Strengthening Health Systems: A robust approach to doing good

This post is intended to sketch out my thoughts on taking a health systems approach to making an impact in global health, and biosecurity.

Acknowledgments: Thanks to Charlie Jeong for his comments and edits.

Summary

I advocate for strengthening health systems as a strategy to improve global health and mitigate global catastrophic biological risks (GCBRs). Drawing parallels to climate change solutions, I argue that fortifying health systems is akin to promoting clean energy, offering multiple benefits such as direct health improvements, economic efficiencies, enhanced disease surveillance, and effective platforms for health interventions. While focused programs like the GOBI-FFF strategy have shown measurable success, they often lack the sustainable and equitable infrastructure a robust health system can provide. Using examples such as malaria prevention and genomic surveillance for GCBRs, I aim to illustrate how a strong health system can amplify the impact and efficiency of specific health initiatives. Despite challenges in quantifiability and evaluation, I conclude that strengthening health systems is a crucial yet relatively neglected avenue to impact for those concerned with the near- and long-term future.

Introduction

Prevention of global catastrophic biological risks (GCBRs) and improving global health and wellbeing are top priorities for those concerned with doing good effectively. My aim is to make the argument for a focus on health systems as a means of facilitating a broad range of cost-effective interventions, and thereby serving an avenue to, and multiplier of, potential impact for those interested in both the near- and long-term future. There are likely enough effective interventions within the space of GCBR mitigation and global health and wellbeing that would benefit from a strong baseline health system for a focus on strengthening health systems to present a robust, and diversified means of doing good. To draw a parallel from the issue of climate change, Will MacAskill in his interview with 80000 hours identified clean energy as “weirdly, robustly good”, a “win-win-win-win-win”. The wins included:

  • Direct health improvements from particulate reduction

  • Improvements for climate change

  • Investments in innovation speeding up technological progress

  • Reductions in energy poverty in low income countries

  • Preserving coal stores which may be needed by future generations

I believe that a focus on strengthening health systems offers similar robustness in its potential to do good, including:

  • Direct health improvement from access to health care

  • Reduction in economic costs associated with disease and infirmity

  • Scale up of evolving surveillance for novel pathogens

  • A platform from which to launch therapeutic or vaccination campaigns

The health systems approach to improving global health and wellbeing is not novel. However, in framing this approach as a means of amplifying the impact of more traditions ‘cost-effective’ interventions, I believe it is relatively neglected. The importance of health systems to the mitigation of risks from GCBRs likewise presents a relatively neglected point of view, that adds to the expected value of a health systems approach from a long-termism perspective. For those who hold high levels of uncertainty surrounding the relative importance of the near- and long-term futures, a focus on health systems could offer an approach working toward positive impact on both fronts. As such I believe focussing on health systems strengthening is a robust means of doing good, which is currently particularly neglected amongst EAs.

Strengthening Health Systems: Historical perspective

There is a long history of conflicting approaches to improving health metrics in global health. Should a broad health systems approach be used, or should we focus the limited available resources on cost-effective interventions with measurable outcomes? The Alma Ata Declaration1 at the 1978 International Conference on Primary Health Care aimed to set the trajectory toward focus on horizontal, broad health service development as a means to closing the gap on inequalities in global health. Criticised for being too broad and without actionable goals, the focus was shifted to selective primary health care (SPHC). Guided by principles of cost-effectiveness and measurable progress the GOBI-FFF strategy (programs focussing on Growth monitoring, Oral rehydration solution, Breastfeeding, Immunisation, Family planning, Female education, Food supplementation) implemented by UNICEF in 19832 provided a rebuttal, an example of a focused, and cost-effective series of health interventions that positively impacted many across the developing world. Rates of immunisation in the developing world increased from 20% to 40%, and more children than ever had access to simple yet lifesaving interventions such as oral rehydration solution.3 Whilst undoubtedly effective in achieving its goals, the approach did little in terms of bolstering the health systems of the nations it benefitted. The nature of the interventions which were invested upon meant the approach provided little foundation from which an effective, equitably distributed health system could be built.

