Objective: The aim of this study was to evaluate the safety and efficacy of de-escalation in patients under treatment with carbapenems and its impact on clinical outcomes. Methods: A prospective observational study was conducted for 1 year. Patients administered active carbapenems for at least 24 h were included. Primary outcomes were in-hospital mortality, mortality at 30 days after carbapenem prescription, and infection-related readmission within 30 days. De-escalation was defined as the substitution of carbapenem with narrower spectrum antimicrobial agents or its discontinuation during the first 96 h of treatment. Results: The study included 1161 patients, and de-escalation was performed in 667 (57.5%) of these. In the de-escalation group, 54.9% of cultures were positive. After propensity score matching, 30-day mortality was lower (17.4% vs. 25.7%, p = 0.036), carbapenem treatment was 4 days shorter (4 vs. 8 days, p < 0.001), total antibiotic therapy duration was 2 days longer (12 vs. 10 days, p = 0.003), and length of hospital stay was 5 days shorter (8 vs. 13 days, p = 0.008) in the de-escalated versus non-de-escalated patients. Inhospital mortality and 30-day readmission rates did not differ significantly between these groups. Conclusion: Carbapenem de-escalation is a safe strategy that does not compromise the clinical status of patients.

The current morbimortality of serious infections is unacceptable and there is a need to promote the increase in the efficacy of empirical and targeted antibiotherapy. This could be achieved by initiatives coming from ASP teams aimed at promoting increased efficacy of antibiotic therapy .In the optimization of the antibiotic therapy there are several critical points in which an adequate timing could achieve benefits in the survival of patients with severe infections: prompt initiation of empirical treatment; de-escalation performance, appropriate targeted treatment; and finally, curtail antibiotic duration.