Abstract
Observational data suggest maternal handwashing with soap prevents neonatal mortality. We tested the impact of a chlorhexidine-based waterless hand cleansing promotion on the behavior of mothers and other household members. In rural Bangladesh in 2014, we randomized consenting pregnant women to chlorhexidine provision and hand cleansing promotion or standard practices. We compared hand cleansing with chlorhexidine or handwashing with soap before baby care, among mothers and household members in the two groups, and measured chlorhexidine use in the intervention arm. Chlorhexidine was observed in the baby's sleep space in 97% of 130 intervention homes, versus soap in 59% of 128 control homes. Hand cleansing before baby care was observed 5.6 times more frequently among mothers in the intervention arm than in the controls (95% CI = 4.0-7.7). Hand cleansing was significantly more frequently observed in the intervention arm among women other than the mother (RR = 10.9) and girls (RR = 37.0). Men and boys in the intervention arm cleansed hands before 29% and 44% of baby care events, respectively, compared with 0% in the control arm. The median number of grams consumed during the neonatal period was 176 (IQR = 95-305 g), about 7.8 g/day (IQR = 4.2-13.8 g). Promotion of waterless chlorhexidine increased hand cleansing behavior among mothers and other household members. Discrepancy between observed use and measured chlorhexidine consumption suggested courtesy bias in structured observations. A waterless hand cleanser may represent one component of the multimodal strategies to prevent neonatal infections in low-resource settings.