The debate between narrow and horizontal approaches to improving health across the globe remains active to date. Whilst the amount of resources dedicated to either such approach is near impossible to aggregate,4 giving adequate thought to both the focussed (intervention specific) and broad (systems based) approaches is necessary in any meaningful discussion on how best to allocate resources within the field.

Health Systems and Cost-Effective Interventions

A comprehensive health system, sensitive to local context, with the ability to provide high-quality, timely and affordable healthcare is the utopic goal for many in global health. The World Health Organisation’s 2007 report Everybody’s Business: Strengthening Health Systems to Improve Health Outcomes makes the case for a health systems approach to achieving the goals outlined in the Millennium Development Goals, and outlines 6 building blocks of an effective health system:5

1. Health services

2. A health workforce

3. Health information systems

4. Access to medical products, vaccines and technologies

5. Health financing

6. Leadership and governance

These points are incredibly broad, and perhaps somewhat difficult to draw specific, tractable action points from. They signal the importance of institutional strength and the reliance of a healthcare system on logistics, technology, budgeting and human resources. As such, effective health sector reform is contingent on many factors, such as political considerations and public policy, environmental conditions and economic development.6 The interplay of these factors mean monitoring progress can be difficult, and the impact of interventions can be unclear. These reasons can make taking a health systems approach less appealing for those, such as EA’s, who highly value high levels of certainty in measuring and maximising impact. I believe it would be an unfortunate unintended consequence for broad interventions, with a strong theory of change, to be underutilised because they’re more challenging to evaluate quantitatively.

It is important to note that drawing a line delineating narrow from horizontal approaches to improving global health is a false dichotomy. Narrowly focussed interventions can pave the way for a broader system to be established, and a focus on health systems can make current and future cost-effective programs more efficient to implement. It is a symbiotic relationship. Given the focus on interventions with more vertical approaches, an opportunity for greater efficiency, as well as a more robust and sustainable positive impact may be achieved by turning some focus and resources toward an approach that seeks to strengthen health systems. A diagonal approach,7 one which leverages the efficiency gained by a focus on both cost-effect and narrowly focused interventions, and the underlying systems that strengthen their implementation and impact, should be the focus.

Let us consider two examples of cost-effective and impactful interventions from top EA cause areas to further explore the idea.

Malaria Prevention

In 2021, there were approximately 247 million cases of malaria worldwide, with the disease being responsible for approximately 619 000 deaths world-wide.8 Insecticide treated bed nets (ITNs) are a low cost and effective means of preventing malaria infection in endemic regions.9 The Against Malaria Foundation (AMF) is a GiveWell top rated charity,10 providing bed nets at scale, monitoring its impact, globally saving the lives of hundreds of thousands each year, and reducing the economic impacts of malaria to society more broadly.11

Considering pragmatically what is required for carrying out the programme on the ground, the AMF has a number of necessary criteria when considering areas in which to roll out ITNs for malaria prevention,12 including:

  1. Accurate household-level data on the demand or need for nets

  2. Independent supervision at the ‘moment of net distribution’

  3. Post-distribution monitoring of net use and condition

  4. Monthly malaria case rate data after distribution

Whilst there is no specific requirements for established health care system architecture to be in place, the AMF does stipulate the following considerations in evaluating potential partners:12

· Would the National Malaria Control Programme (NMCP) agree to the four main requirements we have of a distribution?

· Are there potential distribution partners with the necessary resources and experience with whom we could collaborate?

· Is there a strong leader in charge of the distribution partner(s) with whom we feel we can work closely?

