Quantifying Gram-Negative Resistance to Empiric Treatment After Repeat ExpoSure To AntimicRobial Therapy (RESTART).

ABSTRACT

Background: Antibiotic exposure is a primary predictor of subsequent antibiotic resistance; however, development of cross-resistance between antibiotic classes is also observed. The impact of changing to a different antibiotic from that of previous exposure is not established. Methods: This was a retrospective, single-center cohort study of hospitalized adult patients previously exposed to an antipseudomonal ß-lactam (APBL) for at least 48 hours in the 90 days prior to the index infection with a gram-negative bloodstream or respiratory infection. Susceptibility rates to empiric therapy were compared between patients receiving the same (repeat group) versus a different antibiotic from prior exposure (change group). Results: A total of 197 patients were included (n = 94 [repeat group] and n = 103 [change group]). Pathogen susceptibility to empiric therapy was higher in the repeat group compared to the change group (76.6% vs 60.2%; P = .014). After multivariable logistic regression, repeat APBL was associated with an increased likelihood of pathogen susceptibility (adjusted odds ratio, 2.513; P = .012). In contrast, there was no difference in susceptibility rates between the repeat group and the subgroup of change patients who received an empiric APBL (76.6% vs 78.5%; P = .900). Longer APBL exposure duration (P = .012) and chronic kidney disease (P = .002) were associated with higher nonsusceptibility to the exposure APBL. In-hospital mortality was not significantly different between the repeat and change groups (18.1% vs 23.3%; P = .368). Conclusions: The common practice of changing to a different APBL from that of recent exposure may not be warranted.