UNICEF’s seven step approach to health systems strengthening

SPEAKER: In 2016, UNICEF introduced its approach for health systems strengthening. A core part of this approach is an evidence based equity focused process for conducting a situation analysis and identifying priority actions for health systems strengthening. This process has seven distinct steps. There are many health system frameworks. WHO uses the building blocks. The World Bank model is based on control knobs. Each has their uses and focus. Many health system frameworks begin with thinking about the inputs, the health workers, or the suppliers, or the financing. Within the UNICEF approach, there’s a critical difference. The assumption is that there’s a health system in place, but that some people are missing out. And coverage of known lifesaving interventions is inequitable.

It’s likely to be serving some parts of the population better than others. And within the UNICEF approach, the starting point for any health system strengthening is identifying who is missing out on essential health interventions, why those gaps exist, and what strategies, at what cost, could be put in place to address the gaps and increase the intervention coverage. Imagine you are Amal and have been given the task of advising the Minister of Health on what would be the best way to improve newborn survival in country X.

Step one: So the first step is to identify the underserved populations. So how do we do that? The most robust way is to undertake an equity analysis. An equity analysis involves identifying the potential socioeconomic factors that result in some sub-populations not receiving the same access to health services as other sub-populations. So for your assignment in country X, you need to find out who is missing out on newborn care services. There are some very well-known equity markers, gender or wealth, ethnicity and geographical location.

We can analyse these using regular surveys, such as the Demographic and Health Surveys, and measure whether urban residents or richer families have better health service access and better health outcomes than their poorer rural counterparts. Inequities can take many forms, though. There can be multiple inequities in one particular setting. And we need to tease out and understand which are the important factors that lead to inequity. We often have to look harder for inequities that are not so easy to measure. A commonly overlooked equity marker is disability. We know that many people living with a disability do not have the same access to health services.

And yet, because our routine surveys do not measure the numbers of people living with a disability, we often overlook this important equity marker.

Step two, the second step is to ensure that your health planning focuses on the burden of disease. You need to know what the main causes of mortality and morbidity are for the identified population in your particular context. Without knowing what the current leading causes of death and disability are we can easily miss opportunities to address emerging illnesses. For many years, we spoke of non-communicable diseases, such as diabetes and heart disease, as diseases of the rich. We didn’t look for it or document the rise of non-communicable disease in low income settings. Yet we now know that non-communicable diseases are as important in low and middle income countries as they are in high income countries.

Because of our focus on infectious diseases, we missed an important cause of death and disability. Remember, we manage what we measure. So we need to ensure we accurately identify the major causes of death and disability. We also risk repeating what we’ve done in the past. Take the example of Bangladesh. When people are asked to consider causes of under-five deaths in Bangladesh, they think of diarrhoea and pneumonia. Think again. The causes of death in children under five years of age have changed. 29% of all deaths between age one and four-year-old children are due to injury and the vast majority by drowning, some 83 deaths per 100,000 children per year.

While we still need to fund programs for diseases like pneumonia and diarrhoea, we need to measure and understand the importance of injury deaths, so that the country is prepared to also fund injury prevention programs or treatment programs.

In fact, the Bangladesh Health and Injury Survey, conducted in 2003, demonstrated that more children aged one to four died from drowning than from diarrhoea. Examining the data and not relying on our own preconceived ideas is critical. Step two is to know your causes of death and disability and recognise that these causes change over time.

Step three: Once you’ve determined who’s missing out, and you know the major causes of mortality and morbidity in this target group, the next step is to identify the evidence based interventions that will address those causes in the target population. This is the easy step because we have a growing body of evidence that tells us what health interventions can address specific causes of mortality. And we can often measure the number of lives that can be saved with increasing coverage of that health intervention. In the case of newborn mortality, for example, we know that clean deliveries, skilled birth attendants, neonatal resuscitation, early skin to skin contact, and breastfeeding all improve newborn survival.

But knowing what works is not enough, which brings us to step four. The real system challenges come in working out how to overcome the bottlenecks to high intervention coverage. A bottleneck is a blockage. We could think of it as a traffic jam or a blockage, that thing that stops essential services getting to those who need it. There are many potential bottlenecks in a health system, which reduce effectiveness at different points.

These health system or implementation level bottlenecks often fall into categories of supply, having sufficient staff and supplies in the right place, utilisation or the demand for services, and quality of care. We need to identify what the bottlenecks are and what the underlying determinants of those gaps are. So for example, thinking back to Amal’s challenge to improve newborn outcomes, we know that providing skilled personnel at a delivery who can resuscitate a sick newborn is an evidence based intervention. The bottlenecks might vary, though. They might range from not having a sufficiently trained staff, to staff not being deployed in the birthing centres or insufficient numbers, so 24-hour care is not possible.

The bottlenecks are those barriers to effective coverage of a known lifesaving intervention. In a complex system, these barriers are likely to be many and multifaceted and may be difficult to identify and understand.

Step five: The other challenge is that you might have multiple ways to address the bottleneck. And so you’ve got to work out what’s the best approach. We have, for example, evidence of the effectiveness of postnatal care in early diagnosis and treatment of a sick newborn. But we still really struggle to know how to implement that intervention. Do we keep women and their newborns in a facility for longer, so we can monitor them and check the babies well before they go home? Or do we provide home-based care? If we provide home-based care, who provides that care? Who supervises the person who provides that care?

And in every different context, the answer of what is the best strategy to overcome the bottleneck will be very different. In fact, because these bottlenecks occur in a complex health system, we may need multiple strategies across many parts of the health system to address a bottleneck.

Step six: To make good choices, we need to estimate the costs involved for the different ways we could overcome the bottlenecks. Do we train more midwives to do outreach? Or do we train our community health workers to do early post-natal visits and refer sick children? Two strategies, both potentially effective, but differing in the costs involved in training, supervising and evaluating each strategy to overcome the bottleneck. And as you well know, you don’t get funded without a clear and costed plan. The next step in the framework addresses this by including the cost estimates of any strategies, designed to reduce system bottlenecks and to improve the coverage of lifesaving interventions.

Step seven, this brings us to the final step, which is to ensure that you track to see whether what you said would overcome a particular bottleneck actually did. There’s not much evidence for what works in implementing strategies to improve coverage of interventions. So this step’s really critical.

And even if a strategy fails to provide the improved coverage anticipated, that in itself is a useful piece of information that could inform others who are also making their plans to overcome similar bottlenecks. For those interested, this is the heart of implementation science or implementation research, providing the evidence of the how and why change occurred. Monitoring and evaluation are critical components of a strong health system. Finally, so that brings us back to the beginning. And we might restart this whole process again for a new planning cycle, might be the following year or the following five year plan. And the steps begin again.