Consider two alternative scenarios, one in which the AMF ITN program is run in partnership with a local health centre offering contextually appropriate primary care (scenario A), and one in which the program is run in partnership with an organisation meeting only the minimum essential requirements (scenario B). For the narrow purposes of malaria prevention, scenario B is likely to be the more cost-effective of the two. Costs will not need to be distributed across competing priorities, and the program can optimise for data collection, net distribution, and outcome evaluation. If we consider a broader view of the overall health of the community, there are a number of factors that favour the effectiveness of scenario A, despite a likely higher cost. The efficiency of the ITN program implementation is likely to be greater given availability of basic resources such as brick and mortar infrastructure, and the ability to communicate the occurrence of the program to the local community. The uptake of the program may be more acceptable in a community where the local health centre is a trusted and integrated resource the community regularly utilises. There will be opportunity for gains to be made from educating the community about malaria. Community members may have a greater awareness of signs or symptoms of the disease and be more likely to present to the health centre and complete treatment in the future. Data collection of future cases of malaria may be more robust in a setting where the data collection and treatment point are the same physical location. Importantly, these benefits are more dispersed in time and place than those of the leaner scenario B, and as such are more difficult to quantify and integrate into a cost-effectiveness analysis. This should however not detract from its benefits, measured or not, they do exist. As such, I believe there is a strong argument to be made for a synergist approach in which the impact and sustainability of cost-effective, narrowly focussed public health interventions is supplemented and amplified by an underlying health system infrastructure.

Genomic Surveillance for Prevention of GCBRs

Emerging novel pathogens or the spread of high consequence existing pathogens pose an existential threat to humanity, and we as a species have an interest in their early detection, and in a prompt response to potential threats. These pathogens pay no attention to political lines or nation states and thus, technological strategies cannot be isolated to reference laboratories in high-income nations, they must necessarily be present globally. A wide variety of approaches exist to mitigate the potential consequences of GCBRs. Early detection of novel pathogens, broad spectrum therapeutics or vaccinations and quarantine facilities are all potentially useful and impactful means to reducing the existential risk associated with GCBRs. At each of these steps, there is a potential dependence on an underlying health infrastructure, a platform from which pathogens can be detected, therapeutics can be dispensed, and individuals can be quarantined. In a similar fashion to the argument outlined above, the ability to leverage an existing health system to implement a wide variety of mitigation approaches to GCBRs seems like a robustly good approach to doing good.

As an example, pathogen agnostic genomic surveillance systems have been suggested as a global public health priority in reducing the risk from pathogens with pandemic or epidemic potential. Pathogen agnostic systems survey biological data across a geographical region, screening for genomic patterns suggestive of an emergent or novel disease. A recent strategy report from the WHO outlines objectives required to achieve the goal “that genomic surveillance for pathogens with pandemic and epidemic potential is strengthened and scaled for quality, timely and appropriate public health actions within local to global surveillance systems.” The objectives are as follow:

  1. Improve access to tools for better geographic representation

  2. Strengthen the workforce to deliver at speed, scale and quality

  3. Enhance data sharing and utility for streamlined local to global public health decision-making and action

  4. Maximise connectivity for timely value-add in the broader surveillance architecture

  5. Maintain a readiness posture for emergencies

Objectives 1, 2, and 5 are obvious candidates for greater efficiency if an underlying healthcare architecture is available. In a setting where resources are already limited, there may be an obstructively high bar to the proper implementation of a genomic surveillance system, which could at least partially be overcome by the presence of the architecture of a health system. In LMICs without centralised sewerage systems, surveillance of biological materials may prove difficult if not carried out in the context of a health centre or dispensary, where respiratory swabs, urine, or faecal samples could be collected and analysed at the point of care. A readiness posture for an epidemic or pandemic similarly, will require human resources and likely be more effective if a locally, contextually effective health care system can be leveraged to deliver such an escalation plan.

Actioning a health systems approach

By nature, a ‘health systems’ approach is broad and as such can encompass a variety of potential areas of focus. Taking such an approach should not ignore arguments of cost-effectiveness, cost-effectiveness should remain an important means of arguing for the prioritisation of specific interventions within this approach. Utilising the 6 building blocks identified by WHO, we have a reasonable breakdown of potential focus areas that are somewhat more specific and actionable. I have included some brief thoughts on potential actionable areas for action within each of the building blocks:

Health services

In actioning a health systems approach, a primary focus may be on establishing and enhancing primary healthcare infrastructure, especially in regions with significant needs or geographical isolation. This could involve investments in the fundamental requirements for primary care delivery, such as facilities, equipment, and necessary medications (including vaccinations). Additionally, exploring innovative care delivery models, such community-based investigation and treatment for prevalent diseases (e.g. HIV, TB, antenatal syphilis or hepatitis B), can lead to more efficient and effective healthcare services. Evaluating these novel models through research can provide insights into their scalability and replicability in different contexts.

