusioned with the prospects of
quality improvement (Langabeer et al., 2009). In recent years, Six Sigma has been
introduced to establish and sustain competitive advantage and achieving long-term goals.
Six Sigma offers time and cost savings, revenue enhancement, metrics to indicate current
process, outcome or service quality, and a focus on improving critical to quality
characteristics vital to internal and external customers (Black and Revere, 2006).
Six Sigma was first introduced by Motorola in the late 1980s. It was further developed by
General Electric in the late 1990s (Breyfogle, 2003; Pyzdek and Keller, 2003).
Six Sigma advocates seeking high quality results having as an ultimate goal of
virtually all products, attributes, or services producing fewer than 3.4 defects per million.
A 3.4 defect per million is equivalent to a Six Sigma quality level (SQL). This goal might
be difficult to achieve; however, applying Six Sigma methodology and tools will help
improve quality of service nonetheless.
Six Sigma started on the shop floor and then moved into the front offices (Munro,
2009). Nowadays, Six Sigma is being used in many organisations and environments (for
example, manufacturing, services, schools, and banks).
Hospital personnel experience time constraints in providing optimal patient care.
Workplace stress and overload occur when the job expectations are high and the ability to
make decisions and problem-solve are low (Bojtor, 2003). The impact on safety can be
measured by treatment delays, higher error rates (JCAHO, 2004). Thus, Six Sigma
 Six Sigma applied to reduce patients’ waiting time in a cancer pharmacy 107

projects in this respect can be implemented to reduce patient over-waiting and other
symptoms of the overcrowding problem.
The Six Sigma principles has been utilised in a number of cases in the healthcare
environment to improve both patient care and business processes (Sehwail and Deyong,
2003; Southard et al., 2012). A comprehensive review of Six Sigma applications in
healthcare industry by Taner et al. (2007) reveals that Six Sigma principles have
improved the overall operations of the healthcare organisations utilising them. Since then,
others have also applied Six Sigma framework in healthcare (Kumar and Bauer, 2009;
Mandahawi et al., 2010, 2011; Kumar and Kwong, 2011, Taner et al., 2012, Taner,
2013).
This work presents an application of the Six Sigma DMAIC methodology to improve
processes in an outpatient pharmacy in a cancer hospital. Specifically, the work aims at
increasing patient satisfaction by reducing prescription preparation time. DES and design
of experiments are utilised to examine the different factors affecting the performance of
pharmacy operations.

2 Methodology

The define, measure, analyse, improve, and control methods (DMAIC) present a clear
strategy for the deployment of Six Sigma projects (Sehwail and Deyong, 2003).
In the define phase, we start by identifying improvement opportunities, clarifying
scope, and defining goals. Identifying improvement opportunities is accomplished by
capturing the voice of the customer (VOC). Customers may be classified as internal or
external. In healthcare, internal customers are directors, physicians, clinicians, nurses,
and employees, external customers are patients (Carey and Lloyd, 2001). In the measure
phase, data collection is carried out and the initial performance is measured against
customer requirements. Sampling data are compared with the baseline and the SQL is
calculated. The analyse phase consists of analysis of the root cause(s) accounting for the
errors or defects that are quantified by the data collected. In the improve phase,
alternative solutions to eliminate the root cause(s) or errors or defects are examined and
the optimal one(s) are selected; the system performance is then evaluated after
implementing the process improvements. The control phase entails development and
implementation of a monitoring system to reduce future errors, documentation of results,
and recommendations for additional action.

3 Case study

The Six Sigma project is implemented at a local hospital specialised in cancer treatment.
The outpatient pharmacy at the hospital serves approximately 33,350 patients each year.
Prescriptions are classified depending on the number of medications requested in each
prescription. AB prescriptions are prescriptions with one or two medication types.
Table 1 illustrates the different prescription categories.
108 M. Arafeh et al.

Table 1 Prescription categories

Category Number of medications prepared

AB 1 or 2
CD 3 or 4
EF 5 or more

Figure 1 presents a swimlane flowchart describing the prescription preparation work

flow. The reception process begins when a patient walks through the doors (walk-in
patient), a patient calls in for a refill, or when a prescription is received by fax for a
discharged patient. All types of prescriptions are handled by one staff member, who
collects the patient’s personal information, gives the walk-in patient a queue ticket, and
attaches a copy of the ticket to the prescription itself. Each ticket contains a letter
(prescription category as explained above) and a queue number.

Figure 1 Prescriptions preparation swimlane flowchart (see online version for colours)

Following the reception process, the prescription waits in a queue for the data entry
process, where the patient’s profile is updated. Data entry is performed by two staff
members. The prescription then waits in another queue before it is picked up and moved
to the filling area, where actual preparation of the prescription takes place. In this work,
this step will be referred to as the ‘filling step’. Because of the busy work schedule, the
prescription waits in a third queue before a pharmacy staff member starts the filling step.
Three staff members are available for this step. Once the prescription is filled, it waits in
a queue before it is checked manually for the correct dosage and by a computer in order
to confirm the right prescription is prepared for the right patient. Checking is performed
by one staff member. Prescriptions are then dispensed to the patients by one staff
member.
 Six Sigma applied to reduce patients’ waiting time in a cancer pharmacy 109

4 Application of Six Sigma DMAIC methodology

Figure 2 illustrates the details of the used methodology and the inputs and outputs of each
of the DMAIC phases.

Figure 2 Basic tools used in each phase of the Six Sigma project (see online version for colours)

4.1 Define stage

The project black belt and champion described and scoped the project. They also met
with selected team members, explained to them the project objectives and importance,
discussed with them their roles, and listened to their feedback.
110 M. Arafeh et al.

The team prepared a detailed process map as shown in Figure 1. One of the main
purposes of drawing a process map is to clarify processes, roles, and set a clear scope for
the project. The process map enlightened the pharmacy staff members because they have
never seen their processes modelled end-to-end. They only knew fragments of the
process and were unaware of how what they did fits into the overall outcome.
Figure 3 presents the scope, operational definitions, and project objectives. The
purpose of the project is to achieve the objectives stated within the scope and under the
operational definitions, as agreed by the improvement team.

Figure 3 Project scope and objectives

To collect the VOC, a survey was prepared and distributed to a random sample of
patients at the pharmacy. The random sample included patients from the three
prescription categories. The survey aims at identifying the important factors ‘drivers’
directly affecting patients’ satisfaction. In the survey patients are asked to specify the
wait time acceptable to them. Figure 4 shows how patient satisfaction decreased with
increasing wait time. The figure is used to select upper specification limits for
prescription preparation time. For example, to achieve customer satisfaction of 80% or
more for patients waiting for AB prescription, wait time must be 15 minutes or less. After
consultation with the project champion, 80% patient satisfaction was selected for all
prescription categories.
Analysis of the survey results identified three major patient satisfaction drivers.
Patients’ satisfaction was driven by the patient receiving fast, professional, and quality
service. Figure 5 presents a critical-to-quality characteristics (CTQC) tree diagram which
is used to translate the general drivers (What’s) into specific (How’s). Using this tool,
each of the major drivers is broken down into secondary drivers.

4.2 Measure phase

In order to assess the initial state of the pharmacy's system, the measure phase starts with
data collection, followed by analysing the initial state and calculating the pharmacy’s
initial SQL.
Data collection points were identified after studying the pharmacy layout and
observing the pharmacy in operation. The data collection points identified in Figure 1 are
selected such that minimum interference with work flow could occur. Two members of
the improvement team were assigned data collection duty; one at the data entry area and
one at the filling area. Data collection forms were designed, tested, and modified as
needed. Data was collected from 8:00 a.m. to 5:00 p.m. (pharmacy working hours) for
13 workdays. The data collected includes prescription categories and arrivals, processing
 Six Sigma applied to reduce patients’ waiting time in a cancer pharmacy 111

times at each step (data entry, filling, and checking), and time spent in each queue,
in addition to the number of prescriptions returned or delayed because of errors.
Data collection revealed that half the prescriptions handled are AB prescriptions as
shown in Table 2.

