Evolution of coverage assessments: the next generation loading…

In 2024, under the USAID funded ELEVATE Nutrition activity, an extensive technical review was conducted to understand the reasons behind this setback. This initiative was supposed to culminate in the development of a new methodology and/or improvement of existing methodologies, but the project came to an unexpected halt due to funding cuts. Nevertheless, the review remains a valuable resource as it looked not only at the tools used in nutrition programmes, but also at how other health sectors, including vaccination, reproductive health, and primary care, measure coverage.

The main challenges with existing coverage monitoring methods

Coverage monitoring methods in nutrition grew out of a very particular context: early community-based management of acute malnutrition (CMAM) programmes operating in small geographical areas. Methods such as CSAS, SQUEAC, and SLEAC were designed to help programmes diagnose barriers to access and improve outreach. Over time, these tools became the default instruments for estimating coverage.

But the world has changed:

• wasting treatment is now increasingly integrated into national health systems
• governments – not NGOs – are expected to monitor and finance service quality
• donors request comparable estimates across regions and over time
• programme implementers need more than a one-off estimate – they need trend data for early warning, planning, and accountability

Current tools were not designed for these needs – and are facing numerous barriers.

1. They struggle in low prevalence settings

Most nutrition programmes operate in contexts where severe wasting is relatively rare at any given moment. This makes it extremely difficult to reach the sample sizes needed for reliable estimates. Tools that rely on identifying all, or nearly all, cases are not suitable in all contexts. They can be particularly demanding in urban or mobile populations, and therefore more logistically demanding.

2. Very few tools are scalable for governments

Methods like SQUEAC require highly skilled assessment leads, Bayesian analysis, and relatively long data collection periods. Even simpler methods require logistical support and levels of technical supervision that most health systems cannot sustain. As a result, coverage assessments are conducted sporadically, often driven by donor demand, not government planning needs.

3. Cross-sectional surveys cannot track change

Nutrition programmes need to see how coverage fluctuates seasonally, in response to shocks, or as a result of programme improvements. New methodologies being built on SMART surveys only provide a snapshot in time. They do not allow real-time or near-real-time monitoring. This makes it impossible to course-correct early or assess the impact of adaptations.

4. Indicators and estimators are not harmonised

Organisations conducting coverage assessments report the key finding – the coverage estimate – using point, period or single coverage estimator. These are not comparable, leading to confusion, contradictory results, and difficulties in global reporting. At the same time, some estimates are being done indirectly, using admission data and global burden estimates as denominators. Yet, these remain uncertain, because the average duration of untreated SAM is still poorly understood – leading to over estimation of coverage.

5. Weak integration with other health interventions

Nutrition has tended to operate in isolation. Other health sectors — particularly immunisation — have made major strides in sampling approaches, digital tools, and methods for identifying missed populations. Because there is little cross-sector learning, nutrition programmes continue to use methods that have not evolved alongside wider public health practice.

6. Little attention to gender and equity

Most tools can quantify coverage, but they cannot explain why children are not reached. Qualitative components in some methodologies address this, but inconsistently. Barriers linked to gender norms, social relations, mobility, or insecurity remain poorly captured.

7. Data quality systems remain weak

Administrative systems such as DHIS2 are improving, but nutrition data is still not standardized, routinely checked, or complete. Without good routine data, it becomes difficult to triangulate survey results, build models, or link facility and household information to estimate “effective coverage.”

What a new method must deliver

The review is clear: improving or simplifying existing methods will not be enough. We need to re-define coverage measurement objectives and then build a method to meet those objectives. Based on the available evidence, a new methodology should prioritise.

1. Consensus on what coverage should measure

Before designing tools, we must agree:

• Should the output be an estimate, a classification, or both?
• Should it include quality (“effective coverage”), not just availability (“contact coverage”)?
• Should it track trends or purely measure levels?
• Should it be nationally scalable and affordable?

Without this, the sector will continue to generate incompatible and inconsistent results.

2. Adaptability across contexts

A new method must work in: 

• low- and high-prevalence contexts
• rural, urban, and displaced settings
• secure and insecure areas

3. Use of mixed data sources

Instead of relying only on surveys, the method should combine: 

• routine administrative data
• simplified household sampling
• geospatial data
• community-generated information

4. Built-in qualitative and gender analysis

Understanding why children are not reached is as important as how many are reached. Qualitative tools – particularly those capturing gender and social barriers — must be integrated, not optional.

5. Advances in digital tools and geospatial methods

There is strong potential to integrate: 

• GIS-based sampling
• mobile data collection
• satellite imagery
• machine learning to optimise sampling points

These technologies can reduce cost, improve precision, and increase feasibility.

6. Alignment with WHO recommendations

Any new approach must reflect the 2023 WHO guideline on wasting, especially around: 

• admission criteria
• protocol flexibility
• the need to monitor co-coverage of prevention and treatment interventions

7. Government ownership

The new method must: 

• integrate with DHIS2/NHIS
• be open-source
• be easy to train
• require no specialised statistical expertise

If governments cannot deploy the methodology themselves, it will not scale.

Conclusion

Coverage measurement in nutrition is at a crossroads. Current tools, while still valuable for research or evaluation of impact, are no longer adequate for the needs of modern, integrated health systems. They are costly, complex, difficult to repeat, and often do not produce the information needed for programme adaptation or policy planning.

We now have a clear understanding of what needs to change. The next step is to use this evidence to design and test a new coverage methodology – one that is simple, reliable, scalable and grounded in the realities of today’s health systems. 

Ultimately, this is not only a technical exercise. Better coverage measurement means better programme design, more equitable access, and more children reached with life-saving care. It is time for our tools to catch up with our ambitions.