AJPH METHODS

The World Health Organization STEPwise

Approach to Noncommunicable Disease
Risk-Factor Surveillance: Methods, Challenges,
and Opportunities
Leanne Riley, MSc, Regina Guthold, PhD, Melanie Cowan, MPH, Stefan Savin, MD, MPH, Lubna Bhatti, MD,
Timothy Armstrong, PhD, and Ruth Bonita, PhD

Objectives. We sought to outline the framework and methods used by the World Health number of premature deaths from NCDs by
Organization (WHO) STEPwise approach to noncommunicable disease (NCD) surveillance one third by 2030.3
(STEPS), describe the development and current status, and discuss strengths, limitations, The key to controlling the global epidemic
and future directions of STEPS surveillance. of NCDs and meeting these ambitious but
Methods. STEPS is a WHO-developed, standardized but flexible framework for countries achievable NCD targets is primary prevention
based on comprehensive population-wide
to monitor the main NCD risk factors through questionnaire assessment and physical
programs. Effective prevention of NCDs is
and biochemical measurements. It is coordinated by national authorities of the imple-
possible through identification of the major
menting country. The STEPS surveys are generally household-based and interviewer-
common risk factors and their prevention
administered, with scientifically selected samples of around 5000 participants. and control.4–6
Results. To date, 122 countries across all 6 WHO regions have completed data col- A few common and preventable risk fac-
lection for STEPS or STEPS-aligned surveys. tors underlie most NCDs. These NCD risk
Conclusions. STEPS data are being used to inform NCD policies and track risk-factor factors are the leading cause of the death and
trends. Future priorities include strengthening these linkages from data to action on NCDs disability burden in all countries, regardless
at the country level, and continuing to develop STEPS’ capacities to enable a regular and of their economic development status. The
continuous cycle of risk-factor surveillance worldwide. (Am J Public Health. 2016;106:74– leading behavioral risk factors for NCDs
78. doi:10.2105/AJPH.2015.302962) are tobacco use, harmful alcohol consump-
tion, unhealthy diet including high salt and
sodium intake, physical inactivity, and
overweight and obesity, and the leading

T he global burden of chronic, non-

communicable diseases (NCDs)—largely
heart disease, stroke, cancer, chronic re-
voluntary targets linked to the Global
Monitoring Framework to prevent and
control NCDs by 2025, including a target to
physiological risk factors are raised blood
pressure, raised blood glucose, and abnormal
blood lipids.7,8
spiratory disease, and diabetes—is increasing reduce premature mortality from the 4 main Recognizing a global need for risk-factor
rapidly and will have significant social, eco- NCDs by 25%, and targets for the main data on these key NCD risk factors, WHO
nomic, and health consequences unless behavioral and metabolic NCD risk factors initiated the STEPwise approach to surveil-
urgently addressed. In 2012, the major NCDs and 2 health systems targets.2 Furthermore, in lance (STEPS) in 2002. The key goals of
accounted for 63% of all deaths, representing 2015, the 2030 Agenda for Sustainable STEPS are to guide the establishment
38 million deaths a year. Eighty percent of Development recognizes the importance of of risk-factor surveillance systems in countries
these deaths are already occurring in low- and addressing NCD issues with the inclusion by providing a framework and approach; to
middle-income countries.1 Because NCDs of a similarly ambitious target to reduce the strengthen the availability of data to help
are largely preventable, these deaths can be
significantly reduced.
To address this global health problem, in ABOUT THE AUTHORS
2013, the World Health Assembly—the Leanne Riley, Regina Guthold, Melanie Cowan, Stefan Savin, Lubna Bhatti, and Timothy Armstrong are with Department
decision-making body of the World Health for Prevention of Noncommunicable Diseases, World Health Organization, Geneva, Switzerland. Ruth Bonita is with
University of Auckland, Auckland, New Zealand.
Organization (WHO)—adopted a Global Correspondence should be sent to Leanne Riley, World Health Organization, 20 Avenue Appia, CH1211, Geneva, Switzerland
Monitoring Framework for NCDs with (e-mail: [email protected]). Reprints can be ordered at http://www.ajph.org by clicking the “Reprints” link.
25 key indicators to track progress in pre- This article was accepted October 26, 2015.
Note. The authors alone are responsible for the views expressed in this publication and they do not necessarily represent the decisions,
vention and control of NCDs. The World policy, or views of the World Health Organization.
Health Assembly also agreed on a set of global doi: 10.2105/AJPH.2015.302962