Figure 4 Patient satisfaction vs. wait time (see online version for colours)

Figure 5 CTQC tree diagram for the case study

Table 2 Prescription categories

Category Percentage of total prescriptions

AB 51
CD 26
EF 23
112 M. Arafeh et al.

After data collection is completed, individual-moving range (I-MR) control charts are
created to identify the presence of special cause variations in each of the pharmacy
processes. Special causes of variation include events that the pharmacy staff, when
properly alerted can remove or adjust. Figure 6 presents an I-MR for AB prescriptions
where points identified by 1 are special causes. Special cause events will be addressed in
the analyse phase.

Figure 6 I-MR control charts (see online version for colours)

Process capability analysis (PCA) is performed using Minitab 15 to assess the pharmacy
performance in the initial state. The main purpose of a capability study is to determine
whether a process is capable of meeting certain requirements (Borror, 2009). Figure 7
shows a sample result for PCA performed for AB prescriptions. Capability analysis
involves the calculation of percentage of defect and their corresponding SQL.
The calculated SQL at and the percentages of defective prescriptions are summarised
in Table 3. For example, 58% of the pharmacy’s AB prescription customers (patients or
family members of patients) had to wait for more than 15 minutes. This is equivalent to a
process running at 1.27 SQL.
 Six Sigma applied to reduce patients’ waiting time in a cancer pharmacy 113

Figure 7 Process capability for AB prescriptions (see online version for colours)

Table 3 Process capability analysis (initial state)

PPT (min)
Category Mean Standard deviation % >USL SQL
AB 17.09 5.11 58 1.27
CD 31.62 20.3 34 1.94
EF 58.46 39.2 70 1.03
PPT: Prescription preparation time.

4.3 Analyse phase

After observing the processes at the pharmacy and collecting data, efforts in the analyse
phase are focused on investigating the root causes of the problems in the processes.
Brainstorming sessions were held to examine the I-MR charts generated for the different
processes focusing on the special cause events leading to long prescription preparation
times. A summary of the findings is presented using a cause and effect (C&E) diagram as
illustrated in Figure 8.
Errors are mainly found in data entry and in filling areas. Data entry staff are
answering phones calls and receiving new prescriptions from patients while entering
patients’ and prescriptions’ information in the hospital medication system. This leads to
committing mistakes, fixing them, and thus prolonged prescription preparation times. The
analysis of the data collected reveals that one third of EF prescriptions contained at least
one error. Table 4 presents the proportion of prescriptions with errors in data entry and
filling for each category.
114 M. Arafeh et al.

Figure 8 Cause and effect diagram of possible delays in prescription preparation

Table 4 Entry and filling error percentages

Prescription category Percentage of data entry errors Percentage of filling error

AB 8.2 4.4
CD 19.5 9.0
EF 33.5 16.2

EF prescriptions have the highest error percentages. This can be explained by the fact that
EF prescriptions have the largest number of medications to be prepared, which raises the
likelihood of errors occurrence.
Table 5 presents a brief description of the causes identified in the C&E diagram.

