I’ve written about health systems in a few other places recently: a post on Humanosphere talks about health systems in general, and a post on Reboot’s blog looks at lessons from governance reform. Both posts were sparked by a recent event hosted by the Center for Global Development, Population Services International, and others.

One of the big takeaways from the event was the need for the global health sector to renew its focus on health systems strengthening. Experts from several organizations and donors acknowledged the successes of disease-specific approaches in recent years, but highlighted how enabling environments and system-level interventions hold promise for improving health outcomes across multiple conditions. Thinking about systems is important in global health and, more broadly, international development.

However, there are several distinctions to be made as we zoom out from health to the systems surrounding it. At the risk of being pedantic, I want to slice up our thinking into four levels.

Level 1: Thinking about healthcare

We see healthcare happen most clearly in the interaction of a medical professional and a patient. This is how we most directly experience healthcare. Depending on the context, this professional might be a doctor, pharmacist, or community health worker. The interaction might be focused on routine care or serious conditions. Regardless, this clear picture of healthcare in action is also the most obvious place to target improvements: put better medicines in the hands of the professionals, equip them with new technologies, or improve their skills and knowledge.

Health interaction

Of course, we know that this an incomplete picture. Doctors don’t just appear out of nowhere, ready to heal. Patients don’t magically find themselves in hospitals. To see how we got to this interaction, let’s zoom out one level.

Level 2: Thinking about healthcare systems

There are hundreds of other people and dozens of institutions working to make that interaction possible. They constitute the healthcare system. The World Health Organization identifies six system building-blocks that are needed to achieve health goals: service delivery, health workforce, health information systems, access to essential medicines, financing, and leadership/governance.

That makes our picture of healthcare provision look something more like this:

Health system elements

Each of these system elements feeds into or somehow sets the stage for the healthcare that takes place at the center. Therefore, each of these elements also provides opportunities to improve healthcare. When we talk about systems strengthening, we’re broadening our thinking to cover all of these elements. Beyond the medical research and development that will produce new vaccines or treatments, the sector is increasingly looking for innovations in areas like supply chains (such as how the informed push model is reducing clinic stock-outs in Senegal) or financing (such as Changamka’s outpatient smart cards in Kenya).

How we actually assess system-level improvements is still a challenge: we need better measurement of these system elements in order to target the optimal leverage points, and the best interventions for a given system will remain highly context-dependent and difficult to design.

Level 3: Systems thinking about healthcare systems

There’s a further level of complexity that only gets surfaced through systems thinking. This requires that we go beyond merely seeing the impact of system elements on healthcare, and also see the complex impacts that system elements have on other system elements.

Building on the above, we can see a few simple links:

Systems thinking for health systems

Each of these gets more complex. Systems thinking is highly concerned with flows/stocks, feedback loops, sustainability, resilience, and hierarchy of organization. We need to keep in mind one of the key lessons of systems thinking: elements often interact in ways that are unexpected, and even counterintuitive.

For example, the influence of policy on the availability of medical professionals goes through multiple channels: education funding, compensation of public sector employees, regulations allowing for private sector healthcare, emigration opportunities for skilled professionals, and more. The lag time needed to train a workforce can result in oscillations, as is often the case in systems: adjustments in supply overshoot the need, leading to a glut, followed by a reduction in supply, leading to scarcity.

Similarly, patient access depends on the existing transportation infrastructure and networks. It can be improved through a combination of capital investment (either in roads or new health facilities), financing, or information systems (i.e. telemedicine). Adjustments to each play out on different timelines, complicating our assessments of their effectiveness and efficiency.

The interactions of these elements are unique to context. Other elements might also be relevant in some contexts, including cultural and linguistic diversity, public financing models, politics, or technology literacy.

Applying systems thinking to healthcare systems has the potential to identify greater opportunities than are generally fostered by a linear, value-for-money mindset. However, it requires a structured method to make sense of this complexity in any particular context. This requires a deeper level of engagement that draws on the diverse knowledge of a wide range of stakeholders. (For more, see the WHO report on systems thinking for health systems.)

