Nosocomial infections in a cohort of extracorporeal life support patients.

OBJECTIVES

To examine nosocomial infections in a cohort of patients receiving extracorporeal life support (ECLS) at our institution and to identify the types of infections, impact of prophylaxis, and any apparent risk factors for infection.

METHODS

In a retrospective cohort study, we examined the records of all patients who received ECLS at our institution between August 2009 and March 2011. A prospective, daily, multidisciplinary assessment of all microbiological issues in these patients was carried out, including assessment of microbiological culture positivity and clinical evidence of infection. The results of these assessments were analysed in relation to HELICS (Hospital in Europe Link for Infection Control through Surveillance) and CDC (Centers for Disease Control and Prevention) diagnostic criteria. The use of antimicrobials in these patients was also assessed, as well as the overall bloodstream infection rate in ICU patients.

RESULTS

Seventeen patients received ECLS during the study period, with a total of 445 ECLS days. Of these patients, 13 received respiratory (venovenous) ECLS and four received cardiac (venoarterial) ECLS. There were 17 infections in the cohort: 11 ventilator-associated pneumonias; four bloodstream infections (likely all catheter related, yielding a rate of 9.0 infections/1000 ECLS days); one skin and soft tissue infection; and one urinary tract infection. The bloodstream infection rate in the ICU population as a whole was 9.30/1000 bed-days in 2009 and 7.21/1000 bed-days in 2010. Resistant organisms were identified in 3/17 infections: one methicillin-resistant Staphylococcus aureus, one multidrug-resistant strain of Pseudomonas and one extended-spectrum Β-lactamase-producing Escherichia coli. The median time to acquiring nosocomial infection was 25 days (interquartile range, 13-33 days). The first four ECLS patients received antibacterial (vancomycin) and antifungal (caspofungin) prophylaxis for the duration of ECLS, whereas the later cohort of 13 did not. In patients who received prophylactic antimicrobials, the defined daily dose (DDD) per 100 ECLS days was 49.54 for vancomycin and 49.63 for meropenem. In patients who did not receive prophylaxis, the corresponding DDDs were 25.31 and 37.73, respectively. In ICU patients overall, the DDD per 100 bed-days over the same time period was 13.60 for vancomycin and 19.75 for meropenem. There were 21/445 ECLS days on which antimicrobials were not used.

CONCLUSION

Although ECLS patients are at high risk of acquiring nosocomial infections, the infection rate in our cohort was low. The bloodstream infection rate compared favourably with previously published rates, and was comparable with the bloodstream infection rate among ICU patients as a whole over the same time period. Increased duration of ECLS in this cohort may correlate with an increased rate of infection, consistent with data from other ECLS centres. Antimicrobial use in ECLS patients was high relative to overall use in ICU patients. Larger studies are warranted to evaluate the diagnosis, treatment and overall approach to managing nosocomial infection in ECLS patients.