Understanding limited effectiveness of a set of interventions implemented under continuous quality improvement methodology: a process assessment informed by MRC guidance.

INTRODUCTION: Medication errors remain a major contributor to adverse events (AEs) in surgical services, particularly in low-middle income countries (LMICs), where structural and organizational constraints may limit the effectiveness of quality improvement strategies. Continuous quality improvement (CQI) methods are widely implemented to reduce preventable harm; however, evidence regarding their real-world effectiveness and the mechanisms underlying success or failure in LMIC contexts is limited. This study aimed not only to evaluate the impact of a set of interventions under CQI methodology targeting medication prescribing and administration processes, but also to explain the absence of expected effects through a process-informed assessment.

METHODS: A prospective cohort study using an interrupted time series design without a control group was conducted in a high-complexity teaching hospital in Bogotá, Colombia. The intervention targeted prescribing and administration of antibiotics, anticoagulants, and opioids, and included protocol reinforcement, a mobile application (PROA UNAL), standardized pain assessment, and three Plan-Do-Study-Act cycles. Process and outcome indicators were measured biweekly across baseline, implementation, and sustainability phases. Segmented regression and multivariable logistic regression assessed intervention effects, while a structured process assessment examined fidelity, mechanisms of impact, and contextual factors.

RESULTS: A total of 713 patients were distributed across the three phases of the study (194 baseline; 366 implementation; 153 sustainability). The three structured PDSA cycles were conducted, starting from protocol development to the adherence monitoring for each intervention. Implementation fidelity varied across components (100% for prophylactic antibiotic timing, 80% for PROA UNAL application use, and 40% for standardized postoperative pain scale implementation). No significant improvements were observed in process indicators, medication-related AEs, or clinical incidents (CIs). Adjusted models confirmed the absence of intervention effects. Process evaluation identified key contextual barriers-high staff turnover, fragmented governance, limited frontline engagement, and lack of real-time feedback-that hindered implementation.

DISCUSSION: The absence of measurable effects appears attributable to limited fidelity and contextual misalignment rather than intrinsic ineffectiveness of the interventions under the CQI methodology. Future initiatives in LMICs surgical settings should incorporate readiness assessments, governance alignment, real-time feedback, and sustained frontline.