The World Health Survey Plus
The overwhelming and crucial importance of achieving Universal Health Coverage (UHC), and other Health Related Sustainable Development Goals (HR-SDGs) in health policy discussions within the global health community is driving a proliferation of new estimates to track progress from increasingly complex analytical models. However, considerable limitations and uncertainty about quality remain regarding the primary input data on which these modelled estimates are based. This uncertainty extends to data to measure and evaluate many of the HR-SDGs. The UN agencies, such as World Health Organization (WHO), United Nations Children’s Fund (UNICEF), The United Nations Population Fund (UNFPA) and the World Bank, institutions like the Institute for Health Metrics and Evaluation and Measure, and various donor agencies are contributing to the increasing volume and frequency of new data generation and estimates. These global health estimates are much needed and valuable tools in monitoring trends, predicting future health challenges and guiding interventions. However, while considerable advances have been made, the efforts to generate results which will help to monitor progress towards the HR-SDGs (which go beyond SDG 31) and UHC also obscure a rather critical weakness in the analytical foundations: namely, there is sparse empirical data that can be used as the basis for the sophisticated modelling required – or to inform national policymaking. Additionally, the WHS+ will collect data on the determinants of health outcomes, including social determinants and household economic status, in order to provide the appropriate explanations for the differences in the health outcomes. The WHS+ will also collect data on subjective well-being to inform the relationship between health and well-being. While many existing surveys, such as the World Bank’s Living Standards and Measurement Surveys, might collect some of this information in a piecemeal manner, the WHS+ will collect this data in an omnibus surveys to examine the relationships within the household and individuals who form the sample to provide insights that are otherwise unavailable.
The proliferation of health-related estimates is also driven in part by the growing number of organizations with their own programmatic goals, indicators and targets, including WHO’s 13th General Programme of Work (GPW13) Impact Framework. This framework includes the triple billion targets, one billion more people benefitting from universal health coverage, one billion more people better protected from health emergencies, and one billion more people enjoying better health and well-being by 2023, and 46 indicators for progress towards achieving the HR-SDGs, with a key focus on achieving UHC. These GPW13 indicators are aligned with the SDGs and UHC indicators but, nonetheless, require WHO member states to develop data reporting mechanisms. WHO recognizes that monitoring the multitude of indicators across many international agreements and national priorities will require robust data generation systems. WHO’s GPW13 therefore places a strong focus on the importance of open data systems, with emphasis on country-led and -owned data collection.
One proven method to improve population health estimates is through strengthening routine and periodic data generation efforts and national analytical capacity. WHO’s recent SCORE assessment (https://www.who.int/data/stories/score-global-report-2020–a-visual-summary) has clearly demonstrated that a large number of low- and middle-income countries still lack strong data and health information systems to monitor the impact of interventions in a timely and effective manner. Strengthening these systems takes time and sustained attention.
One element of health systems strengthening includes implementation of multi-topic health surveys that provide robust and high-quality data for monitoring health trends and determinants. Household health surveys are a reliable method of addressing many data gaps and form a critical component of national data and health information systems. For instance, 80 of the 232 SDG indicators come from household surveys, according to the Inter-Secretariat Working Group on Household Surveys.2 These include 29 of the 59 health-related SDG indicators. The option of using any additional surveys to assess progress towards SDGs should, ideally, be integrated into a nationally-led, regular and coordinated system and schedule of household health surveys.
While most countries have some form of household health surveys as part of their data and health information system, especially in low resource settings these are heavily donor dependent and not nationally owned and institutionalized.
WHO’s World Health Survey Plus (WHS+) is proposed as the next generation of an integrated multi-topic, multi-mode household health survey for all member states learning from the experience of conducting household health surveys over the last two decades.
The WHS+ protocol and implementation plan builds on experiences from the 2003 World Health Survey (WHS) and longitudinal WHO Study on global AGEing and adult health (SAGE), which have provided national and sub-national level data for a number of countries. The WHS and SAGE studies were reviewed by WHO’s ethical review committee as well as national institutional review boards. The data from these surveys still remain invaluable today in providing internationally comparable information in large datasets not available in other omnibus household health surveys. The foundations of the WHS+ will also aim to build on and work with other large survey programmes, including the World Bank’s Living Standards Measurement Study (LSMS) or Household Income and Expenditure Survey (HIES), USAID’s Demographic and Health Surveys (DHS), UNICEF’s Multiple Indicator Cluster Survey (MICS), and WHO’s World Mental Health Survey.
Aims and objectives
The aim of the WHS+ is to generate valid, reliable, comparable and timely information on a range of health and well-being outcomes and determinants of public health importance in nationally representative target populations. The results from the WHS+ will support WHO member states to track progress on country-specific objectives and towards UHC and HR-SDG targets, and WHO’s Impact Framework approved by Member States.
- Create a robust multi-functional data collection platform for generating data to monitor progress on national priorities, HR-SDGs and UHC
- Support countries to generate, analyse and use high-quality data to fill data gaps
- Develop standardized, computerized data collection strategies that enable speedier data collection and more reliable comparison across countries.
- Develop capacities in using evidence for policy
- Institutionalize a regular data generation platform in countries
- Provide tools and methods for better sequencing and coordination between studies
- Pursue improvements in integrating survey data into health information systems
- Provide methods for the inclusion of neglected topics and population groups
- Build on post-census surveys with up-to-date sampling frames
- Introduce a suite of data collection and analysis tools to use alongside National Statistics Offices and the United Nations Statistics Division
The standard sampling design for the WHS+ will be a nationally representative stratified multi-stage cluster sampling of households when traditional area-based sampling frames derived from a recent population census or large-scale household sample survey are available.
The sample size will be approximately 6000 households including a margin to support an expected non-response of up to 20 percent. A larger sample size will be considered if sub-national estimates are required.
The typical design will stratify the areas in the frame by regions that are specified as core domains of interest. For each of these regions, further area stratification may or may not take place, depending on the quality of the available frame information.
Sampling of areas defined as primary sampling units (PSUs) will be carried out using probability proportional to size (PPS). Typical size measures are numbers of households or residents from the most recent population census or some other similar source.
The number of PSUs to be selected in each stratum are obtained by dividing corresponding stratum sample sizes by the expected number of households an interviewer can cover during the time allocated for completing data collection work within a sampled PSU.
Sub-sampling of households in each sampled PSU will be carried out after updating the list of households in the PSU.
The WHS+ will explore an alternate sampling frame approach – gridded population datasets – to improve and extend representativeness of the results. Especially where up-to-date sampling frames are unavailable, this approach helps to establish robust, representative study samples.
The sampling frame is defined based on micro-level population estimates. Satellite imagery and its derived products can be used to construct a population layer, using an easy, workable method that divides the area of interest into grids. The sampling design reflects the characteristics of the random field by combining contextual stratification and is proportional to population size sampling. The parameters of the grid samples include coverage, frame, design, strata, target and sample size2.
The WHS+ team will work with in-country colleagues to determine the most appropriate way of identifying the target population and then advise on how best to identify, select and contact the sample. Through this innovative and tailored approach, the WHS+ teams will be able to collect detailed, granular information about different population groups.
The units of analysis will be the household and individual. The target population is adults, 18 years and older, residing within individual households, institutionalized population excluded. Inclusion criteria are age 18 years and older (unless countries want younger populations included and can obtain ethical approvals and appropriate procedures for the WHS+ implementation) and ability to provide informed consent.
Prospective household surveys will use interview modes that best suit the country situation. The WHS+ will provide options to collect data in a number of different ways, taking advantage of the latest technologies for face-to-face computer assisted personal interviews (CAPI), computer-assisted telephone interviews (CATI) and online surveys. Translation and back-translation protocols, as per WHO’s instrument translation guidelines, will be used to ensure a variety of main languages are available for the interviews in each country.
The type of data to be collected includes interview and biomarker data. The data collection period is open-ended, starting in 2022 for baseline surveys. A 5-year survey cycle will be pursued for repeat cross-sectional design surveys.
The household member who is most knowledgeable about the household will be invited to complete the household questionnaire as the household informant. The interviewers will be trained to identify the household member through a set of probing questions that will be clarified during the training and in the interviewer manual for the WHS+. This person will then be interviewed following informed consent. Eligible household members will be listed in the household roster and will be randomly selected by the CAPI or by the CATI programme, including approaches such as a random-digit dial, to complete the individual interview. Typically one eligible respondent will be selected to complete the individual questionnaire per household. When the module on reproductive and child health is included in the individual questionnaire, a second respondent, a woman of reproductive age, will be randomly selected to complete this module only if the initial eligible respondent is not a woman of reproductive age. The individual respondent will be an adult 18+ years in age and who gives consent to be interviewed.
When a country wants to implement a longitudinal cohort study the WHS+ design will take this into account and the sample will be designed accordingly. In countries where there has been previous implementation of a longitudinal study like SAGE attempts will be made to follow up the previous sample and account for attrition.
Standardized survey content, set of methods, interviewer training and translation protocols will be used.
The survey content will be selected in close collaboration with countries from multiple modules developed for the WHS+ survey instrument, which have been guided by HR-SDG, UHC and WHO Impact Framework indicators, as well as by the goal of filling data gaps in assessing health, determinants, and wellbeing.
A suite of survey modules that meet international standards, include reliable and valid interview questions, appropriate measurement approaches, and that incorporates the latest health and biological tests has been developed, with direct input from technical experts within and outside WHO. Short and long versions of the individual modules will also be made available depending on technical needs and expert advice. Many of the questions have been harmonized with validated instruments, building on work by groups such as the International Household Survey Network (http://www.ihsn.org/health-modules), but where specific cross-cultural elements arise, questions may need to be tested in cognitive interviews and pilot studies before the main surveys start.
Member states will generate a household questionnaire and individual questionnaire from a set of standardized modules covering a wide range of topics (see example table below). A suite of modules covering a wide array of priority health issues could be used to create the WHS+ instrument as a stand-alone survey.
Duration: December 2022 to July 2023
National-level Household survey