Outbreak of Colistin Resistant Enterobacteriaceae at Detroit Medical Center

Background: Carbapenem resistant enterobacteriaceae (CRE) have spread in Southeastern Michigan. Colistin is used extensively to treat these organisms in the Detroit Medical Center (DMC).  We describe a cluster of colistin-resistant CRE infections that occurred at DMC and an attached long-term acute care facility (LTAC).

Objective: To conduct a comprehensive epidemiologic investigation of colistin-resistant CRE strains outbreak at DMC, including risk factors and outcomes assessment.

Methods: A cluster of 5 cases of colistin-resistant Klebsiella spp. emerged at DMC. Epidemiologic data were collected and transmission opportunities were analyzed. A transmission opportunity (TOP) was defined as two case patients staying on the same ward at the same time. Data regarding the usage of colistin were obtained from pharmacy records. Colistin defined daily doses (DDDs) were based on the assumption of a daily dose of colistin being 3 million units/day (90 mg base activity). Resistance to colistin was defined as MIC > 4.  Cases were defined as patients with colistin-resistant CRE and controls were defined as patients with colistin-susceptible CRE isolated from 09/2008 to 09/2009.

Results: The first case of colistin-resistant CRE occurred in 07-27-2009, followed by 4 additional cases occurring between 08-16 and 08-22, 2009. Four of the cases were Klebsiella pneumoniae (2 from urine and 2 from wounds) and one case was of K. oxytoca bacteremia. The strains were isolated in 2 hospitals and one LTAC (attached to one of the affected hospitals).  All cases, at a certain point, had stayed at a single involved hospital, or in the LTAC attached to it. The mean TOPs between cases was 2.3±0.5, and each case had at least one TOP with one of the other cases. When comparing the five colistn-resistant CRE cases to the 60 colistin-susceptible CRE controls, the MIC to imipenem for cases was significantly higher for cases (p<0.001).  The mean age of case patients was 77±6 years, significantly older compared to controls (62±8 years, p=0.05). The mortality rate and length of stay were higher among cases than controls, though these differences did not reach statistical significance (40% and 26%; and 33±23 and 30±23 days, respectively). Colistin utilization during the quarter of the year immediately prior to cluster emergence (04-01 to 06-30, 2009), did not differ from the previous quarter (01-01 to 03-31, 2009), with a mean DDD/1,000 patient days of 20.7±3.4 for the entire 6 months studied.

Conclusions: This is the first report to our knowledge of a colistin-resistant CRE outbreak in the U.S.  Geonotyping studies are ongoing.  Although raw colistin exposure was not associated with the outbreak, further data regarding colistin dose and dosing intervals are needed to optimize killing and to prevent the emergence of resistance.  This outbreak was controlled through implementation of strict infection control practices and effective communication at the 2 hospitals and LTAC.