Colistin versus polymyxin B for the treatment of carbapenem-resistant Klebsiella pneumoniae bloodstream infections

To assess the effectiveness of colistin (administered as colistimethate sodium-CMS) and polymyxin B (PMB) for the treatment of bloodstream infections (BSIs) caused by carbapenem-resistant Klebsiella pneumoniae (CRKP).This retrospective cohort included hospitalized adult patients with CRKP BSIs from a single tertiary-care hospital. A univariate analysis comparing CMS and PMB groups was carried out and an inverse-probability propensity score (IPPS) was created. An IPPS-adjusted Cox regression model for 30-day mortality was performed including covariates potentially associated with mortality.A total of 100 patients with CRKP BSI (87 were KPC-producing isolates) were included. The 30-day mortality was 42.0 %:17/46 (38.8 %) and 25/54 (44.6 %) patients of CMS and PMB groups, respectively, P = 0.54 (incidence rate, 18.9 and 21.7/1000 patients-day in CMS and PMB groups, respectively, P = 0.62). No statistically significant difference in 30-day mortality rate was observed in a model adjusted for Pitt bacteremia score, high-risk primary site and IPPS, which included age, intensive care unit admission, minimal inhibitory concentration, previous colonization by CRKP, diabetes mellitus, malignancy, neutropenia, meropenem use before BSI, adjuvant therapy with meropenem and amikacin, and time to start polymyxin. Acute kidney injury (AKI) occurred in 52.0 % of patients, with no significant differences between groups (47.8 % and 57.4 % for CMS and PMB, respectively, P = 0.83). In-hospital mortality was 47,7 % and 50.0 % in CMS and PMB groups, respectively, P = 0.82.There was no difference in 30-day mortality and AKI rates among patients with CRKP BSI treated with PMB or CMS.