The MapSan trial was a major 4-year research project which aimed to evaluate the health impacts of a shared sanitation intervention delivered by WSUP in the slums of Maputo. It’s the largest ever high-rigour study of the health impacts of urban sanitation. And now the long-awaited results are out.

By Guy Norman, WSUP’s Director of Research and Evaluation

Jumping straight to the topline finding: no, there was no clear impact on health.

On surface reading this seems disappointing. But in this blog, I’ll aim to read below the surface…

First, though, a short summary for busy people:

  • Faecal-oral disease is a major cause of death, disease and poor life-outcomes among children in slums.
  • The findings of the MapSan trial strongly suggest that slum sanitation improvements are necessary but not sufficient for combating faecal-oral disease.
  • In other words, we need to continue investing in slum sanitation: it’s absolutely necessary. But we also need to understand what other types of improvement are required to combat faecal-oral disease in any given context. And we need to work together to deliver that package of improvements.

About the intervention

MapSan evaluated the health impacts of a shared sanitation intervention delivered by WSUP in Maputo (Mozambique). Between March 2015 and March 2016, WSUP constructed about 50 communal sanitation blocks (CSBs) and about 250 smaller shared toilets, all with septic tanks, in low-income compounds across various districts of Maputo. Typical before-after pictures below: as can be seen, the existing situation was pretty unpleasant (that’s an open cess-pit in the first photo)…

Before the intervention
After the intervention

About the trial

This was a major research study with total budget of about $1.3m. It was a controlled before-and-after trial: child health and other metrics were assessed immediately before the intervention, and then in both intervention compounds and control compounds at 12 and 24 months after the intervention. Intervention compounds received a new CSB or shared toilet; control compounds didn’t.

The primary outcome measure was prevalence of enteric infections among children under 5 years of age. Specifically, the researchers took stool samples from children; used molecular biology techniques to test for the presence of a range of bacterial and protozoan pathogens; and recorded “enteric infection” if one or more of those pathogens was present.

But the researchers also looked (before and after) at multiple other types of metric, including helminth parasites in stools, and various biochemical markers of bacterial and viral infection; carer-reported diarrhoea; and child height and weight. They also looked at pathogens in the immediate environment (for example, in samples taken from soil at the toilet entrance, or swabs from food preparation surfaces); and at the abundance of flies in different locations. Finally, they looked at diverse non-health outcomes, including psychosocial stress.  I told you this was a major study…

Findings

Findings were presented at the University of North Carolina Water & Health Conference in October 2019: detailed publications in scientific journals are expected over coming months.

The study did not detect an impact on the primary outcome measure: the prevalence of enteric infections did not decline as a result of the intervention. Multiple other analyses likewise indicated no effect, whether on other infection markers, on diarrhoea prevalence, or on child height and weight. [For more technical readers: this is unlikely to be attributable to inadequate statistical power, because there were no clear trends in plotted data.]

However, in sub-analyses considering only children born in an intervention compound after the intervention, significant declines were detected in the prevalence of the diarrhoea-causing bacterium Shigella and the helminth parasite Trichuris. So this indicates that the intervention did have a specific effect on two pathogens among very young children: this is an interesting and very encouraging finding.

Environmental sampling indicated some reduction in pathogen concentrations in soil at the toilet entrance. Furthermore, there was a significant reduction in the abundance of flies around toilets. The intervention therefore appears to have had some impact on faecal contamination in the immediate household environment, though not a strong impact.

So the findings provide evidence of some impacts on faecal-oral disease transmission, but no clear impact on health.

This lack of strong effect is unfortunately par for the course among recent major trials of sanitation interventions in rural contexts: see Cumming et al. (2019).

