Abstract
Latrine cleanliness increased in the intervention group compared to the control group (increase from 21 to 31 % of latrines classified as clean in intervention [N = 198] and decrease from 37 to 27 % in control [N = 91]). Improved habitual latrine cleaning lead to latrines being 3.5 times more likely to improve in observed latrine cleanliness (χ2 = 16.36, p < .001) and so did improvements in quality of latrine construction, eg households that had installed a lid were 7.39 times more likely to have a cleaner latrine (χ2 = 4.46, p < .05). Changes in psychosocial factors, namely forgetting, personal norm, satisfaction with cleanliness, explained much of the change in habitual latrine cleaning (adj. r2 = .46). Behaviour change interventions targeting psychosocial factors and quality of latrine construction seem promising to ensure clean and hygienic latrines.
Acknowledgements
First and foremost, we would like to thank Andrea Stocker for her great work organising and supervising data collection. We also thank Christoph Häfliger, Christine Oberbeck and the whole data collection team for their hard work, support and enthusiasm. Further, we are very grateful for the support and good collaboration with our partners in Burundi.