Methods: An Infectious Diseases (ID) physician-driven prospective audit and feedback strategy to evaluate the effectiveness of an AMSP in 2 intensive care settings of a tertiary care hospital was performed from January 2016 to July 2017 in 3 phases: baseline, intervention and follow up each consisting of 6 months. In the baseline and follow up period relevant data was recorded. In the intervention phase a patient on antibiotics for > 48 hours was assessed by an ID physician and recommendations made. Primary outcome was days on antimicrobial therapy (DOT) and other secondary outcomes were assessed.
Results: A total of 401, 381 and 379 patients were recruited in the baseline, intervention and follow up phases. Baseline characteristics of the 3 groups were similar. Antimicrobial use decreased from 831.5 during baseline to 717 DOT per 1000 patient days in the intervention (p value<0.0001) and the effect was sustained in the follow-up period (713.6 DOT per 1000 patient days). Among the study antimicrobials, DOTs were significantly lower in the intervention vs. baseline phase for Quinolones(21.5vs.33.3),Carbapenems (340.2 vs.426.0) and Colistin (131.5vs.155.9) (p-<0.0001). De-escalation according to culture susceptibility was significantly higher in the intervention group compared to the baseline (42.7% vs.23.6%-p<0.0001).Compliance to hospital-based antibiotic guidelines significantly improved in intervention and follow up phases compared to the baseline (19.5%, 21.8%, 33.2%; p-<0.0001). We found that 73.3% of antibiotic prescriptions were inappropriate and commonly occurred in the absence of an appropriate clinical indication. Recommendations by the ID team were accepted in 60.7% of the cases. All-cause in hospital mortality rates were 22.4% and 27.6% in the baseline and intervention phases respectively (p-0.093).
Conclusion: An ID physician-driven antimicrobial stewardship programme was successful in reducing antibiotic utilization without compromising patient safety in low and middle-income countries; however this needs further validation.