Impact of a sanitation intervention on quality of life and mental well-being in low-income urban neighbourhoods of Maputo, Mozambique: an observational study.

OBJECTIVES: Toilet users often report valuing outcomes such as privacy and safety more highly than reduced disease, but effects of urban sanitation interventions on such outcomes have never been assessed quantitatively. In this study, we evaluate the impact of a shared sanitation intervention on quality of life (QoL) and mental well-being. DESIGN: We surveyed individuals living in intervention and control clusters of a recent non-randomised controlled trial, and used generalised linear mixed regression models to make an observational comparison of outcomes between arms. SETTING: Low-income unsewered areas of Maputo City, Mozambique. PARTICIPANTS: We interviewed 424 participants, 222 from the prior trial's intervention group and 202 from the control group. INTERVENTIONS: The control group used low-quality pit latrines. The intervention group received high-quality shared toilets, with users contributing 10%-15% of capital cost. OUTCOMES: Our primary outcome was the Sanitation-related QoL (SanQoL) index, which applies respondent-derived weights to combine perceptions of sanitation-related disgust, privacy, safety, health and shame. Secondary outcomes were the WHO-5 mental well-being index and a sanitation Visual Analogue Scale. RESULTS: The intervention group experienced a 1.6 SD gain in SanQoL compared with the control group. This adjusted SanQoL gain was 0.34 (95% CI 0.29 to 0.38) on a 0-1 scale with control mean 0.49. Effect sizes were largest for safety and privacy attributes. Intervention respondents also experienced a 0.2 SD gain in mental well-being. The adjusted gain was 6.2 (95% CI 0.3 to 12.2) on a 0-100 scale with control mean 54.4. CONCLUSIONS: QoL outcomes are highly valued by toilet users and can be improved by sanitation interventions. Such outcomes should be measured in future sanitation trials, to help identify interventions which most improve people's lives. Since SanQoL weights are derived from respondent valuation, our primary result can be used in economic evaluation.