Table 5 Brief description of causes of long prescription preparation time and suggested
solutions

Brief description Suggested solutions (Improve-phase)

Process • The lack of process 1 Establish a mentoring program;
standardisation standardisation is a major where each well trained and
contributor to high variability in experienced staff is teamed with new
process times. Special cause novice staff. This created an
events are traced to an environment of cooperation between
unexperienced pharmacist staff members and ensured
handling the prescription standardisation in processes and
proper training to new pharmacists
Pharmacy layout and • Filing cabinets are not located 2 Organise the pharmacy and modify
office organisation where needed the layout to ensure ease of access to
medications and minimise
• Medication counting instruments congestions in the pharmacy. Office
are not used reorganisation also ensures that the
• Finished prescriptions are first-in-first-out queue system is
misplaced implemented
 Six Sigma applied to reduce patients’ waiting time in a cancer pharmacy 115

Table 5 Brief description of causes of long prescription preparation time and suggested
solutions (continued)

Brief description Suggested solutions (Improve-phase)

Queue system • The pharmacy staff should be • Solutions: 1, 2
processing prescriptions in a first-
in-first-out mode. However, due
to the lack of office organisation
and the fact that not all staff
members adhered to this. Many
prescriptions were left
unprocessed for long periods of
time
Interruption • Pharmacy staff at data entry are 3 Hire a full-time receptionist to
continuously interrupted to handle all patients submitting their
handle new walk-in prescriptions prescriptions, thus freeing data entry
and answer phone calls staff to focus on data entry. The
receptionist is trained on how to
greet patients and collect necessary
information
Errors • Error are found in data entry and 4 Solution 3
filling 5 Use medication counting instruments
• Data entry errors are caused by
continuous interruptions as
explained earlier
• Filling error are mainly counting
errors, caused by human error

Another important potential cause not mentioned in the table is understaffing. In the case
of this pharmacy, understaffing may occur:
• when pharmacy staff are absent for the whole day
• during lunch break
• when one or more step out of the pharmacy (tea break, toilet, etc.).
Unorganised breaks create instability in the pharmacy operations leading to long queues
and long prescription preparation times. The pharmacy management presented the
improvement team with the following two important questions:
• Is the staffing policy adequately meeting the staffing requirements?
• What are the dynamics and effects of changing staff assignment to the different
pharmacy main processes (data entry, filling, and checking) on prescription
preparation time?
Answering these two questions is detailed in the improve phase.

4.4 Improve phase

The goal of the improve phase is to implement the changes to the system that are needed
to improve it (Pyzdek and Keller, 2003). Several brainstorming sessions are held to find
116 M. Arafeh et al.

the solutions to the various potential causes presented in Figure 8. We recommend

several solutions as explained in Table 5.
Staffing requirements
Traditionally, organisations will implement process improvements on varying levels of
sophistication. At the lowest level, organisations implement what executives feel is the
right thing to do (Miller et al., 2003). The pharmacy management suggested adding one
more filling staff (increase from two to three) and to keep the data entry at two staff
members and the checking at one staff member.
To test if these changes will provide the best solution, one can try the different
scenarios on the pharmacy processes and observe the outcomes. Experimenting on an
actual system, however, without knowing the consequences can cause unneeded
disruptions and costs. Simulation provides a way to understand implementation
ramifications without disrupting the actual system (Southard et al., 2012). Simulation is
one of the tools deserving special mention as an emerging technique that can play an
important role in Six Sigma initiatives and is considered by some authors to be vital to
the long-term success of Six Sigma projects (Tjahjono and Ball, 2010; El-Haik and
Al-Aomar, 2006). Simulation techniques are usually used within the improve phase
(Tjahjono and Ball, 2010). Once validated, they yield accurate estimates of the behaviour
of the real system.
Ruohonen et al. (2006), Mandahawi et al. (2010) and Celano et al. (2012) utilised the
discrete-event simulation (DES) as decision supporting tool in the management of
improvement projects.
There are several benefits gained by including simulation in Six Sigma projects:
reduced experimentation costs, reduced project time, and improved results (Mccarthy and
Stauffer, 2001).
Computer simulation models can be integrated with other analysis methods such as
design of experiment (Yang and El-Haik, 2003). Experimenting with simulation models
is a typical practice for estimating performance under various running conditions,
conducting what-if analysis, testing hypothesis, and comparing alternatives (El-Haik and
Al-Aomar, 2006).
Design of experiment (DOE)
Design of experiments is a statistical approach to designing and conducting experiments
such that the experiment provides the most efficient and economical methods of
determining the effect of a set of independent variables on a response variable.
Knowledge of this relationship permits the experiments to optimise a process and
predict a response variable by setting the factors at specific levels (Kubiak and Benbow,
2009).
To study the effect of changing the number of staff members at each major process
(data entry, filling, and checking), we design an experiment with three factors.
Each factor will be run under different levels. Data entry will be run under two levels: 2
staff members (current) and 3 staff members (new). Filling will be run under three levels:
3 staff members (current) and 2 or 4 staff members (new). Finally, checking we will run
under two levels: 1 staff member (current) and 2 staff members (new). The full factorial
design including the different factors with their levels is presented in Table 6.
 Six Sigma applied to reduce patients’ waiting time in a cancer pharmacy 117

