Study Measures. The outcomes of interest were program coverage by gender and age, and other factors contributing to disparities in MDA treatment among sub- groups of the population (reasons for non-participation, drug distribution location and source of MDA information). While the specific questions within each coverage survey varied by country, all collected gender, age, drug coverage and reasons for non-participation. Other variables of interest included the source of MDA information (how people were informed of MDA) in the Burkina Faso coverage survey, and drug distribution location (location people received MDA) measured in the Uganda coverage survey. Drug coverage was measured for ALB and IVM among all ages in Burkina Faso, Malawi and Uganda approximately six months after an MDA occurred in each country and drug coverage was measured for PZQ and ZITH in Uganda approximately four months after MDA occurred. Program coverage (proportion of the eligible population that is treated) was calculated by including surveyed individuals that were treated in MDA in the numerator and including surveyed individuals who were eligible for the drug at the time of MDA in the denominator (individuals who reported ineligible for receiving the drug at the time of MDA were excluded).
Study Measures. On each well-child visit chart, goal setting was assessed by documentation of at least one weight-related diet or activity goal in each clinical chart. Communication of each evidenced-based weight-related strategy was assessed by documentation (presence versus absence) of each health message (fruit and vegetable consumption, sugar- sweetened beverages, out-of-home consumption, physical activity, screen time). Alternatively, use of the Healthy Habits Assessment Form indicated documentation on all five health messages. In every chart, documentation of each physical exam measure, including height, weight, BMI percentile, and blood pressure were also assessed. At each time point, pediatrician counseling frequency for each counseling component was calculated as: number of charts with the counseling component documented/total number of charts reviewed. For each counseling component, counseling frequency was categorized to identify the proportion of consistent counselors (≥80% charts with documentation) at each time point. Continuous practice and pediatrician characteristics were categorized at the median since no a priori cut-points have been consistently used. Practice characteristics included use of EMR, practice size (≤7 versus >7 physicians employed at the practice), and percent Medicaid (≤12% versus >12%). Pediatrician demographics included pediatrician gender, age (≤50 versus >50 years), employment status (full-time versus part-time), and prior Strong4Life training (yes versus no). Statistical analysis. Descriptive statistics (medians and inter-quartile ranges [IQR] or sample sizes and frequencies) of demographics and pediatrician practice variables were computed. For each counseling component, ten charts six-months before baseline, ten charts three-months before baseline, and fifteen charts at baseline were combined to denote before- MOC counseling frequency (n=35 charts). Histograms of counseling frequency before the MOC, at MOC completion, and six months after MOC completion were produced for each counseling component. The proportion of pediatricians counseling 80% or more of patients was calculated before the MOC, at MOC completion, and six months after the MOC.
