Multi-drug resistant organism infections in a medical ICU: Association to clinical features and impact upon outcome.

OBJECTIVE: To define clinical features associated with Intensive Care Unit (ICU) infections caused by multi-drug resistant organisms (MDRO) and their impact on patient outcome. DESIGN: A single-center, retrospective case-control study was carried out between January 2010 and May 2010. SETTING: A medical ICU (MICU) in the United States. PATIENTS: The study included a total of 127 MDRO-positive patients and 186 MDRO-negative patients. INTERVENTIONS: No interventions were carried out. RESULTS: Out of a total of 313 patients, MDROs were present in 127 (41.7%). Based on the multivariate analysis, only infection as a cause of admission [OR 3.3 (1.9-5.8)]), total days of ventilation [OR 1.07 (1.01-1.12)], total days in hospital [OR 1.04 (1.01-1.07)], immunosuppression [OR 2.04 (1.2-3.5)], a history of hyperlipidemia [OR 2.2 (1.2-3.8)], surgical history [OR 1.82 (1.05-3.14)] and age [OR 1.02 (1.00-1.04)] were identified as clinical factors independently associated to MDROs, while the Caucasian race was negatively associated to MDROs. The distribution of days on ventilation, days in hospital and days of antibiotic treatment prior to infection differed between the MDRO-positive and MDRO-negative groups. The MDRO-positive patients showed a greater median number of days in hospital and days of antibiotic treatment before infection, with a greater median number of days in hospital, days of antibiotic treatment and days of ventilation after infection, compared to the MDRO-negative patients. The mortality rate was not significantly different between the two groups. Appropriate empirical antibiotic therapy was prescribed in 82% of the MDRO-positive cases - such treatment being started within 24h after onset of the infection in 68.5% of the cases. CONCLUSION: Defining clinical factors associated with MDRO infections and administering timely and appropriate empirical antibiotic therapy may help reduce the mortality associated with these infections. In our hospital we did not withhold broad spectrum drugs as empirical therapy in patients with clinical features associated to MDRO infection. Our rate of appropriate empirical therapy was therefore high, which could explain the absence of excessive mortality in patients infected with MDROs.