Table 6 Factor and their corresponding levels

Factor DOE level 1 DOE level 2 DOE level 3

Data entry 2 3
Filling 2 3 4
Checking 1 2
The underscored levels are the currently applied.
The italicised levels are recommended by pharmacy management.

Discrete event simulation

The conceptual model is documented with Microsoft Visio® as shown in Figure 9. The
model included trigger objects to show prescription arrival. It also shows activities, which
describe inputs, outputs, required resources, activity durations, and business rules.
Finally, the model included decision points for routing prescriptions and objects to show
the end of the process. The programming phase included coding the process model into
Promodel® software. Blocks B1–B3 in Figure 9 represent the arrival process, whereby an
arrival is made by the maximum arrival during the day and the excess entities are routed
out in Block B3; thus only the required number of entities is entered into the system.
Following the arrival process, blocks B3–-B17 shows the processing of the entities from
the reception to the dispensing. The ‘USE’ statement allocates the necessary resource for
each operation and the duration of use. The entity name is changed to reflect the
occurrence of error and the difference in processing time. Blocks (B4–B17) present the
various processing operations for the prescription. AB prescriptions are given priority for
any resource or location by defining a priority attribute. This attribute is placed for each
location and for each use statement to reflect the preference for AB over CD and EF.

Figure 9 Conceptual model with ProModel simulation statements (see online version for colours)
118 M. Arafeh et al.

Selecting a distribution model to represent operational data in the model is an important

step in creating an accurate representation of the real world (Mccarthy and Stauffer,
2001). Times for the various tasks within the process are randomly selected by the model
from time distributions in order to represent the stochastic nature of the pharmacy. The
parameters of the time distributions for the tasks are based on operational data collected
in the measure phase. EasyFit® software is used for fitting data distributions.
The number of resources available for each task is manipulated according to the full
factorial design explained earlier.
Hypothesis testing
Hypothesis testing is used to compare the simulated alternatives with the goal of
identifying the superior alternative relative to prescription preparation time. The
hypotheses to be tested for each prescription category are shown in Figure 10.

Figure 10 Hypothesis testing

The null hypothesis (H0) states that changing the number of data entry staff from two to
three will not affect prescription time. On the other hand, the alternative hypothesis (H1)
states that changing the number of data entry staff from two to three will have a
significant effect on prescription preparation time.
Simulation results
The simulation model is run and the generated prescription preparation time data is
analysed using Minitab 16®. The results of the analysis are summarised in Table 7 where
all p-values less than 0.05 indicate a factor with significant effect on prescription
preparation time.