74 Surveillance Peer Reviewed Riley et al. AJPH January 2016, Vol 106, No. 1
 AJPH METHODS

countries inform, monitor, and evaluate their provision of general lifestyle advice to tackle including the level of confidence, the
policies and programs; to facilitate the NCDs) is collected through self-report. acceptable margin of error, the estimated
development of population profiles of NCD Physical measurements of height and weight design effect, the estimated baseline levels of
risk-factor exposures; to enable comparability to measure the body mass index (defined as the behaviors to be measured, the desired
across populations and across time frames; weight in kilograms divided by the square number of age–gender estimates, and the
and to build human and institutional capacity of height in meters), waist circumference, and anticipated nonresponse rate.12 A typical
for NCD surveillance.9,10 blood pressure are undertaken in step 2, STEPS survey includes between 5000 and
Since its inception, the STEPS approach and step 3 consists of biochemical measure- 6000 participants to be able to generate
has advocated that small amounts of good- ments of fasting blood glucose, total age- and gender-specific estimates according
quality data are more valuable than large cholesterol levels, and urinary sodium. to strata relevant by the implementing country.
amounts of poor-quality data.10 The STEPS Within each step, countries are encour-
approach supports monitoring a few modi- aged to focus on the “core” or most essential
STEPwise Approach to Surveillance
fiable NCD risk factors that reflect a large part information on each risk factor. “Expanded”
Fieldwork
of the future NCD burden and that can in- items, also in standardized format, are
STEPS surveys are implemented at the
dicate the impact of interventions considered available for those countries requiring more
country level as national household surveys
to be effective in reducing the leading NCDs. nuanced information on any of the risk
via trained interviewers who undertake
Because STEPS also promotes the collection factors, resources permitting. Expanded items
face-to-face interviews at households with
of data on a number of different risk factors, it include additional information for the core
selected survey respondents of behavioral risk
has the benefit over single risk factor surveys questions on each of the behavioral risk factors
factors and physical measurements such as
in that it allows an understanding of how of step 1, measurement of hip circumference
blood pressure and height and weight
risk factors cluster within a population and and heart rate in step 2, and biochemical
measurements (steps 1 and 2). Step-3
offers an opportunity for countries to estimate assessment of triglycerides and high-density
biochemical assessments for blood glucose,
the small proportion of the population with lipoprotein cholesterol levels in step 3. Finally,
blood lipids, and urinary sodium usually take
high overall risk of a cardiovascular event standardized “optional” modules have been
place at a local clinic or health center.
for referral for possible treatment.9,10 developed in collaboration with the respective
eSTEPS (World Health Organization,
Our objective was to outline the frame- technical departments in WHO and topic
Geneva, Switzerland), a suite of software that
work and methods used by STEPS, to area experts in violence and injury, mental
allows for the preparation and implementa-
describe the development and current status health and suicide, oral health, sexual and
tion of data collection with a handheld PC,
of STEPS, to discuss the strengths and reproductive health, and tobacco policy.9,10
has been used widely since 2009. This method
limitations of STEPS surveillance, and to The STEPS instrument has undergone
of data collection has proved to be a significant
highlight future directions. a number of revisions to accommodate new
improvement on the previous paper-based
and emerging evidence on the various risk
method in that it allows automatic skip pat-
factors over time and to ensure that STEPS
terns to be programmed into the question-
surveys track the most important aspects of
naire, performs immediate error checking,
METHODS NCD risk factors and determinants. The latest
provides a simplified and documented means
The STEPS framework is based on the revision of the STEPS instrument accommo-
of selecting participants from within each
concept that surveillance systems require dates 6 risk-factor targets and 1 health systems
selected household, and requires no data-
standardized data collection, but also suffi- target in the Global Monitoring Framework.11
entry work following data collection.
cient flexibility to be appropriate in a variety
The introduction of eSTEPS allows a
of country situations and settings.10 The
more streamlined and standardized approach
key element of this framework is the dis- STEPwise Approach to Surveillance
to quality control and substantially reduces
tinction between different levels of risk-factor Sampling
the time from data collection to production of
assessment that allow greater or lesser levels The basis of STEPS risk-factor surveillance
the final report. eSTEPS can be adapted to
of detail, depending on the resources is repeated cross-sectional, population-based
suit any survey and allows implementation
available, without compromising the household surveys.12 Multistage cluster
in a large number of different languages.
comparability of the data. sampling is used in most countries to draw
As of late 2015, eSTEPS data collection is also
In step 1, information on demographics a nationally representative sample of adults
possible with Android devices such as tablets
and behavioral risk factors (tobacco use, aged 18 to 69 years. However, sampling
and smartphones.
alcohol consumption, dietary behaviors such strategies vary across countries depending on
as fruit and vegetable intake and salt and the local context, and exceptionally, there
sodium intake, and physical inactivity, as well are a few very small or well-resourced Coordination, Technical Support,
as history of NCDs and related conditions countries that have performed simple random and Resourcing
such as raised blood pressure, diabetes, raised sampling or censuses. STEPS surveys are usually initiated and
cholesterol, cardiovascular diseases; cervical Sample sizes for STEPS surveys are cal- coordinated at the country level by ministry
cancer screening coverage in women; and culated on the basis of a number of factors, of health officials, working in collaboration

