At the African Population and Health Research Centre, where all three authors of this article work, there is a running joke of our 19-year old baby, who just eats and eats. The baby, without whom the Center would not exist, is the Nairobi Urban Health and Demographic Surveillance System (NUHDSS). The NUHDSS is a bi-annual survey conducted among 33 000 households in two of Nairobi’s slums – Viwandani and Korogocho – and collects data on the residents’ lives. From births, deaths, migration patterns, vaccinations, education levels, incomes and employment, and marital status, amongst others. The main driver behind the study is the lack of inclusion of slum residents’ in national data collection exercises. This means that for the longest time, there were people who were born, lived and died in Nairobi (and in many other informal settlements across the country), but according to official figures, they simply did not exist.
So how can a government plan for people whom it does not recognise?
The problem of accuracy of national statistics is one of many that beleaguers the use of data in planning and development. If whole swathes of the population are left out of national statistics, national priorities, policies and program plans based on the data will be wrong. Health data, most especially the routinely collected service data are often beset with data quality issues, such as missing values and errors in data entry and computation. Errors related to inaccurate entries, incompleteness and inconsistencies in data mean wrong results, wrong conclusions and wrong recommendations. Good quality data are essential for proper planning, budgeting and implementation of development activities, particularly those in the essential services sector, such as public health. In the absence of quality data, investments of limited public resources are often based on guesstimates leading to costly wastage.
There are also other issues with collecting and using health data.
Coordination of multiple sources of data: Whereas most countries in Sub-Saharan Africa (SSA) have taken major steps and made significant progress towards improving availability, access, analysis and use of health data, there still exists a need for more coordinated and collaborative efforts of all stakeholders in order to fully exploit the rich potential of their health information systems. For most countries in the region, there exist multiple sources of health data related to HIV/AIDS, TB and malaria, including health and other household surveys, census, health facility data, disease surveillance data, facility assessments, administrative data (such as human resources and financing), service availability and readiness assessment, policy data and research studies. Increasingly, many of these datasets are spatially referenced and would be valuable in informing health programming and monitoring performance. However, they remain relatively underutilised resources in many countries in SSA. Scarcity of spatially-explicit population and health datasets is often a gap in discovery. In several cases, datasets are available but need to be repurposed for new needs. Facilitating mechanisms to bridge this gap and increase discovery and usage of geospatial data by health managers is therefore important. In such a case, a platform for analytical support and triangulation of available data is needed in order to reduce fragmentation and duplication of efforts, improve the efficiency, and enhance continued learning to drive improvements to program quality, efficiency and impact.
Frequency of analysis: The basic premise of evidence-based decision-making is that health data and information lack value unless they are analysed and actually used at all levels of the health system to inform decisions. Therefore, coordinated systematic analysis and review of all available data at regular intervals is an essential part of the program performance evaluation and management cycle. This could be linked with a comprehensive or disease-specific program review, which is a more intensive and structured opportunity for reflection to identify key issues and concerns at policy and strategic levels, as well as at operational levels, and to make informed decisions for effective program implementation. Regular assessment of successes, bottlenecks and opportunities are critical to maximise efficiency. However, these are often lacking or insufficient in most low-developed countries.
Data structures: Population and health survey data often comprise of quantitative, qualitative and geospatial characteristics – all of which can even be disaggregated to different levels of measurement. These can be at cross-sectional level, hierarchical and/or longitudinal with repeated measurements for the unit of analysis. Moreover, unlike data from surveys, the routinely collected service data (RCD) or register-based data that is common in the health sector is traditionally used for reporting purposes at the national and sub-national levels other than research. They can be used to understand the effectiveness of services and to improve decision-making in the healthcare system as they have wider coverage both geographically and in terms of recorded items/parameters. However, utilisation of RCD in most African countries has been far from optimal, due to lack of analytical capacity as well as low government demand for such data. There have been calls for research considerations to be made to maximise the utility of RCD in order to benefit from the funds and efforts to collect these data.
Data quality: Although country-specific health data management systems in African countries have continued to improve, information collected still remains limited. This further hinders data analysis, dissemination and use in the sector, where use of information sources beyond routine health management information is already weak.
Data cost: In a world that is already moving onto the use of artificial intelligence and Big Data as the next frontier, low- and middle-income countries (LMICs) are still grappling with institutionalising ‘small data’. Data collection, handling, archival and analysis is still an expensive affair in terms of capacity, logistics and financial implications for most LMICs in SSA. National statistical offices do not have the necessary and sufficient technological, financial and human resource capacities to collect, process and disseminate the data required for the pursuit of the Sustainable Development Goals, Agenda 2063 and national development plans of countries, and often have to weigh donor priorities against their own national concerns, with scales tilting in favour of one end of the balance much to the detriment of the other.
African governments at all levels need to invest in relevant, timely and accurate data for decision-making at all levels. Countries that practice a devolved or federal system of governance, need to ensure that national-level health data systems are not detached from the realities of life at the grassroots. Social service provision needs to be data driven to ensure greater efficiency, by matching public policies with the needs and desires of the end users.
Data-for-development needs to be empirically sound, right from conceptualisation of the problem, through the design of the research methodology, to the data collection, handling, analysis and dissemination. Without basic metrics, it is not possible to get an accurate picture of how a population is developing or how to target health investments to meet the real needs of the people. In order to establish and maintain functional public health systems, governments would need information on trends in mortality, health of special interest groups, trends in development and implementation of public health policy and information on the general the health status of the population and the status of health services – in an open and transparent manner.
The African Population and Health Research Centre (APHRC) is committed to maximising existing country-level efforts to improve data availability, data quality and data use at the national, local and community level through coordinated investments in national data systems and advocacy for open data. For instance, all APHRC’s data since 2002 are publicly available on a Microdata Portal for external researchers and policy-makers.
Authored by Ms. Michelle Mbuthia, Dr. Damazo T. Kadengye and Dr. Caroline Kabaria from the African Population and Health Research Centre.
The views expressed in published blog posts, as well as any errors or omissions, are the sole responsibility of the author/s and do not represent the views of the Africa Evidence Network, its secretariat, advisory or reference groups, or its funders; nor does it imply endorsement by the afore-mentioned parties.