Effectiveness of clinical algorithms combined with point-of-care testing to improve antibiotic prescribing practices of physicians in outpatient setting of a low- to middle-income country: A pre-post–intervention study

Objectives Antimicrobial resistance (AMR), driven by misuse of antibiotics, is a major global health threat. This study evaluated the impact of clinical algorithms and point-of-care C-reactive protein (POC-CRP) testing on antibiotic prescribing for acute watery diarrhea (AWD), upper respiratory tract infections (URTI), and acute undifferentiated fever (AUF) in children and adults at Aga Khan University Hospital (AKUH), Pakistan. Methods A pre-post–intervention study (January 2022 to September 2024) was conducted at AKUH and two affiliated sites in Karachi. Children (aged ≥6 months) and adults with AWD, URTI, and AUF were enrolled. A semi-quantitative POC-CRP test was integrated into disease-specific algorithms as the intervention. The primary outcome was change in prescription of specific antibiotics; secondary outcomes included hospitalizations, emergency visits, inappropriate antibiotic prescriptions, and physician adherence. Results A total of 457 and 460 patients were enrolled in the pre- and post-intervention phases respectively. Two-thirds were children (pre: 296 of 457 [64.8%], post: 269 of 460 [58.5%]), mostly male (pre-intervention: 275 of 457 [60.2%]; post-intervention: 248 of 460 [53.9%]). Appropriate antibiotic use improved by 11.9%, with overall antibiotic prescriptions rising from 28.9% to 38.5%. The use of azithromycin (% point difference: −11.4; 95% confidence interval: −33.1 to 10.3) in children with URTI declined significantly. At day 7 follow-up, clinical improvement did not differ between groups (95.0% vs 95.9%; P = 0.51). Clinic revisits were lower in the post-intervention period (27.8% vs 20.0%; absolute difference −7.8%, 95% confidence interval −13.3 to −2.3; P = 0.006). Emergency visits (6.3% vs 5.4%; P = 0.79) and 28-day hospitalization rates (2.0% vs 1.5%; P = 0.61) were low and showed no significant differences between groups. Overall, 72% (330 of 460) of patients were referred by physicians for POC-CRP testing, with the highest adherence at AKUH (81%). All physicians who participated in the feedback session (14 of 14) reported being satisfied and agreed that the intervention supported clinical decision-making. Conclusion The intervention improved appropriate antibiotic prescriptions. Broader studies are needed across diverse and rural health care sectors.