Level 4: Systems thinking about health

There’s one final level as we zoom out from healthcare: systems thinking about health itself. The healthcare provision system is only one element of a much larger system that influences the health of an individual.

Our health outcomes are also shaped by other elements:

Systems thinking about health


Each of these — water and sanitation, environmental conditions, food security, security and conflict, and more — have influence over health outcomes that are at least as significant as the actual healthcare system. From an “ounce of prevention is worth a pound of cure” perspective, these other systems provide greater opportunities to improve health, as they’re often the source of negative health outcomes.

Level 5 and beyond? Shall we keep going?

This is as far out as I want to zoom. Systems thinking requires that you set boundary conditions at some point. At one extreme, there’s a risk that a narrow and mono-disciplinary analysis misses important connections; at the other extreme, there’s a systems thinking approach that gets so multi-disciplinary that it can’t decide what to leave out, becomes unwieldy, and faces analysis paralysis. Striking the balance is delicate.

Practically, the implication for health systems strengthening is that we have to think in levels. There are trees and forests and everything in between. To recap:

  • Level 1: Thinking about healthcare
  • Level 2: Thinking about healthcare systems
  • Level 3: Systems thinking about healthcare systems
  • Level 4: Systems thinking about health

Generally, the higher-level analysis will point the way to the critical intervention points at the lower levels, but only the lower-level analysis will let us see what’s working and what’s not. For example, systems thinking about healthcare systems (level 3) will help us see the critical elements in the healthcare system in question (level 2), and even identify the likely knock-on effects of adjustments. Yet the intervention is still done at a particular element of the system.

Similarly, systems thinking about health (level 4) might suggest that poor health outcomes spring from other surrounding systems (e.g. food insecurity), in which case the healthcare system (levels 2 & 3) offers only marginal opportunities for improving health outcomes. Often this analysis is done intuitively and automatically — i.e. we hardly need systems thinking to know when conflict or famine are the greatest threats to health outcomes.

However, we increasingly need better analytical tools and frameworks for the situations where the interactions between system elements are more nuanced. That starts with distinguishing between levels of thinking. After all, thinking about systems is not the same as systems thinking.


Further resources:

  1. Dear Dave,

    I think we should first begin to change the language around “health”. it might just be semantic to most people but I think the first portion of your blog / comments focuses on a Ill-care system — A system that is structured to receive people that “ill”.

    The last part of your blogs….really speaks to “health care”. Outside of conditions caused by our gene pool (the bad one), 90% of ill care as nothing to do with what we know as “healthcare”. Once we start changing that language maybe people will pay attention to government policies and actions.

    I think people would be more conscious if the government would come out and say “We have invested $ xxxx billions of dollars in Ill-care and a few hundred thousands in healthcare. Then we may get to the “thinking” part.


  2. […] great, straightforward guide to systems from “Find What Works” (by Dave Algoso), that explains pretty well the different […]


  3. […] great, straightforward guide to systems from “Find What Works” (by Dave Algoso), that explains pretty well the different layers […]


  4. One thing I’ve noticed in international health is that local public health institutions buy into small vertically funded interventions more readily if they tie in directly to core public health activities of the local authorities like rational drug use surveillance and interventions, rather than merely having a line item for health systems strengthening included in the budget. Creative thinking about placebo medicine is needed in settings where the out of pocket cost of basic health services is so high relative to ordinary income levels that placebo administration is a medical ethics issue of note where diagnostic resources are scarce and cost extra. Use of antimalarials and vitamin B injections and injectable aspirin for non-specific symptoms in outpatient medicine is pervasive, and including an antibiotic cocktail in such cases also increases the risk of community acquired MDR infections. Skills training in patient communication is a core service need that is emphasized by local health authorities and clinical staff at all levels – placebo medicine is often a matter of reluctance to explain to a patient from a background of low educational attainment that no medical intervention is warranted, for fear that patient will opt out of biomedical services in the future if they leave the consultation empty handed. These kinds of core competency gaps don’t sound exotic enough to inspire a vertical intervention in international health, but they represent key stumbling blocks in affordable health care provision for many developing countries.


Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: