The ability of primary healthcare (PHC) providers to practice in accordance with evidence-based guidelines and norms is a critical component of improving the quality of primary healthcare. Implementation science seeks to promote the routine use of evidence-based practices by identifying barriers to their implementation and developing strategies, such as audit and feedback (AnF), to overcome them. However, because the effects of AnF show significant heterogeneity across studies, this research focuses on systematically developing an optimized AnF strategy and rigorously evaluating its effectiveness in improving clinical practice compared to no intervention. The development of an optimized AnF strategy involves a preparation phase, which utilizes expert consultation and a Best-Worst Scaling (BWS) survey to identify key candidate components and assess resource constraints, followed by an optimization phase utilizing a 2×2×2×2 factorial design randomized controlled trial (RCT) to determine the most effective combination of AnF components. Subsequently, in the evaluation phase, a two-arm, multicentre RCT will be conducted across four nations (Nepal, Mozambique, Tanzania, and China). Primary healthcare providers (PHPs) will be 1:1 randomly assigned to either the optimized AnF intervention group or a no intervention control group based on randomly permuted blocks (sizes 2, 4, and 6), stratified by country. Care quality will be assessed using the gold standard method of Unannounced Standardized Patients (USPs). The primary outcome is the proportion of completed guideline-recommended quality checklist items for the consultation of hypertension and Type II diabetes cases among all items. This outcome will be expressed as a continuous score ranging from 0% to 100%. Furthermore, a mixed-methods research strategy will be employed to extract Context-Mechanism-Outcome (CMO) elements and construct a Causal Pathway Diagram (CPD). This study will provide a robust empirical foundation for using an optimized AnF strategy to improve the quality of primary healthcare in developing countries. By deconstructing "why, for whom, and under what circumstances" the intervention works through the CMO framework and CPD, this study will provide vital mechanistic evidence for the future scale-up of this model, contributing a comprehensive and universal research paradigm to the field of implementation science.

Inclusion Criteria:

• In Mozambique, the inclusion criteria are nurses, general medicine technicians, and physicians in PHC facilities.
• In Zanzibar, Tanzania, the inclusion criteria are clinical officers, nurses, and other healthcare providers involved in consultation, examination, diagnosis, and treatment in PHC facilities.
• In Nepal, the inclusion criteria are medical officers, health assistants, and senior auxiliary health workers in PHC facilities.
• In China, the inclusion criteria of PHC facilities are primary and secondary hospitals, community health centers (stations) and clinics, as well as township health centers and village health clinics; and the inclusion criteria of research participants are practicing (assistant) physicians and rural physicians working in healthcare facilities that meet the above criteria, with a scope of practice only including general practice and internal medicine.

Exclusion Criteria:

• In Mozambique, the exclusion criteria are healthcare providers from other departments, such as tuberculosis, malaria, human immunodeficiency virus/acquired immune deficiency syndrome, or emergency services, as well as students currently on internship.
• In Nepal, the exclusion criteria are interns or students present during the visit.