Audit And Feedback Strategy in Maternal And Child Health Services in A Low-Resourced Setting – A Process and Impact Evaluation of the IDEAs Strategy in Central Mozambique

Abstract

University of WashingtonAbstract Audit And Feedback Strategy in Maternal And Child Health Services in A Low-Resourced Setting – A Process and Impact Evaluation of the IDEAs Strategy in Central Mozambique Aneth Antonio Dinis Chair of the Supervisory Committee: Kenneth Sherr, PhD, MPH Department of Global Health, Epidemiology, and Industrial and Systems Engineering Audit and feedback strategies have been used to improve healthcare quality. Scarce evidence exists on audit and feedback implementation processes and their potential to improve delivery of maternal and child health services in low-resource health systems. The Integrated District Evidence to Action (IDEAs) is a multi-component audit and feedback strategy designed to improve the implementation of maternal and child guidelines in Mozambique and was implemented in 154 health facilities across 12 districts in Manica and Sofala provinces between 2016 and 2020. This dissertation aims to evaluate the implementation process of IDEAs and report implementation outcomes, evaluate the effectiveness of IDEAs on health service outcomes, and examine the association between measures of implementation fidelity and the strategy's effectiveness.In process evaluation, the Reach, Effectiveness, Adoption, Implementation, and Maintenance (RE-AIM) framework was used to guide the definition and measurement of implementation outcomes. Reach was defined as the proportion of pregnant women attending IDEAs facilities.; adoption as the proportion of facilities initiating audit and feedback meetings; implementation as fidelity to the strategy components (readiness assessments, meetings (frequency, participation, action plan development); and targeted financial support and supervision); and maintenance as intervention sustainment at 12, 24, and 54 months. We found high reach, adoption, and maintenance. Implementation fidelity was high for all strategy components with the exception of implementation of readiness assessments and action plans. The impact evaluation assessed 1) ten service delivery outcomes across antenatal, maternity, postpartum, childcare, and reproductive health services and 2) five service readiness outcomes (medicines, infrastructure, equipment, care, and staffing availability). We used propensity score matching to minimize bias and a controlled interrupted time series with a negative binomial mixed effects model for analyses of improvements in service delivery outcomes. We created composite scores for each domain of service readiness outcomes and employed a difference-in-difference analysis using an ordinal mixed effects model to assess score improvements. Significant associations were found with the monthly number of first at-risk child appointments (IRR = 1.06 [1.04, 1.07]), first PCR tests for HIV-exposed children (IRR= 1.02 [1.01, 1.03]), new contraceptive users (IRR = 0.95[0.94, 0.96]), women starting long-lasting contraceptives (IRR= 0.94 [0.93, 0.95]), availability of infrastructure (OR= 5.84 [1.32, 25.88]) and essential care (OR = 0.13 [0.03, 0.54]). Null effects were found on six of the remaining ten service delivery outcomes. To explore fidelity and its association with effectiveness, we used the conceptual framework for implementation fidelity to guide the descriptive assessment of adherence to the IDEAs strategy components. An exploratory principal component analysis was applied to develop the primary measures of fidelity. Regression modeling was used to study patterns of association between these fidelity measures and the previously described outcome measures (1) ten service delivery outcomes and 2) five service readiness outcomes). High fidelity was found in 12 out of 15 measures of fidelity. The principal component analysis identified four main components interpreted as "measure of audit and feedback structure," "measure of participation," "measure of insufficient supervision," and "measure of unsatisfactory completion of micro-interventions”. In regression modeling, service delivery measures were positively associated with participation, supervision, and degree of micro-intervention completion and negatively associated with more action plans and problems identified in meetings. For readiness outcomes, delivery of essential care was positively associated with participation and micro-intervention completion, and staff availability was negatively associated with the number of supervisions.