ABSTRACT
Objectives: Adherence to supplementation provided during an intervention trial can affect study outcomes. We compared different approaches for estimating adherence to SQ-LNS and dispersible tablets in a randomized clinical trial in Burkina Faso to evaluate concordance among results and factors associated with reported non-adherence. Methods: 2453 children (9-18 mo) were randomly assigned to receive daily 20 g SQ-LNS with varying contents of zinc and a dispersible tablet (0 or 5mg zinc). During weekly home visits, reported adherence to SQ-LNS and tablets was collected through caregiver interview and disappearance rate was calculated based on unused packages. In a randomly selected subgroup (n=192), 12-h home observations were completed when children were 11 and 16 mo of age, to assess supplement consumption. Results: Average daily reported SQ-LNS and tablet adherence was 97%±6%. SQ-LNS and tablet disappearance rate also showed high weekly adherence (98%±5%). By contrast, home observation found that only 68% and 58% of children at 11 and 16-mo, respectively, received SQLNS during the observation periods (Rho=0.06, P=0.294 reported vs. observed), and fewer (36 and 28%) received a tablet at 11 and 16-mo (Rho=0.11, P=0.05). Fever, diarrhea, malaria, vomiting and loss of appetite reduced significantly reported consumption of SQ-LNS and, to a lesser extent, tablet (P<0.0001). Conclusions: Discrepancies among observed and reported results suggest possible overreporting of adherence to products and/or that consumption occurs outside the 12h home observation period. Child morbidity may change child acceptance or caregiver perceptions regarding the suitability of supplementation. Better methods are needed to assess adherence in community supplementation trials.