Risk Factors for Colonization or Infection with Cefepime-Resistant, Piperacillin–Tazobactam Susceptible Gram-Negative Bacilli. (4th October 2017)

• Record Type: Journal Article
• Authors: Wu, Janet; Mynatt, Ryan; Kaye, Keith S; Pogue, Jason M
• Is Part Of: Open forum infectious diseases. Volume 4(2017)Supplement 1
• Page Start: S148
• Page End: S148
• Publication Date: 2017-10-04
• Subjects: Communicable diseases -- Periodicals; Medical microbiology -- Periodicals; Infection -- Periodicals; 616.9
• DOI: 10.1093/ofid/ofx163.237
• Languages: English
• ISSNs: 2328-8957

Abstract

Abstract: Background: Recent literature has suggested increased nephrotoxicity in patients receiving vancomycin and piperacillin/tazobactam (PT) combination therapy when compared with those on vancomycin and cefepime. The primary objective of this study was to determine independent risk factors for cefepime-resistant, PT susceptible Gram-negative bacilli (GNB) isolates to prioritize those requiring empiric PT. Methods: This was a retrospective case-case–control analysis from January 2014-December 2016. Patients with nosocomial or healthcare associated infections were eligible for inclusion if they had blood or respiratory cultures positive for GNB. Three study groups were included in this analysis; cefepime susceptible (CS), cefepime-resistant and PT susceptible (CRPTS), and cefepime-resistant and PT-resistant (CRPTR). Bivariate and multivariate modeling was performed to determine risk factors in two models. Independent risk factors identified in model 1 (CRPTS vs. CS), but not in model 2 (CRPTR vs. CS) would represent unique predictors for CRPTS and patients with these factors would be targets for empiric PT. Results: Three hundred patients (100 in each arm) were included in the analysis. The table displays the two multivariate models. The only independent risk factor for CRPTS was admission from a long-term care facility, however this was also a predictor for CRPTR. Higher comorbidity indices and receipt of PT or ceftriaxone were independent predictors of CRPTR only. Cefepime use had a stronger association with CRPTR (OR 2.60; 95% CI, 0.97–6.98) than CRPTS (OR 1.75; 95% CI, 0.76–4.00). Conclusion: In the presence of cefepime resistance, unique independent predictors for PT susceptibility could not be identified. However, prior uses of PT or ceftriaxone, as well as ICU admission were unique predictors of CRPTR and thus represent scenarios where both empiric cefepime and PT should be avoided. Disclosures: All authors: No reported disclosures.