Table 7 Analysis of variance test results

p-values
Factor AB CD EF
Data entry 0.104 0.048 0.001
Filling 0.000 0.006 0.000
Checking 0.000 0.000 0.000
Entry-filling 0.806 0.377 0.587
Entry-checking 0.251 0.607 0.676
Filling-checking 0.695 0.769 0.357
Entry-filling-checking 0.977 0.482 0.803
 Six Sigma applied to reduce patients’ waiting time in a cancer pharmacy 119

Observations of Table 7 are in order. We present several observations followed by a brief

discussion.
• Data entry factor has insignificant effect on prescription preparation time for AB
prescriptions.
Adding data entry staff has little effect in reducing prescription preparation time for AB
prescriptions. This is expected, since AB prescriptions are given the highest priority. On
the other hand, the effect is larger in the case of CD and EF prescriptions since they are
assigned lower priority.
• Filling factor has significant effect on prescription preparation time for all
prescription types.
Changing the number of staff at filling will significantly affect prescription preparation
times for all prescription types. However, it does not tell us how. To be able to
understand how, we use the Minitab-generated main factor effect plots presented in
Figure 11. Examining the figure reveals that the removal of a filling staff will increase the
preparation time as expected, particularly for AB prescriptions. However, the addition of
a filling staff will not reduce the prescription preparation time significantly.
• Checking factor has significant effect on prescription preparation time for all
prescription types.
The addition of a checking staff is the most significant on all prescriptions types. It will
reduce the prescription preparation time significantly.
• Factors have no interactions; the effect of one factor is not affected by the level of
other factors.
This result is logical since the addition or removal of a staff member from a location (data
entry, for example) will not affect operations at another location (filling, for example).
Figure 11(d) presents an interaction pot for EF prescriptions. We can see that all the lines
are nearly parallel, indicating absence of interactions.
Table 8 summarises the results of the analysis, indicating the changes that need to be
implemented in order to reduce the prescription preparation time.
Contrary to the expectation of the pharmacy managers, the addition of one more
filling staff will not reduce the prescription preparation time significantly. However,
reducing the staff by one will significantly increase prescription preparation time. So the
current number of filling staff (three) is optimal. Adding one more pharmacist to perform
checking will significantly reduce prescription preparation time for all prescription types.
In addition, adding a staff at the data entry will significantly reduce prescription
preparation time for prescriptions CD and EF.

Table 8 Current, management recommended and optimal recommended staffing levels

Management recommended Optimal recommended

Factor Current staffing staffing levels
Data entry 2 2 3
Filling 3 4 3
Checking 1 1 2
120 M. Arafeh et al.

Figure 11 Main effects and interaction plots (see online version for colours)

(a) (b)

(c) (d)

I-MR control charts are generated for the simulated data under optima staffing levels.
Figure 12 presents an I-MR control chart for AB prescriptions. One can easily observe
the reduction in the mean prescription preparation time and the process variability.
The reduction in prescription preparation times for AB, CD, and EF prescriptions is
shown in Table 9. The table reports a calculated reduction in prescription preparation
time by as much as 48% for AB prescriptions, 47% for CD prescriptions, and 39% for EF
prescriptions. The table also shows the improvement in the process capability expressed
in sigma quality levels.

Table 9 Comparison between the initial and the future states

PPT (min) SQL

Category Initial Improved Reduction (%) Initial Improved
AB 18.47 9.66 48 1.27 3.68
CD 31.62 16.66 47 1.94 4.02
EF 58.46 35.39 39% 1.03 2.94

4.5 Control stage: sensitivity analysis

Pyzdek and Keller (2003) note that the new system should be evaluated for risks, hence a
sensitivity analysis A sensitivity analysis is conducted to study the effects of staff
absenteeism on prescription preparation time. Table 10 shows possible scenarios as a
 Six Sigma applied to reduce patients’ waiting time in a cancer pharmacy 121

result of staff absenteeism. The simulated prescription preparation times for the different
scenarios are shown in Figure 13 with scenario 10 identified as the optimal scenario
circled for each prescription type.

