Impact of mobile health-enhanced supportive supervision and supply chain management on appropriate integrated community case management of malaria, diarrhoea, and pneumonia in children 2-59 months: A cluster randomised trial in Eastern Province, Zambia

BACKGROUND: Despite progress made over the past twenty years, child mortality remains high, with 5.3 million children under five years having died in 2018 globally. Pneumonia, diarrhoea, and malaria remain among the commonest causes of under-five mortality; contributing 15%, 8%, and 5% of global mortality respectively. Recent evidence shows that integrated community case management (iCCM) of pneumonia, diarrhoea, and malaria can reduce under-five mortality. However, despite growing evidence of the effectiveness of iCCM, there are implementation challenges, especially stock out of iCCM commodities and inadequate supportive supervision of community health workers (CHWs). This study aimed to address these two key challenges to successful iCCM implementation by using mobile health (mHealth) technology. METHODS: This cluster randomised controlled trial compared health centre catchment areas (clusters) where CHWs and their supervisors implemented mHealth-enhanced iCCM supportive supervision and supply chain management vs clusters implementing iCCM as per current Zambian guidelines. CHWs in intervention clusters used community DHIS2 platform on mobile phones to report on a weekly basis children with iCCM conditions and make requisitions for iCCM commodities. Their supervisors received electronic reports on disease caseloads and monthly automated supervision reminders. The supervisors on receipt of requisitions, organized the medical supplies and notified CHWs for collection. Intention-to-treat analysis on the primary outcome, the percentage of children aged 2-59 months receiving appropriate treatment for malaria, pneumonia, or diarrhoea from an iCCM trained CHW, was performed using a generalized linear model. Prevalence ratios and 95% confidence intervals comparing the prevalence of appropriate treatment in the intervention and control groups were calculated using log binomial regression with an exchangeable correlation matrix, adjusted for clustering by health facility. RESULTS: In the intervention clusters, 61.3% (98/160) of expected monthly supervision visits took place vs 52.0% (78/150) in the controls. A total of 3690 children 2-59 months old presented with malaria, diarrhoea, or pneumonia. In the intervention group, 65.9% (1,252/1,899) of children received appropriate care for iCCM conditions, compared to 63.3% (1,134/1,791) in the control group. The mHealth intervention was associated with 18.0% improvement in supportive supervision and 21.0% increase in appropriate treatment for pneumonia; these changes were not statistically significant. There was a 2-3-fold increase in the proportion of CHWs receiving supplies ordered: prevalence ratios ranged from 2.82 (confidence interval (CI) = 1.50, 5.30) to 3.01 (95% CI = 1.29, 7.00) depending on the particular commodity. CONCLUSION: This study was unable to determine whether using mHealth technology would strengthen supervision and supply chain management of iCCM commodities for community-level workers. There was no statistically significant effect of mHealth enhanced iCCM on appropriate diagnosis and treatment for children with malaria, pneumonia, and diarrhoea in rural Zambia. Longer term longitudinal studies are required to determine the impact of mHealth enhanced iCCM on health outputs and outcomes. TRIAL REGISTRATION: ClinicalTrials.gov, NCT02866097