January 2016, Vol 106, No. 1 AJPH Riley et al. Peer Reviewed Surveillance 75
AJPH METHODS

with local technical partners. Financial in 2015); templates for an implementation had, by the end of that year, completed
resources to conduct the surveys are the plan and ethical approval; sampling tools, such data collection (Figure A, available as
responsibility of national authorities. The as a sample size calculator and a sampling a supplement to the online version of this
WHO provides resourcing support in spreadsheet; various forms for data collection; article at http://www.ajph.org). Of these
the form of equipment loans such as blood analysis programs for several statistical pack- surveys, 28 were national.
pressure–measuring devices, small amounts ages; and reporting templates.9,12 To date, 129 countries from all regions
of seed funding, and links to potential have attended a total of 166 regional or
international donors. national workshops, and 122 countries
Technical support for the implementation (42 WHO African Region, 21 WHO
of STEPS is provided across WHO head- RESULTS Region of the Americas, 16 WHO Eastern
quarters, regional and country offices, and in Pilot testing of STEPS began in 2002 and, Mediterranean Region, 6 WHO European
collaboration with other technical partners. following several regional workshops, several Region, 11 WHO South-East Asia Region,
To provide overall guidance on the planning countries implemented the first round of 26 WHO Western Pacific Region) have
and implementation of STEPS, STEPS surveys collecting data in 2002 and 2003.10 completed data collection (Figure 1),
training workshops are delivered at different By 2005, 56 countries had attended including 11 STEPS-aligned surveys. In
stages of the surveillance process: planning 17 regional or national workshops, and total, 93 countries have carried out national
and training, data collection, data analysis, and 52 countries (14 WHO African Region, surveys. A total of 48 countries have
reporting, as well as data utilization. 1 WHO Region of the Americas, 11 WHO already undertaken more than 1 STEPS or
For each of these stages, survey tools and Eastern Mediterranean Region, 9 WHO STEPS-aligned survey.9 Of those 48 coun-
training materials are available. They include, South-East Asia Region, 17 WHO Western tries, 21 have either moved from subnational
for example, the STEPS Manual and CD Pacific Region), including 6 that aligned to national, or done several subnational
(updated along with the STEPS Instrument STEPS to their national risk-factor studies, STEPS surveys, and 24 have done 2 STEPS

Note. The boundaries and names shown and the designations used on this map do not imply the expression of any option whatsoever on the part of the World Health
Organization concerning the legal status of any country, territory, city, or area or of its authorities, or concerning the delimitation of its frontiers or boundaries.
Dotted and dashed lines represent approximate border lines for which there may not yet be full agreement.
Source. Data are from the World Health Organization. Map production by Health Statistics and Information Systems.

FIGURE 1—Status of the World Health Organization’s STEPwise Approach to Noncommunicable Disease Surveillance Implementation in 2015

76 Surveillance Peer Reviewed Riley et al. AJPH January 2016, Vol 106, No. 1
 AJPH METHODS

or STEPS-aligned surveys, and 3 have status, the Global Burden of Diseases Project, on alcohol, nutrition, tobacco, and injuries,
conducted several STEPS surveys in a sys- and publications on global and regional levels tend to elicit similar information, leading to
tematic, ongoing way. of selected risk factors.1,8,13 survey fatigue in many countries. Because
Of the 122 countries that have completed More than a decade of implementation of STEPS can be considered the primary data
data collection to date, 10 have done STEPS has also proved that the evolving source of all NCD risk factors, other surveys
steps 1 and 2 (questionnaire and physical STEPS methodology is flexible and adaptable should consider the availability of STEPS
measurements) only, and 112 have completed enough to be used even in the most results before introducing the same items or
all 3 steps, including biochemical resource-constrained settings. indicators.
measurements.

