1818. Negative MRSA Surveillance Tests after Positive Tests: Evidence of Transient Colonization?
Session: Oral Abstract Session: Addressing Healthcare-Associated MRSA Outside of the Inpatient Setting
Saturday, October 5, 2013: 3:15 PM
Room: The Moscone Center: 200-212

Studies of methicillin-resistant Staphylococcus aureus (MRSA) transmission in the hospital setting often assume that once colonized, patients remain colonized for the duration of their hospital stay. The study of temporal patterns of MRSA colonization is complicated by the variable sensitivity of nasal surveillance testing and the lack of a gold standard. We used surveillance data to investigate temporal patterns in MRSA colonization as well their variation.


We used nasal surveillance data from 111 Veterans Affairs (VA) hospitals, collected for all patients at admission, transfer and discharge. The surveillance tests were performed between 10/1/2007 9/30/2010, during which there were a total of 2,973,845 surveillance tests for 1,424,891 unique patient admissions. We assumed that a first positive test result indicated true patient colonization, and analyzed all tests following the index positive. There were 122,101 admissions with an index positive test followed by at least one follow-up test, resulting in 173,839 tests. Of primary interest was the proportion of tests following the index positive that were negative. We looked at the variation of this proportion across the 111 VA hospitals as well as temporal trends based on the time since the index positive.


There was substantial variation in the proportion negative among the 111 VA hospitals, where the median was 33.7% and the range was [14.2% 61.7%]. Additionally, the proportion negative increased with the time since the first positive test. For tests occurring 1, 10 and 20 days after the index positive, the fraction of all subsequent tests that were negative was 28.1%, 37.2% and 44.6% respectively.


Possible explanations for negative tests after index positives include lack of test sensitivity, poor sampling practices and true loss of carriage. The change in proportion negative over short periods of time suggests that carriage may be lost in a substantial proportion of individuals. Our analysis is limited in that we have not yet taken into account the fact that patients with multiple tests tend to have long lengths of stay, transfers within the hospital, and more medical interventions. Additional effort is needed to better understand these temporal dynamics.







Karim Khader, PhD, VA Salt Lake City Healthcare System, Salt Lake City, UT, Makoto Jones, MD, MS, Medicine, University of Utah School of Medicine, Division of Epidemiology, Salt Lake City, UT, Stephen Kralovic, MD, MPH, National Infectious Diseases Service, Department of Veterans Affairs, Cincinnati, OH, Martin Evans, MD, Internal Medicine, University of Kentucky, Lexington, KY and Matthew Samore, MD, University of Utah School of Medicine, Salt Lake City, UT


K. Khader, None

M. Jones, None

S. Kralovic, None

M. Evans, None

M. Samore, None

<< Previous Abstract | Next Abstract

Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 2nd with the exception of research findings presented at the IDWeek press conferences.