![[ Visit Client Website ]](images/banner.gif)
Purpose: Recent analysis of 2010 antibiogram data from our institution suggests equal activity of cefepime (Cef) and meropenem (Mer) against gram-negative (GN) bloodstream and respiratory pathogens with Cef and Mer both being more active than piperacillin/tazobactam (P/T). APACHE IV predicted mortality (A-IV) is a well validated model for predicting patient outcomes in the ICU. We sought to evaluate risk-adjusted mortality of patients admitted to the medical ICU (MICU) who required GN therapy using Cef, Mer, or P/T.
Methods: Adult MICU patients were included who had A-IV prospectively calculated and received Cef, Mer, or P/T on ICU admission from November 2011 to March 2012. A-IV was used to predict mortality and ICU length of stay (LOS). All-cause hospital mortality and actual ICU LOS minus predicted ICU LOS were assessed based on selection of Cef, Mer, or P/T for GN coverage with multivariate analysis and standardized mortality ratios (SMR).
Results: This study included 182 patients (n=24 Cef, n=33 Mer, n=125 P/T) for analysis. A-IV was similar between groups [mean % (SD) for Cef 39 (28), Mer 34 (24), P/T 31 (29), P>0.05 for all]. The SMR for Mer was 1.6 compared with 0.75 and 0.77 for Cef and P/T, respectively. Logistic regression controlling for A-IV, antibiotic choice, suspected respiratory source of infection, and gram-positive therapy revealed that use of Mer was associated with greater odds of death (OR 4.2, 95% CI 1.8-9.8). Cef or P/T was not predictive of mortality in this model. Risk-adjusted ICU LOS was longer than predicted with Mer by 4 days (95% CI 1.6-6.5) and similar to predicted for Cef and P/T.
Conclusion: MICU patients given Mer had greater odds of death and longer than predicted ICU LOS compared with Cef or P/T, even when risk-adjusting for acuity of illness. Further study of this interesting phenomenon is warranted.