Challenges and Limitations Future Directions

Despite this progress, a number of limi- The need for robust, comparable NCD
DISCUSSION tations, challenges, and constraints hamper risk-factor data at the national, regional, and
More than a decade has passed since a few broader progress in advancing NCD risk- global level has been further strengthened
low- and middle-income countries started factor surveillance. Although there has been by the global adoption of the set of 9 vol-
implementing STEPS surveys, and great increased global attention, global targets and untary targets for NCDs and the Global
progress has been made in generating new indicators have been set,2,11 and countries’ Monitoring Framework for NCDs indicators
country-specific risk-factor data.9 Nearly implementation of these have been moni- which include many of the risk factors
100 countries have published the results of tored, NCD risk-factor surveillance is still not monitored by a STEPS survey.11 By imple-
their STEPS surveys in form of summary data, considered a priority in many countries. menting a STEPS survey, a country will
country reports, or journal articles.9 From There is inadequate progress in integrating have the necessary data to monitor and
a bleak picture of very little data available in NCD risk-factor surveillance into national report on 7 of the 9 global targets for NCDs
2000 from most low- and middle-income health information systems. (tobacco, harmful use of alcohol, physical
countries, to 15 years on, we can see from the At the country level, progress in in- inactivity, salt and sodium intake, diabetes
mapping of STEPS activities that many stitutionalizing ongoing NCD risk-factor and obesity, hypertension, prevention of
countries now have robust data on NCD risk surveillance is hampered by a number of heart attack and strokes) and a number of
factors that they are using to report their factors. These include the often high turnover other indicators included in the Global
progress against national targets for NCDs of personnel involved in surveillance; the Monitoring Framework (e.g., total choles-
and to guide their policy and programming in lack of resources available at the country level terol, cervical cancer screening). In addition,
NCD, such as for the development of to support surveillance activities; out-of-date the Global Action Plan for the Prevention
NCD action plans in Benin, Cabo Verde, sampling frames; lack of geographical and Control of Noncommunicable Diseases
Mauritania, and Togo, to name a few. accessibility of some areas in some countries; adopted by the World Health Assembly in
Countries increasingly report that having weak infrastructure (e.g., travel and transport May 2013 calls on WHO Member States to
their own reliable data on risk factors leads systems) that makes household surveys
to an increasing awareness and commitment difficult; weak country capacity in data undertake periodic data collection on the
management, analysis, and report writing; behavioural and metabolic risk factors (harmful
to address NCDs, and is of utmost
use of alcohol, physical inactivity, tobacco use,
importance for setting meaningful NCD and technological problems with step-3 unhealthy diet, overweight and obesity, raised
program evaluation mechanisms at the measurements, to name but a few. In blood pressure, raised blood glucose, and
national level. a number of countries, these limitations hyperlipidemia). . . .2(p52)
Risk-factor surveillance has also served as have resulted in poor data quality.
an entry point to developing programs for Frequently the awareness generated by STEPS therefore remains a key approach
NCDs in a number of countries. Countries conducting a STEPS survey in a country leads and resource to help countries meet this
have started with an initial survey and followed to an increase in demand for preventive obligation.
this with the development of a national NCD and curative health services—this demand is There remains a strong need to continue
policy and program involving key stake- often unmet by the current health system in with the capacity-building aspects of STEPS
holders to set clear and achievable targets for low- and middle-income countries. Other technical support to help build sustainable
action based on their STEPS data. STEPS challenges include the dependence on WHO national capacity for NCD risk-factor sur-
“Data to Action” workshops have helped to for technical support and assistance, when veillance and institutional support for NCD
accelerate this process by focusing on how the WHO has few resources and personnel to surveillance. Increased advocacy efforts are
STEPS data can help a country consider what help meet this demand. needed at the country level to improve
its priorities for NCD programming might be Multiple surveys at national and global linkages with NCD-related data into national
on the basis of reliable, recent data. levels are also being increasingly imple- health information systems to ensure these
At a global level, data generated through mented, addressing some of the risk factors have an integral place.
STEPS has been used in a number of projects, already captured in STEPS. Single-issue The STEPS methodology recommends
including global reporting on risk-factor surveys supported by WHO, including those the implementation of STEPS surveys every

January 2016, Vol 106, No. 1 AJPH Riley et al. Peer Reviewed Surveillance 77
AJPH METHODS