Interpreting these findings

OK, this is complex! There are number of possible explanations for the lack of strong effect:

a) The intervention wasn’t a good sanitation intervention. This seems an unlikely explanation. The researchers rigorously assessed quality of delivery (“fidelity and compliance”), and they found that delivery was excellent: the toilets were all built to good quality, with well-designed septic tanks, and people were using the toilets and in general keeping them clean. The intervention was not perfect (for example, there was probably insufficient attention to infant faeces disposal): but it was a good intervention. As further discussed below, though, insufficient coverage density may have been an issue.

b) 24 months isn’t long enough to show an effect. A longer follow-up period might certainly detect additional effects: but again it seems unlikely that this is a major explanation… if we didn’t see a clear effect within 24 months, this suggests that the intervention didn’t have strong effect and won’t have strong effect in future. [It’s worth discussing this in a bit more detail. Faecal pathogens are abundant in these locations and these populations, and some persist for years in soil. So we might certainly expect even a strong-impact intervention to take several years to reach maximum effect. But the findings of this trial, for example around pathogen levels in soil, do not give reason to think that the effects of this intervention are heading towards a much stronger eventual effect.]

c) The intervention was good but insufficient. This seems by far the most likely explanation. The research findings show that faecal pathogens are abundant in Maputo slum environments, and infections are highly prevalent among children. It seems likely that breaking faecal pathogen transmission pathways will require multiple types of parallel intervention, possibly including education around food hygiene, better infant nutrition, better water quality, better primary healthcare, and maybe things like flood control and compound paving (so that kids are playing on washable concrete surfaces, not soil). It’s also possible that the coverage density of the sanitation intervention wasn’t sufficient: this intervention improved sanitation in many of the poorest compounds in the intervention districts, but it certainly didn’t provide massive blanket improvement across those districts.

What does this trial say about shared sanitation?  

Nothing specific. It wasn’t a comparison of household (private) sanitation and shared sanitation. There is no strong reason to suppose that a household sanitation intervention of similar characteristics would have achieved a better impact. And independently of this, in Maputo as in many other cities in sub-Saharan Africa and South Asia, many people live in dwellings that are simply too small for a household toilet: so short of rehousing, shared sanitation is their only option (see Evans et al. 2017).

What does this mean for future research and future interventions?

I’m going to express a personal view here, though it’s pretty much in line with other researcher views (see Mills et al. 2018, Cumming et al. 2019). What we need to do in any given slum is understand better how different faecal pathogens are being transmitted, from release into the environment (e.g. via septic tank effluent), through the environment (e.g. along drains), to eventual exposure (e.g. a child stepping through that drain, or eating unwashed fruit). Once we understand that, we can work out what types of intervention are likely necessary to break the transmission pathways. And that’s going to require us to think beyond sanitation: because better sanitation is clearly only part of it. The results of this trial strongly suggest that better urban sanitation is necessary but not sufficient.

Closely related to this, WSUP is currently supporting research around tracking faecal pathogen pathways (in Dhaka in Bangladesh): see Mills et al. (2018) and this blog post, and look out for full findings around March 2020.

Ongoing analysis of the findings of the MapSan trial will undoubtedly shed more light on faecal pathogen pathways in Maputo. But this was not the primary focus of this study, and it’s going to need more research. At this stage it’s probably fair to say that we know that additional interventions (not just sanitation) are necessary; but we don’t yet have a clear idea of what those additional interventions are, or a clear idea of how important it is to ensure high-density sanitation coverage. And of course, we can expect variation between contexts: for example, the interventions required in a flood-prone slum may not be the same as those required in a slum not subject to flooding.

Does it matter that this intervention had no strong health impact?

Yes it matters: we want to understand how to tackle faecal-oral disease in slum contexts, and we’re evidently not there yet. It’s critically important, because faecal-oral disease has severe negative impacts on child health and life-chances.

In scientific terms, though, a negative result is not a useless result: as in the present case, it can help us to understand ways forward.

And in more practical terms: health impact is our most important aim, but improving sanitation has multiple other benefits too. For example, the MapSan researchers found that users of the new facilities reported lower stress, related to decreased disgust and increased feelings of privacy and safety (Shiras et al. 2018b). And certainly, if you look back at the before-after photos above, it’s pretty clear that if it was you or me, we’d definitely prefer the “after” version!