Figure 12 I-MR chart for AB prescriptions at the initial (left side) and improved (right side) states
(see online version for colours)

Table 10 Different scenarios tested

1 Data entry Filling Checking

1 2 2 1
2 2 2 2
3 2 3 1
4 2 3 2
5 2 4 1
6 2 4 2
7 3 2 1
8 3 2 2
9 3 3 1
10 (Optimal) 3 3 2
11 3 4 1
12 3 4 2
122 M. Arafeh et al.

Figure 13 Prescription preparation time under the different scenarios (see online version
for colours)

By examining Figure 13, we can see that the absence of a filling staff (scenario 8) has
more significant effect on prescription preparation time than the absence of a data entry
staff (scenario 4). However, the absence of a checking staff (scenario 9) has a greater
effect when compared to the absence of a data entry staff or a filling staff. This
observation is very important since the research team noticed that when staff members
took the day off, or had a temporary leave, obtaining a temporary replacement was not
always a priority for the pharmacy manager. Also when faced with a situation where a
checking staff is absent and no outside replacement is available, it is better to find an
inside replacement, even if one of the filling staff is switched to perform checking duties
than to keep one checking staff alone.
The control phase requires that policies and procedures be established to maintain the
improvements created. Monitoring procedures must be established to ensure that the
prescription preparation time remain within acceptable levels. I-MR control charts can
assist in this control through its ability to detect shifts in process performance, especially
when they are used to gather and report information in real time, providing immediate
visibility into the process. Training and mentoring plan was put into place, new job
description for the receptionist was written, and changes in managerial policies regarding
absenteeism and temporary replacements were implemented.

5 Conclusions

In this paper, the Six Sigma DMAIC framework is used to reduce the prescription
preparation time by developing customer-oriented processes in a cancer outpatient
pharmacy.
 Six Sigma applied to reduce patients’ waiting time in a cancer pharmacy 123

VOC was collected and analysed to establish prescription preparation times

acceptable by majority of patients visiting the pharmacy. The data collected in the
measure phase showed high percentages of prescriptions taking longer time than that
accepted by patients. In-depth analysis of the pharmacy processes unfolded different
causes leading to a long prescription preparation time, specifically pharmacists
understaffing. DES coupled with design of experiment was used to analyse the system
and identify the staffing requirements of the pharmacy. The team recommended adding
one staff at data entry and one staff at checking. The team also recommended giving
priority to AB prescriptions since they are the majority of prescriptions arriving at the
pharmacy. Finally, new policies and mentoring programs are introduced to maintain
improvements.

