Checklist Based Box System Interventions

Overview

Maternal mortality is still high in Ethiopia. Antenatal care (ANC), use of skilled delivery attendants and postnatal care (PNC) services are key maternal health care services that can significantly reduce maternal mortality. However, interventions applied to the continued utilization of these key maternal heath services in a continuum of care approach (i.e. early initiation of ANC and continued utilization up to four plus vists, health facility delivery attended by skilled health care providers and attending three PNC visits) were not well applied and studied. Hence, the purpose of this study is to test the effectiveness of checklist based box system interventions on improving utilization of maternal health service (Antenatal care, skilled birth attendance and postnatal care) utilization.

Cluster Randomized controlled trial study design will be employed. The sample size for this study was calculated based on the recommendations for sample size calculations for cluster randomized controlled trials with fixed number of clusters, by using STATA. The following assumptions were considered: to detect an increase of postnatal care three utilization from 16% to 28% from previous study, number of clusters available-30, with 95% confidence interval and 80% power, intra-cluster correlation coefficient of 0.04849 from similar studies, 15 clusters per arm. The sample size was calculated to determine number of observations required per cluster, for two-sample comparison of proportions (using normal approximation), Assuming individual randomization, sample size per arm is 194. Then allowing for cluster randomization, average cluster size required is 40, and the final sample size is 1200 pregnant mothers (600 in intervention, and 600 in control). Data analysis will take place in two levels (cluster and individual). Risk ration will be computed at cluster level, and the results of this cluster summary will be compared using t-test. Primary and secondary outcomes will be compared between intervention and control groups with random effects logistic regression models, taking account of clustering.