Refining hand washing interventions by identifying active ingredients: A cluster-randomized controlled trial in rural Zimbabwe

Rationale: Consistent hand washing with soap can reduce the risk of diarrhea, but changing hand-washing behavior is difficult. Systematic behavior change approaches promise to enhance hand washing with soap effectively, and allow the identification of active intervention ingredients using mediation analysis. This knowledge can then be used to derive hypotheses for systematically refining the intervention. Objective: We demonstrate this at the example of a behavior change intervention to promote hand washing with soap based on the RANAS approach (risk, attitudes, norms, ability, and self-regulation). Methods: Sixteen wards of Masvingo province in Zimbabwe were randomly allocated to the RANAS-based intervention or a wait-list control group. Hand washing at baseline and follow-up was observed for 224 randomly selected caregivers of young children. They additionally participated in quantitative face-to-face interviews assessing psychosocial factors. Results: At baseline, hand washing with soap was <3% on average, and did not differ between groups (p = .526). At follow-up, intervention participants washed hands with soap more frequently than controls (in 29.4% vs. 8.2% of all stool- and food-related situations, B = 1.88, SE = 0.32, OR = 6.6, p < .001). Mediation analyses revealed that the intervention enhanced several of the targeted psychosocial factors; return, descriptive and injunctive norms, action knowledge, action self-efficacy, maintenance self-efficacy, action planning, and remembering. The intervention effect was mediated through increased remembering. Conclusion: This study supports the efficacy of a systematic approach to behavior change to promote hand washing with soap. The analyses of the mechanisms revealed important insights into the active ingredients of the intervention, which will facilitate its future refinement