Optimize Audit and feedbaCk To Implement eVidence-based prAcTices in Primary Health carE

The investigators will use two phases of Multiphase Optimization Strategy (MOST) - preparation and optimization phases. In the preparation phase, Audit and Feedback (AnF) intervention will be prepared. First, the investigators will use scoping review to develop conceptual frameworks for AnF components. The outcome indicators and resource constraints for intervention will be identified based on the literature reviews. Second, an expert consultation meeting will be conducted to develop a set of candidate components for the AnF intervention. Around 10 relevant scholars and primary healthcare workers will be invited to rank the components that researchers conclude from the literature. The top 7 ranked components will be assessed by Best-Worst Scaling (BWS) questionnaires to finally identify 3 key components for AnF intervention.

In the optimization phase, the investigators will identify AnF intervention that will lead to the best desired results within key resource constraints in terms of effectiveness , efficiency, economy and scalability. First, the investigators will realistically and comprehensively assess the quality of care provided by primary healthcare facilities of the four Low and Middle Income countries (LMICs) using USP. Second, a 2×2×2 factorial design (RCT) will be conducted to determine how the results of quality of care can be fed back to primary healthcare workers in the four LMICs in order to optimize the impact of improving healthcare quality. To achieve this goal, the factorial trial will involve the 3 identified key AnF components at 2 levels each, for a total of 8 intervention groups (i.e. 8 different ways to conduct audit and feedback). By randomly assigning healthcare facilities to one of these 8 different ways to conduct audit and feedback, the investigators can obtain the change in the quality of care after implementing audit and feedback interventions in these facilities. Then, through statistical analysis, the investigators can estimate main and interaction effects for AnF components on improving the quality of primary health care. After that, the optimal combination of AnF components will be determined by trade off of the effects of AnF components and resource constraints in local primary healthcare implementation settings. Study details are as follows.