Effectiveness and safety of a prospective audit and feedback-based antimicrobial stewardship program in hospitalized COVID-19 patients: a quasi-experimental before-and-after study.

BACKGROUND

Antibiotic use among hospitalized patients with confirmed active SARS-CoV-2 infection is notably high (>70%) compared to the incidence of bacterial coinfections and superinfections (3.5% and 14.3%, respectively). Unjustified antimicrobial use poses preventable risks due to its toxicity and potential for long-term negative consequences. In this context, antimicrobial stewardship programs (ASPs) play a pivotal role in optimizing antibiotic therapy for COVID-19 patients.

METHODS

A quasi-experimental, before-and-after study was conducted to assess the impact of the COVID-ASP. The pre-ASP cohort included patients retrospectively, while the ASP cohort included patients prospectively, following the program implementation. The primary outcome was the evaluation of the impact of the COVID-ASP on days of therapy (DOT)/1,000 patient-days. Secondary outcomes included the rate of antibiotic use, rate of inappropriate use, number of recommendations made, their acceptance rate, the incidence of Clostridioides difficile infection, 30-day mortality, and 30-day readmissions.

RESULTS

A total of 1,289 patients admitted were included, 561 (43.5%) in the pre-ASP cohort and 728 (56.5%) in the ASP cohort. The COVID-ASP cohort showed a significantly lower DOT/ 1,000 patient-days (162.96 vs. 105.71; p < 0.001). Additionally, the COVID-ASP was associated with a significant reduction in the rate of antibiotic use for suspected pneumonic coinfections (13.2% vs. 5.9%, p < 0.001), for all causes (17.5% vs. 12.5%, p = 0.012), and for pneumonic superinfection (9.3% vs. 5.9%, p = 0.022). No significant difference was observed in antibiotic use for superinfection from all causes (16.0% vs. 16.2%, p = 0.936). The program also reduced the inappropriate antibiotic use rate for suspected pneumonic coinfection (8.7% vs. 1.9%, p < 0.001), for all causes (10.3% vs. 4.0%, p < 0.001), and for pneumonic superinfection (4.3% vs. 1.9%, p = 0.013). However, it showed no impact on inappropriate antibiotic use in superinfections across all causes (5.9% vs. 5.2%, p = 0.605). The most common recommendations included discontinuing antibiotics (58.6%) and adjusting the antibiotic regimen (30.3%). The program showed no significant effect on the incidence of Clostridioides difficile infections (0% vs. 0.1%, p = 0.379), 30-day mortality (15.2% vs. 18.1%, p = 0.156), or 30-day readmission rates (4.8% vs. 4.9%, p = 0.469).

CONCLUSION

In our experience, the three-step evaluation methodology allows for the evaluation of antibiotic prescriptions in patients with COVID-19 and allows for their optimization.