Active Surveillance Cultures of MRSA on Admission to ICU Do Not Predict Clinical Culture Isolates

Active Surveillance Cultures of MRSA on Admission to ICU Do Not Predict Clinical Culture Isolates

Scott Norville, MD, Charlotte Luce, RN, Melicia Brown, MPH, Calvin White, MPH; JPS Health Network, Fort Worth, TX

Background: Many institutions perform surveillance cultures in order to identify and isolate patients with resistant organisms. This is based on two assumptions: that colonization with a resistant organism can lead to transmission of the organism to other patients or healthcare workers, and that colonization progresses to active infection with the resistant organism.

Objective: to search for correlation or predictive value between surveillance and clinical cultures.

Methods: Retrospective data are reviewed; listings of patients who had surveillance cultures of the nares and positive clinical cultures were retrieved from the microbiology laboratory and compared for congruence.

Results: 208 patients who had surveillance cultures performed also had positive blood cultures, in 28 cases these cultures showed the same isolate. These ranged from being drawn simultaneously to 800 days apart. Within this range, those blood cultures that matched showed a significantly shorter interval between cultures, 7.2 v. 43 days (p=0.0387 by Mann-Whitney test). When specimens drawn more than thirteen days apart (mean LOS was 12.9 for admissions who had ASC performed) were excluded 16 of 111 patients had blood cultures that matched surveillance cultures, not significantly different by Fisher's exact test (p= 0.8663) from the entire group. This calculates as a false positive rate of 86% for the predictive values of ASC for S. aureus bacteremia of all patients with bacteremia.

A similar analysis was undertaken for patients who had S. aureus isolated from both ASC and non-blood clinical cultures. 52 were found who met these criteria, four with urine cultures, the remainder with skin and soft tissue cultures. Of these 32 had matching ASC/clinical cultures, while 20 did not, an insignificant difference by Fisher's exact test (p= 0.7656).

Conclusions: In this retrospective study surveillance cultures did not reliably predict concurrent clinical culture results. While useful as an epidemiologic and infection control tool, ASC is not accurate as a clinical predictor.

Please wait

Image not available