Example: A systematic review by Vaughan et. al. has suggested utilising community health worker (CHW) programs for the delivery of tuberculosis treatment and control, reproductive and child health, and malaria interventions can be more cost effective in comparison to standard care.

A health workforce

Addressing the health workforce involves ensuring accessible education and training avenues for potential health staff in areas of need, such as, for example, remote learning opportunities for aspiring community health workers. Ensuring there are not undue barriers to the training of health staff, and a focus remains on pragmatic requirements for healthcare delivery, (e.g. low-cost options or scholarships for those seeking specialists training from areas of need), is essential. Additionally, incentivization strategies for health workers, like scaled remuneration or other benefits for those serving in remote regions, can address the issue of workforce shortages in critical areas.

Example: A systematic review by Callaghan et. al. found ‘task-shifting’ interventions, the deligation of routine and repetitive tasks from doctors to non-physician clinicians, in HIV treatment and care was a cost-effective means of delivering high-quality and more cost-effective care tose living with HIV in sub-Saharan Africa.

Health information systems

Investment in digital healthcare infrastructure is an important aspect of a strong health system, emphasising efficient data management and sharing, such as centralised reporting systems for notifiable diseases. Moreover, the development and implementation of surveillance systems for novel pathogens are vital in developing health systems, particularly as pathogen-agnostic diagnostic technologies become more affordable. Ensuring the rollout of these technologies in low- and middle-income countries can significantly enhance global disease monitoring capabilities.

Example: An observational study by Secor et. al. highlighted the potential added value of scaling electronic immunisation registries, looking at data collected from a program in Tanzania, with the potential to improve vaccination rates, identify populations at risk for disease and improve care delivery more broadly.

Access to medical products, vaccines and technologies

Developing robust supply chains is essential for the distribution of key equipment, vaccinations, and medications. This can include monitoring systems for stock-outs of crucial diagnostic tests and essential medications. Investment in research and development of supply chain management can also play a significant role in ensuring consistent access to medical products in diverse healthcare settings.

Example: A mixed-methods study by Lavtepatil and Gosh explored the effectiveness of the Pradhan Mantri Bhartiya Jan Ausadhi Pariyojana program, a program seeking to increase the availability of essential medications through the local production and procurement of unbranded generic medications. Whilst work there remains room for optimisation, the approach shows some promise in reducing healthcare expenditure on drugs.

Health financing

A key aspect of health financing is ensuring individuals are protected from catastrophic costs associated with healthcare, advocating for universal health coverage systems in developing economies. Furthermore, making sure that highly cost-effective interventions are cost-free or actively incentivized can lead to more equitable and widespread access to essential healthcare services (this is already a focus area for some EA focussed charities).

Example: The introduction of a community-based health insurance scheme, the Mutuelles program, in Rwanda, where subscriptions of varying premiums are paid by households, with the remainder of the funding being pooled from various NGOs, government funding or development partners. Lu et. al. found the intervention to be an effective step toward universal health coverage.

Leadership and governance

Efficient collaboration between various sectors, such as public and private health systems or NGOs, especially in regions where multiple systems coexist, is important in developing health systems. Working towards a cohesive and integrated approach can optimise resource utilisation and improve overall health service delivery.

Example: Kenya has trialled a decentralised means of healthcare governance, aiming to enable more localised decision making processes catering to regional health needs. A systematic review by Masaba et. al. found there were improvements to the development of the Kenyan health system as a result of the policy, however, did identify a number of key challenges with its implementation.

Again, none of these strategies are particularly novel or ground-breaking in and of themselves. However, given the broad potential good they can facilitate, a greater focus on strengthening these key aspects of health systems should receive greater attention for those aiming to do good in the near- or long-term.

