A major objective of many health systems worldwide is to ensure that all people obtain the high-quality health services they need without risk of financial hardship. Achieving this universal health coverage (UHC) requires reducing inequalities within countries as well as between countries. Within-country inequalities, particularly in low- and middle-income countries, has therefore become the focus of research. In a recent study published in BMC Medicine, Emmanuela Gakidou from the University of Washington, USA, and colleagues focus on maternal, neonatal, and child health (MNCH) interventions across 72 districts in Zambia. Using data collected over 20 years, they produce the first systematic assessment of trends in this area. Here Gakidou explains their findings and what the implications are, not only for Zambia but also for other sub-Saharan African countries.
What was the main aim of your research and why did you decide to investigate this?
I’ve always had a strong interest in measuring health inequalities and evaluating the impact of health programs (The Lancet. 2006, 368, 1920-1935). I deeply care about and recognize the need to assess local health trends and know how powerful subnational benchmarking can be to catalyze action and improve health.
Zambia has been recognized worldwide for its successes in improving childhood survival and tackling many deadly diseases, including malaria and HIV/AIDS. But little was known about whether such progress was evident throughout the country, or if these gains were largely driven by the country’s most well-off.
Over the last few years, the Institute for Health Metrics and Evaluation (IHME) and collaborators at the University of Zambia brought together all of the pre-existing data sources we could locate and carefully standardized them to produce a large database of district-level trends for a full range of health interventions between 1990 and 2010. With these locally focused data we created a resource for tracking health gains and benchmarking components of overall health system performance.
Our knowledge of how well health programs or policies are working can only be as deep as our level of analysis. If we limit it to the surface level – national trends in this case – our insights are likely to be similarly limited. But if we are able to dive deeper, to examine health trends at the district, state or county level, the results are likely to be much more actionable and policy-relevant.
Your study showed relatively even improvements in certain interventions (e.g., malaria control, exclusive breastfeeding), whilst other routine services (e.g., polio vaccination, antenatal care) were variable across districts. What do you think are the reasons for this?
The program structure behind providing these interventions or services is likely a main factor. Malaria control interventions were largely funded and delivered within a more vertical program orientation, particularly during the earlier years of scaling up malaria control. For the most part, households were provided malaria control interventions without additional efforts or resources to acquire them. Insecticide-treated nets (ITNs) were distributed through mass campaigns, and indoor residual spraying (IRS) was conducted house-by-house each spraying season.
More routine services, such as polio vaccination and antenatal care (ANC), require multiple contacts with the health system (i.e., three to receive full polio immunization and four ANC visits to comply with Zambia’s clinical guidelines). If you live far away from a health facility, when and how often you seek care can easily be affected. Other factors, including child care needs, travel expenses, and patient views of health facilities, also directly influence health-care-seeking behaviors.
In comparison with disease-specific programs like malaria control, the number and types of barriers to care facing routine services are both higher and varied across districts.
Socioeconomic status was not found to strongly correlate with performance in various interventions. Were you surprised by this finding? What other factors do you think are at work?
Initially we had expected to see a clearer relationship between higher levels of intervention coverage and socioeconomic status. But once we started looking at which interventions were rapidly scaled up between 2000 and 2010 (and which ones weren’t), it started to make more sense. These interventions – malaria control, exclusive breastfeeding, and the pentavalent vaccine – are purposely targeted to rural, less wealthy populations as they often bear the largest burdens of malaria, malnutrition, and vaccine-preventable diseases. In some ways, this weak correlation between socioeconomic and intervention coverage may reflect the country’s success in providing a subset of health services to the people who needed them the most.
We don’t yet know for sure whether this finding is unique to Zambia. The structure and implementation of Zambia’s health programs varies substantially from many of the other places in sub-Saharan Africa where we work, including Uganda and Kenya. We are finishing up subnational analyses for Uganda and Nigeria, which will help answer this question.
How can the disparities in health outcomes across districts in Zambia be overcome?
An important first step was conducting these district-level analyses and making them publicly available. This kind of information can be used to galvanize greater accountability from local health officials and political leaders. We’ve seen how this kind of data generation and exchange can spur action at local levels in Mexico (The Lancet. 2006), a country with a history of disparities in wealth and health.
Policymakers and development partners can also use these data to identify target areas for additional attention and investment. Specifically, our partners from the University of Zambia have had the opportunity to present and discuss the findings of this project at the biannual meetings of the District Health Officers in Zambia and have continued discussions with key policymakers at the Ministry of Health to help formulate policies for districts that are lagging behind and interventions that require particular attention.
What are the implications for both Zambia and other sub-Saharan African countries?
This work has three main implications for Zambia and other sub-Saharan African countries.
Firstly, the rapid and nearly universal scale-up of several malaria control interventions has likely contributed to Zambia’s gains against malaria to date. As Zambia focuses on eliminating malaria district by district, new strategies and interventions may need to be evaluated for their effectiveness across different settings. What has worked to bring down malaria in Namibia (as published in a study in BMC Public Health), a malaria-eliminating country bordering Zambia, may not work as well as in Zambia or other countries transitioning from a focus on controlling malaria to one directed toward eliminating the disease. Thus, it will be critical for Zambia and other countries considered malaria control success stories (e.g., Rwanda, Kenya, Ethiopia) to determine which intervention packages and programmatic approaches will optimize their transition toward reaching zero malaria.
Second, like many countries in sub-Saharan Africa, Zambia has not seen the same kind of progress in reducing maternal mortality (The Lancet. 2014, 384, 980-1004) as it has for improving childhood survival (The Lancet. 2014, 384, 957-979). It’s become clear that across districts in Zambia the gains documented for interventions that directly affect under-five mortality (including malaria control interventions, immunization coverage) have not been achieved for interventions that promote women’s health. How or why exactly this kind of imbalanced progress has occurred warrants further investigation, especially if Zambia and other sub-Saharan African countries are to support longer, healthier lives for women.
Finally, our findings reflect the critical need for routinely assessing health trends at local levels in a timely, comprehensive way. As Zambia rolls out its health insurance scheme and strives to achieve equity goals established through the Vision 2030 development program, being able to promptly respond to local health needs will become increasingly crucial. Other countries throughout sub-Saharan Africa are implementing similar initiatives and policies and thus could equally benefit from such subnational benchmarking exercises.
What further research is needed in this area?
I would like to see this kind of subnational benchmarking analysis updated at least every few years. As we have seen with Zambia, a lot can happen within a few years: huge gains in access to key interventions such as malaria control, or quick declines in intervention coverage – as several districts experienced with polio immunization. To properly respond to an emerging health need, we need this kind of comprehensive analysis to occur with much greater frequency.
We also need to better understand why or how certain trends in health service provision are occurring. With this work, we are getting a much better evidence base for changes within Zambia’s health system over the last few decades. Nonetheless, we remain a few steps away from fully understanding the impact of the country’s scale-up of different interventions and which programs are driving the largest health gains. We need to conduct more impact evaluations, ideally in a prospective manner (or when new programs are implemented), to identify the best evidence-based policy options in global health.
Questions by Lin Lee, Deputy Editor for BMC Medicine.
The post Assessing trends in healthcare interventions across Zambia appeared first on Biome.