Carbapenem-resistant Enterobacteriaceae (CRE) are extremely drug-resistant organisms. According to the Centers for Disease Control and Prevention’s National Healthcare Safety Network, in 2014 in the United States, 3.6% of Enterobacteriaceae causing hospital-acquired infections were resistant to carbapenems.1 Antibiotic treatment options for CRE infections are severely limited, and mortality for invasive infections can be as high as 40%-50%.2
Resistance to carbapenems can be mediated by several mechanisms. From an epidemiologic standpoint, production of carbapenemases is the most-threatening mechanism because Enterobacteriaceae-harboring carbapenemases are highly transmissible.
Carbapenemase-producing CRE (CP-CRE) have caused large outbreaks throughout the world. Israel experienced a nationwide outbreak of CP-CRE, primarily Klebsiella pneumoniae carbapenemase–producing Klebsiella pneumoniae, in the mid-2000s. At the peak of the outbreak in 2007, there were 185 new cases per month (55.5/100,000 patient-days). A successful intervention at the national level dramatically decreased the incidence to 4.8/100,000 patient days in 2012.3
One component of the intervention (which is still ongoing) is active surveillance of high-risk groups using rectal swabs. Upon admission to the hospital, we screen patients who were recently in other hospitals or long-term care facilities. In addition, when a patient is newly diagnosed with CP-CRE (either asymptomatic carriage or clinical infection), we screen patients who had contact with that index case before isolation measures were implemented.
We recently published a study in Infection Control and Hospital Epidemiology that draws on our experience with CP-CRE screening of contacts at Tel Aviv Sourasky Medical Center.4 Both Israeli and International guidelines do not precisely define which contacts of a CP-CRE index case warrant screening. For example, should only roommates of index cases be screened or should we screen all patients on the same ward as the index case? Likewise, is there a minimum time of contact that should trigger screening?
Identifying which contacts are at high risk of acquiring CP-CRE is important for two reasons: We want to detect contacts who acquired CP-CRE so that they can be isolated before further transmission occurs, and we don’t want to waste resources and screen those at low risk. In our hospital, the criteria for being a contact are staying in the same ward and being treated by the same nursing staff as a newly identified CP-CRE patient.
This strategy appears to lead to overscreening, as we found that from October 2008 to June 2012, 3,158 screening tests were performed to detect 53 positive contacts (a yield of less than 2%). In order to screen more efficiently, our study aimed to determine risk factors for CP-CRE acquisition among patients exposed to a CP-CRE index patient.
We used a matched case-control design. The case group consisted of the 53 contacts who screened positive for CP-CRE. For each case we chose 2 controls: contacts who screened negative for CP-CRE. The basis for matching between the case and the 2 controls was that they were exposed to the same index patient. The benefit of matching this way was that it eliminated the question of whether a contact became positive because the index patient was more likely to transmit CP-CRE (e.g., because of diarrhea), and not because of characteristics of the contact patients themselves.
We found three factors that increased the risk that a contact would screen positive:
• Contact period of at least 3 days with the index case.
• Being on mechanical ventilation.
• Having a history of carriage or infection with another multidrug-resistant organism (such as methicillin-resistant Staphylococcus aureus).
Unexpectedly, sharing a room with the index patient or being debilitated did not significantly increase the risk of acquiring CP-CRE.
Many studies have identified antibiotic use as a risk factor for acquiring CP-CRE. In our study, no class of antibiotic increased the risk of CP-CRE acquisition, probably because only a small number of patients received each class. We were surprised to find that contacts who had taken cephalosporins were less likely to acquire CP-CRE. On further examination, when we compared patients who received only cephalosporins with patients who received no antibiotic, this protective effect disappeared. Nevertheless, compared with other antibiotics, it appears that cephalosporins might pose less of a risk for CP-CRE acquisition. More studies are needed to confirm our findings.
Our findings have practical implications for infection control. Using the risk factors we identified could help us to avoid excessive screening. We calculated that selective screening, based on our three risk factors, would have decreased the number of contacts screened by 30%, but 2 out of 53 positive contacts would have been missed. Institutions need to decide whether that is a trade-off they are willing to make.
Another way to apply our findings could be to add an additional layer of infection control by preemptively implementing contact precautions for patients at highest risk, for example, those with more than one risk factor.