References
Black, K. and Revere, L. (2006) ‘Six Sigma arises from the ashes of TQM with a twist’,
International Journal of Health Care Quality Assurance, Vol. 19, pp.259–266.
Bojtor, A. (2003) ‘The importance of social and cultural factors to nursing status’, International
Journal of Nursing Practice, Vol. 9, pp.328–335.
Borror, C.M. (2009) The Certified Quality Engineer Handbook, ASQ Quality Press, Milwaukee
53203.
Breyfogle, F.W. (2003) Implementing Six Sigma: Smarter Solutions Using Statistical Methods,
John Wiley & Sons, Inc., Hoboken, New Jersey.
Carey, R.G. and Lloyd, R.C. (2001) Measuring Quality Improvement in Healthcare: A Guide to
Statistical Process Control Applications, ASQ Quality Press, Milwaukee 53203.
Celano, G., Fichera, S., Costa, A. and Tringali, G. (2012) ‘Linking Six Sigma to simulation: a new
roadmap to improve the quality of patient care’, International Journal of Health Care Quality
Assurance, Vol. 25, pp.254–273.
El-Haik, B. and Al-Aomar, R. (2006) Simulation-Based Lean Six-Sigma and Design For
Six-Sigma, John Wiley & Sons, Hoboken, New Jersey.
JCAHO (2004) Managing Patient Flow, Joint Commission Resources: Oakbrook Terrace, Joint
Commission on Accreditation of Healthcare Organizations, Illinois.
Kubiak, T.M. and Benbow, D.W. (2009) The Certified Six Sigma Black Belt Handbook, ASQ
Quality Press, Milwaukee 53203.
Kumar, S. and Bauer, K. (2009) ‘The business case for implementing electronic health records in
primary care settings in the United States’, Journal of Revenue & Pricing Management,
Vol. 10, pp.119–131.
Kumar, S. and Kwong, A.M. (2011) ‘Six Sigma tools in integrating internal operations of a retail
pharmacy: a case study’, Technology and Health Care, Vol. 19, pp.115–133.
Langabeer, J., Dellifraine, J., Heineke, J. and Abbass, I. (2009) ‘Implementation of lean and
Six Sigma quality initiatives in hospitals: a goal theoretic perspective’, Operations
Management Research, Vol. 2, pp.13–27.
Mandahawi, N., Al-Araidah, O., Boran, A. and Khasawneh, M. (2011) ‘Application of lean
Six Sigma tools to minimise length of stay for ophthalmology day case surgery’, International
Journal of Six Sigma and Competitive Advantage, Vol. 6, pp.156–172.
Mandahawi, N., Al-Shihabi, S., Abdallah, A.A. and Alfarah, Y.M. (2010) ‘Reducing waiting time
at an emergency department using design for Six Sigma and discrete event simulation’,
International Journal of Six Sigma and Competitive Advantage, Vol. 6, pp.91–104.
Mccarthy, B.M. and Stauffer, R. (2001) ‘Six Sigma: enhancing Six Sigma through simulation with
igrafx process for Six Sigma’, Proceedings of the 33nd Conference on Winter Simulation,
2001, IEEE Computer Society, pp.1241–1247.
 124 M. Arafeh et al.

Miller, M.J., Ferrin, D.M. and Szymanski, J.M. (2003) ‘Simulating Six Sigma improvement ideas
for a hospital emergency department’, Simulation Conference, 2003. Proceedings of the 2003
Winter, IEEE, pp.1926–1929.
Munro, R.A. (2009) Lean Six Sigma for the Healthcare Practice: A Pocket Guide, ASQ Quality
Press, Milwaukee 53203.
Pyzdek, T. and Keller, P.A. (2003) The Six Sigma Handbook: A Complete Guide for Green Belts,
Black Belts, and Managers at All Levels, McGraw-Hill New York.
Ruohonen, T., Neittaanmaki, P. and Teittinen, J. (2006) ‘Simulation model for improving the
operation of the emergency department of special health care’, Simulation Conference, 2006.
WSC 06. Proceedings of the Winter, 3–6 December, pp.453–458.
Sehwail, L. and Deyong, C. (2003) ‘Six Sigma in health care’, Leadership in Health Services,
Vol. 16, pp.1–5.
Southard, P.B., Chandra, C. and Kumar, S. (2012) ‘Rfid in healthcare: a Six Sigma dmaic
and simulation case study’, International Journal of Health Care Quality Assurance, Vol. 25,
pp.291–321.
Taner, M.T. (2013) ‘Application of Six Sigma methodology to a cataract surgery unit’,
International Journal of Health Care Quality Assurance, Vol. 26, pp.768–785.
Taner, M.T., Sezen, B. and Antony, J. (2007) ‘An overview of Six Sigma applications in healthcare
industry’, International Journal of Health Care Quality Assurance, Vol. 20, pp.329–340.
Taner, M.T., Sezen, B. and Atwat, K.M. (2012) ‘Application of Six Sigma methodology
to a diagnostic imaging process’, International Journal of Health Care Quality Assurance,
Vol. 25, pp.274–290.
Tjahjono, B. and Ball, P. (2010) ‘Six Sigma: a literature review’, International Journal of Lean
Six Sigma, Vol. 1, p.19.
Yang, K. and El-Haik, B. (2003) Design for Six Sigma, McGraw-Hill, New York.

View publication stats