Valuing an Index of Sanitation-Related Quality of Life in Urban Mozambique: A Discrete Choice Experiment

Abstract

Objectives: A total of 1.5 billion people live without basic sanitation. A 5-attribute index of sanitation-related quality of life (SanQoL-5) has been applied in 9 countries. SanQoL-5 attributes and their levels require weighting (valuation), with the resulting index ranging from 0 to 1. To date, SanQoL-5 valuation applied simple methods such as rank sum, not robust methods such as discrete choice experiment (DCE). We aimed to value SanQoL-5 using a DCE in urban Mozambique.

Methods: We enrolled 601 adults in the cities of Maputo and Dondo, sampling women and men equally. The DCE task was a choice of which was "better" among 2 combinations of SanQoL-5 attribute levels (always, sometimes, never). Each respondent completed 10 tasks and a dominance test. After fitting a mixed logit model, we rescaled coefficients to derive the index.

Results: The highest-valued attribute was disgust ("never feel disgusted while using the toilet") at 0.25. The other attributes had similar values (ranging 0.18-0.19). People valued "sometimes" levels at approximately 60% of "never" levels. Therefore, moving from the middle level to the worst involves a larger decrement than moving from the best to the middle. Mean SanQoL-5 by toilet type followed a gradient with Sustainable Development Goal 6 categories: "open defecation" 0.30, "unimproved" 0.45, "limited" 0.60, and "at least basic" 0.70.

Conclusions: To our knowledge, this is the first DCE-based valuation of any index of sanitation-related quality of life, enabling SanQoL-5 to be used in economic evaluation. Identifying sanitation service transitions associated with the greatest quality of life gains could inform more efficient resource allocation.

Keywords: discrete choice experiment; quality of life; resource allocation; sanitation.