Antimicrobial escalation is not beneficial for Gram-negative bacteremia in adults who remained critically ill after appropriate empirical therapy.

Abstract To study whether antimicrobial escalation is beneficial for the outcome of bacteremia patients receiving appropriate but less responsive antimicrobials as empirical therapy, adults with community-onset Gram-negative bacteremia and remained the critical illness after appropriate empirical therapy with third-generation cephalosporins were retrospective enrolled. Clinical outcomes included incidental nosocomial infections, breakthrough bacteremia, and in-hospital crude mortality were compared between patients receiving escalation and non-escalation therapy, after propensity-score (PS) matching at a ratio of 1:3 using independent predictors of 30-day mortality recognized by the multivariate regression model. Initially, the higher proportion of fatal comorbidities (McCabe classification) and 30-day mortality rates was exhibited in the escalation group (51 patients), compared to the non-escalation group (de-escalation, 81; non-switch, 95). After appropriate PS-matching, similar proportions of clinical variables between the escalation (45 patients) and no-escalation (135) groups, in terms of patient demographics, bacteremia severity at onset, severity and types of comorbidities, and bacteremia sources, were observed. Consequently, poorer clinical outcomes, such as the higher rate of incidental nosocomial infections and in-hospital crude mortality as well as the longer length of intravenous antimicrobial administration and hospitalization, were statistically evidenced in the escalation group, compared to the non-escalation group. Conclusively, for patients exhibiting poor responses to appropriate empirical therapy, antimicrobial escalation did not significantly result in improved outcomes; otherwise, clinicians should pay more attention to the strategy of supportive care or adequate control of septic complication.