Carbapenem-Resistant Enterobacteriaceae, Long-Term Acute Care Hospitals, and Our Distortions of Reality

L. Munoz-Price

DOI: https://doi.org/10.1086/671264

2013-06-27

Abstract:Our knowledge about long-term acute care hospitals (LTACHs) and their role in the regional epidemiology of multidrug-resistant organisms has markedly evolved during the past decade." In this issue of the journal, Lewis et al report findings from their performance of rectal surveillance cultures among all patients admitted to a LTACH. From October 2010 through July 2012, 12 of 262 patients were positive for carbapenem-resistant Enterobacteriaceae (CRE) on admission, 6 of whom were known carriers of this strain at the time of their transfer from various acute care hospitals. Additionally, 7 patients who were negative on admission had positive rectal cultures later during their hospitalizations at the LTACH. An eighth patient was positive by rectal surveillance during his stay at the LTACH but failed to undergo culturing on entry to the facility. Only 1 of 20 CRE-positive patients had clinical cultures that grew CRE during their hospitalizations. These findings suggest that there was an influx of patients colonized with CRE from acute care hospitals to the LTACH (12 [63%] of 19). However, active horizontal transmission of CRE was also occurring within the facility (7 [37%] of 19). Let us go through some potential scenarios of how these data could have been interpreted depending on the timing and location of surveillance cultures and the presence or absence of communication between facilities (see Figure 1). Scenario 1. What if cultures would not have been performed on admission but only later during the stay at the LTACH? Then we would have detected 13 positive patients (6 unknown as positive at entry and 7 cases acquired in the facility) and might have concluded that all 13 cases were associated with—or even caused by—the facility. Furthermore, if transfer of information between facilities would have been deficient, then the status of the 6 patients with a previous history of CRE might have been unknown by the LTACH, further increasing the number of positive cases detected in the facility to 20 (100%). Where did these acquisitions occur? Scenario 2. What if surveillance cultures would not have been performed at all at the LTACH? Then detection would have been limited to passive surveillance of clinical cultures. In this scenario, the facility would have detected only 1 patient positive for CRE (1 [5%] of 20). Scenario 3. Let us imagine that all 20 cases (19 unrecognized and 1 clinical infection) are transferred back to a single acute care hospital (for simplicity) and that this hospital performs active surveillance for CRE on admission. Then all 20 CRE cases would be associated—with causal implications—to the stay at the LTACH. Scenario 4. Let us go back to the beginning of scenario 3 and now assume that the 20 cases (19 unrecognized and 1 clinical infection) are transferred to a single acute care hospital that does not perform active surveillance cultures on admission and that there is limited transfer of information. Now imagine that 2 or 3 of the 20 develop positive clinical cultures during their acute care hospital stay. These cases would be the only ones identified unless point prevalence surveillance was performed in the facility. Point prevalence surveillance would identify the unrecognized CRE-positive patients in addition to any other cases acquired within the facility. A case-control study done at this point would likely determine that stay at a LTACH is associated with CRE colonization. What if all of these 17 or 18 unrecognized cases end up going to other LTACHs or long-term care facilities that do not conduct surveillance? Then the cycle repeats all over again.