Background: Health risks associated with poor sanitation behaviours continue to be reported mostly from low-income countries (LICs). Reports show that various factors limit many people from accessing and using improved latrines, forcing some to opt for sharing latrines with neighbours, others practicing open defecation. Meanwhile, debate prevails on whether shared latrines should be categorised as unimproved according to WHO/UNICEF-JMP criteria. We contribute to this debate based on results from a study undertaken in three regions, Tanzania.
Materials and methods: Data were collected through observations in 1,751 households with latrines, coupled with collection of opinions from heads of such households regarding the latrine-sharing practices. Bivariate and multivariate logistic regression analyses were performed to assess associations between the outcome and possible predictor variables.
Results: Of all 1,751 latrines, 14.6% were shared. Among the shared latrines, 74.2% were found being generally clean as compared to 69.2% of the non-shared ones. Comparing the shared and non-shared latrines, the non-shared latrines were significantly less likely to be found with floors built with permanent materials (OR = 0.73, 95% CI: 0.55, 0.98); washable floors (OR = 0.69; 95% CI: 0.51, 0.93); and lockable doors (OR = 0.73; 95% CI: 0.56, 0.95). Shared latrines were less likely to have floors with faecal matter, functional handwashing facilities (HWFs), HWFs with running water, and roofs; albeit the differences in all these scenarios were not statistically significant. Respondents expressed desire for improved latrines, but also did not find it wrong to share latrines if cleanliness was maintained.
Conclusion: Having an 'improved' latrine remains important as JMP recommends, but based on our study findings, we argue that possessing a non-shared latrine neither guarantees safety to its users nor its categorisation as 'improved'. Instead, the state of the latrine, the construction technology used and the behaviours of the users may be more important.