Risk factors for extreme-drug resistant Pseudomonas aeruginosa infections in patients with hematologic malignancies

Yin-Xin Liew,Thuan-Tong Tan,Winnie Lee,Jern-Lin Ng,De-Qing Chia,Gee-Chuan Wong,Andrea L Kwa

DOI: https://doi.org/10.1016/j.ajic.2012.02.025

Abstract:Background: Extreme-drug resistant Pseudomonas aeruginosa (XDRPa) infection has emerged in our hematologic unit and is associated with increased mortality. Treatment options are limited, and infection prevention is currently the best option. We aimed to identify the risk factors for XDRPa infection in patients with hematologic malignancies to facilitate institution of stringent infection control measures. Methods: This was a retrospective matched case-control study (26 cases and 53 controls). XDRPa was detected using pulsed-field gel electrophoresis. Data were analyzed by conditional logistic regression and conditional multivariate logistic regression using Stata MP10. Results: Seventy-nine patients (49 males and 30 females; median age, 47 years; range, 16-79 years) were evaluated between December 2006 and December 2008. Independent predictors were use of carbapenem (odds ratio [OR], 10.63; 95% confidence interval [CI], 1.88-59.94; P < .01), Foley catheter insertion (OR, 11.24; 95% CI, 1.075-117.62; P = .04), and bone marrow aspiration (OR, 16.41; 95% CI, 1.57-171.73; P = .02). PFGE revealed 4 patterns. These XDRPa infections were borderline susceptible or resistant to polymyxin B and resistant to all other classes of antibiotics. The infection-attributed mortality of XDRPa was 53%. The median duration of hospitalization was 41 days (range, 19-83 days) in cases after XDRPa isolation and 28 days (range, 5-97 days) in controls. Conclusions: The independent predictors for XDRPa infection were carbapenem use during the current admission, Foley catheter insertion, and bone marrow aspiration. More stringent infection control measures should be applied for patients with these predictors.