3 to 5 years, and even this cycle is proving evaluate their policies and programs—and analysis for the Global Burden of Disease Study 2013.
Lancet. 2014;384(9945):766–781.
challenging in low-resource countries. Sup- future STEPS work should remain firmly
14. World Health Organization. WHO Global InfoBase.
port is needed to encourage countries to focused on this goal.
Available at: https://apps.who.int/infobase. Accessed
move from having conducted a baseline September 9, 2015.
CONTRIBUTORS
survey to having a regular and continuous L. Riley designed the study, with input from the other
cycle of risk-factor surveillance reflected in authors. L. Riley and R. Guthold led the writing of the
their national NCD plans of action. Similarly, first draft of the article. R. Guthold and L. Riley designed
the figure graphics. All others contributed to subsequent
other countries need encouragement to drafts.
move from subnational, piecemeal imple-
mentation of STEPS surveys to capture
HUMAN PARTICIPANT PROTECTION
national prevalence data where this would be Human participant protection is not applicable because this
preferable. article is not reporting any individual-level data.
The questionnaire items and measures
used in STEPS and the indicators reported REFERENCES
from STEPS surveys need to be periodically 1. Global Status Report on Noncommunicable Diseases
2014. Geneva, Switzerland: World Health Organization;
reviewed to adapt to the latest scientific 2014.
standards and policy needs in countries. 2. 2013–2020 Action Plan for the Global Strategy for the
Recent discussions have focused on whether Prevention and Control of Noncommunicable Diseases.
items such as objective measurement of Geneva, Switzerland: World Health Organization; 2013.
physical activity patterns using wearable 3. United Nations Sustainable Development Knowledge
Platform. Transforming our world: the 2030 Agenda for
technology could be routinely captured in Sustainable Development. New York, NY: United
population-based risk-factor surveys such as Nations Department of Economic and Social Affairs;
STEPS. Similarly, measures such as hemo- 2015. Available at: https://sustainabledevelopment.un.
org/post2015/transformingourworld. Accessed
globin A1c as an indicator for diabetes could September 10, 2015.
be included if and when more reliable, 4. The World Health Report 2002—Reducing Risks, Pro-
cost-effective, and portable measuring moting Healthy Life. Geneva, Switzerland: World Health
equipment is developed. Organization; 2002.
Efforts are needed to make the data 5. Preventing Chronic Diseases: A Vital Investment. Geneva,
Switzerland: World Health Organization; 2005.
collected from STEPS more widely available,
6. Kontis V, Mathers C, Rehm J, et al. Contribution of six
so that it is fully used for strengthening NCD
risk factors to achieving the 25·25 non-communicable
programming and policymaking. Aggregate disease mortality reduction target: a modelling study.
data are routinely entered into the WHO Lancet. 2014;384(9941):427–437.
Global InfoBase,14 but this relies on countries 7. Global Health Risks. Mortality and Burden of Disease
providing their published reports to WHO Attributable to Selected Major Risks. Geneva, Switzerland:
World Health Organization; 2009.
for this purpose. Further mechanisms are
8. Lim SS, Vos T, Flaxman AD, et al. A comparative risk
needed to make STEPS data more available assessment of burden of disease and injury attributable
and ideally accessible through a central to 67 risk factors and risk factor clusters in 21 regions,
repository or portal. Participation in such 1990–2010: a systematic analysis for the Global Burden
of Disease Study 2010 [Errata in Lancet. 2013;381(9867):
a mechanism should be agreed on at the start 628; Lancet. 2013;381(9874):1276]. Lancet. 2012;380
of a survey. (9859):2224–2260.
Finally, future efforts to strengthen the 9. World Health Organization. WHO STEPwise
linkages from data to action on NCDs at the approach to surveillance (STEPS). Available at: http://
www.who.int/chp/steps/en. Accessed August 10, 2015.
country level remain a firm priority. One
10. Armstrong T, Bonita R. Capacity building for an
method that has shown promise in this area is integrated noncommunicable disease risk factor
the implementation of pilot workshops in surveillance system in developing countries. Ethn Dis.
selected African countries to strengthen the 2003;13(suppl 2):S13–S18.
link between having the STEPS data and 11. World Health Organization. NCD Global Moni-
toring Framework. Available at: http://www.who.int/
moving to action. There is enormous value in
nmh/global_monitoring_framework/en. Accessed
linking good national NCD risk-factor data August 20, 2015.
to processes that inform policy and program 12. WHO STEPS Surveillance Manual: The WHO
making at the country level and help with STEPwise Approach to Chronic Disease Risk Factor Surveil-
target setting and monitoring of progress. This lance. Geneva, Switzerland: World Health Organization;
2008.
is after all one of the key goals of STEPS
13. Ng M, Fleming T, Robinson M, et al. Global, re-
surveillance—to strengthen the availability of gional, and national prevalence of overweight and obesity
data to help countries inform, monitor, and in children and adults during 1980–2013: a systematic

78 Surveillance Peer Reviewed Riley et al. AJPH January 2016, Vol 106, No. 1