Empirical antibiotic prescribing in COVID-19 patients: patterns and predictors in a resource-constrained Central Indian setting

Background Empirical antibiotic prescribing during the COVID-19 pandemic may have accelerated antimicrobial resistance, particularly in resource-constrained settings with limited diagnostic and stewardship capacity. This study characterized antibiotic prescribing patterns and identified predictors among hospitalized COVID-19 patients in rural Central India. Methods This retrospective observational study included 951 laboratory-confirmed COVID-19 patients (aged ≥18 years, hospitalized ≥48 hours) admitted between March 2020 and December 2021. Antibiotic use was assessed using WHO ATC/DDD methodology and Drug Utilization 90% (DU90%) analysis. Predictors of antibiotic prescription were evaluated using multivariable logistic regression. Empirical prescribing was defined as antibiotic initiation without microbiological confirmation. Results Among 951 patients (mean age 53 ± 15.8 years; 64% male), 54% had ≥1 comorbidity and 74% required oxygen; 20% required mechanical ventilation. Antibiotics were prescribed to 90% (858/951); 46% received two agents and 8% ≥3 agents. The DU90% segment consisted of WHO “Watch” antibiotics, mainly amoxicillin-clavulanate (40.87 g/100 patient-days), doxycycline (18.26), and azithromycin (17.12). However, the use of meropenem (4.04 DDD gm/100 patients) a WHO Reserve group carbapenem is particularly alarming. Elevated C-reactive protein (>10 mg/dL) was the strongest predictor of antibiotic use (adjusted OR 29.28, 95% CI 17.53–48.89; p<0.001). Antibiotic recipients had longer median hospital stays (11 vs 8 days). Overall mortality was 26%. Conclusions Empirical antibiotic use was extremely high and driven by WHO “Watch” and “reserve” group agents, largely associated with elevated CRP in the absence of culture testing. Strengthening diagnostics and stewardship programs is urgently needed to reduce unnecessary broad-spectrum antibiotic exposure.