Limitations

Neglectedness

Some may consider that a health systems focus is already the focus on many players in the global health landscape. For instance, local government bodies are likely to already be contributing resources to the elements of health system development, rather than focussing on more specific interventions that are particularly important, tractable, and neglected. The comparative advantage of an actor in the global health and development space may therefore have a large marginal impact focussing elsewhere. An argument may also be made that health systems emerge as a result of economic development, and as such in developing nations the most effective means of improving health systems is by supporting development more broadly. In rebuttal to these points, I believe that those with a lens toward doing good effectively in both the near- and long-term, could have an impact on shaping this progress with a particular focus on a more cost-effective bent. For instance, an actor with a specific concern for the risks of GCBRs could advocate for the inclusion of pathogen monitoring systems in developing health systems.

Complexity of approach

Given the interdependence of health systems on other aspects of society, and the required sensitivity to local context, two potential issues may arise:

1) Difficulty in accurate monitoring and evaluation

2) Difficulty in rapid scale up and ability to absorb funding

These are precisely the points which more narrow and quantifiable approaches to issues in public health have their strengths. In response to these points:

1) Difficulty in monitoring and evaluation does not mean the actual impact of such projects does not exist. This is an issue of monitoring and evaluation methodology, rather than program failure.

2) The development of health systems is highly contextual and resource intensive, as mentioned above. Whilst this is true, given the broad array of potential avenues to positive impact, the benefit here still outweighs these potential costs.

Opportunity cost

Numerous cost-effective programs exist outside this space with no reliance on an underlying healthcare delivery platform. It is difficult to directly compare the utility of allocating funds toward what may be considered more certain avenues for impact, rather than a high-level systems-based approach. Systems based approaches are also likely to have their impact over longer time horizons. As such, one may prefer to allocate their resources to more ‘certain’ routes of positive impact, from a traditional expected value point of view.

Conclusions

For those concerned with doing good in the near- and long-term future, global health and wellbeing, and mitigating risks from GCBRs are key issues of concern. Historically, the discourse on progress in global health has fluctuated between vertical and horizontal approaches. Arguments for progress from a cost-effectiveness point of view tend to favour vertical approaches. I argue that by taking a traditionally horizontal, or health systems approach, one can robustly contribute to progress on global health and wellbeing, and mitigating risks from GCBRs, with a reasonable level of confidence. Whilst this approach is not without its own limitations or difficulties, and noting that the two approaches are mutually beneficial, I believe the health systems approach is a relatively neglected stance taken by those concerned with doing good effectively.

Crosspost from my blog at: thoughtfulexperiment.blog

References

1. Decleration of Alma-Ata. International Conference on Primary Health Care, Alma-Ata, USSR: World Health Organisation; 1978.

2. Cash R, Keusch G, Lamstein J. Child health and survival. The UNICEF GOBI-FFF program. Beckenham, UK: Croom Helm Ltd.; 1987.

3. UNICEF. 1946-2006 Sixty Years for Children. New York, NY USA, 2006.

4. McCoy D, Chand S, Sridhar D. Global health funding: how much, where it comes from and where it goes. Health Policy and Planning 2009; 24(6): 407-17.

5. Everybody’s business—strengthening health systems to improve health outcomes: WHO’s framework for action: World Health Organization, 2007.

6. Kim J, Porter M, Rhatigan J, et al. Scaling Up Effective Delivery Models. In: Farmer P, Kleinman A, Kim J, Basilico M, eds. Reimagining Global Health : An Introduction: University of California Press; 2013.

7. Ooms G, Van Damme W, Baker BK, Zeitz P, Schrecker T. The ‘diagonal’ approach to Global Fund financing: a cure for the broader malaise of health systems? Globalization and Health 2008; 4(1): 1-7.

8. World Malaria Report 2022: World Health Organisation, 2022.

9. Pryve J, Richardson M, Lengeler C. Insecticide‐treated nets for preventing malaria. Cochrane Database of Systematic Reviews 2018; (11).

10. GiveWell. Our Top Charities. 2023. https://​​www.givewell.org/​​charities/​​top-charities (accessed 02 Mar 2023).

11. Sachs J, Malaney P. The economic and social burden of malaria. Nature 2002; 415(6872).

12. AMF. How we make decisions. 2023. http://​​www.againstmalaria.com/​​Distribution_DecisionMaking.aspx (accessed 03 Mar 2023).