Equally, it’s clear that replacing those disgusting open-pit latrines is necessary for public health, even if not sufficient.

So the findings of the MapSan trial certainly DO NOT indicate that we shouldn’t be investing in slum sanitation. We definitely should! That means the Municipal Council of Maputo, and WSUP, and multiple other actors worldwide. Improving slum sanitation has multiple benefits, and is unarguably a centrally necessary component of any set of improvements to combat faecal-oral disease. We just need to get better at identifying what other stuff needs to be done alongside, and at delivering whole-district improvements.

 

Acknowledgements. The intervention studied by MapSan was delivered by WSUP, in partnership with the Municipal Council of Maputo, under funding from the Japan Social Development Fund (managed by the Water & Sanitation Programme of the World Bank).  Top-up funding to ensure that the intervention met the needs of the research was provided by the Stone Family Foundation. Special recognition is due to Vasco Parente, Sanitation Lead at WSUP Mozambique, who had primary responsibility for the excellent delivery of this intervention. The MapSan trial with 12-month follow-up was funded by USAID; additional funding to allow 24-month follow-up was provided by the Bill & Melinda Gates Foundation. The MapSan trial was delivered by a consortium led by Dr Joe Brown and Dr Oliver Cumming, comprising the London School of Hygiene & Tropical Medicine, Georgia Institute of Technology, University of North Carolina Gillings School of Global Public Health, University of Florida, and Instituto Nacionale de Saúde Mozambique.

References    

Brown J, Cumming O, Bartram J, Cairncross S, Ensink J, Holcomb D, Kolsky P, Knee J, Liang K, Liang S, Nala R, Norman G, Rheingans R, Stewart J, Zavale O, Zuin V, Schmidt WP.  2015. A controlled, before-after trial of an urban sanitation intervention to reduce enteric infections in children: research protocol for the Maputo Sanitation (MapSan) study, Mozambique.  BMJ-Open.

Brown J & Cumming O. 2019. Stool-based pathogen detection offers advantages as an outcome measure for water, sanitation, and hygiene trials. American Journal of Tropical Medicine and Hygiene [in press].

Capone D, Adriano Z, Berendes D, Cumming O, Dreibelbis R, Holcomb DA, Knee J, Ross I, & Brown J. 2019. A localized sanitation status index as a proxy for fecal contamination in urban Maputo, Mozambique. Plos One [in press].

Cumming O, Arnold BF, Ban R, Clasen T, Mills JE, Freeman MC, et al. The implications of three major new trials for the effect of water, sanitation and hygiene on childhood diarrhea and stunting: a consensus statement. BMC Med 2019;17:173.

Evans B, Hueso A, Johnston R, Norman G, Pérez E, Slaymaker T, & Trémolet S. (2017). Limited services? The role of shared sanitation in the 2030 Agenda for Sustainable Development. Journal of Water Sanitation and Hygiene for Development 7(3), 349-351.

Knee J, Sumner T, Adriano Z, Berendes D, de Bruijn E, Schmidt W-P, Nalá R, Cumming O, & Brown J. 2018. Risk factors for childhood enteric infection in urban Maputo, Mozambique: a cross-sectional study. PLoS Neglected Tropical Diseases 12(11).

Mills F, Willetts J, Petterson S, Mitchell C, Norman G. 2018. Faecal pathogen flows and their public health risks in urban environments: a proposed approach to inform sanitation planning. International Journal of Environmental Research and Public Health 15(2).

Shiras T, Cumming O, Brown J, Muneme B, Nala R, & Dreibelbis R. 2018a. Shared sanitation management and the role of social capital: findings from an urban sanitation intervention in Maputo, Mozambique. International Journal of Environmental Research and Public Health 15(10).

Shiras T, Cumming O, Brown J, Muneme B, Nala R, & Dreibelbis R. 2018b. Shared latrines in Maputo, Mozambique: exploring emotional well-being and psychosocial stress. BMC International Health and Human Rights 18(1):30.