Predictive Model Discerns Patients at Risk for CRE

Carbapenem-resistant Enterobacteriaceae (CRE), gram-negative bacteria resistant to the carbapenem class of antibiotics, are responsible for high morbidity and mortality internationally. Prompt identification of patients who may harbor CRE helps health care facilities refine their infection control and treatment interventions.

A new case-control study, published in Open Forum Infectious Diseases, found that a prediction model using prior health care exposure information could discern patients likely to harbor CRE at time of hospital admission.

Investigators performed the case-control study utilizing the Illinois hospital discharge database, as well as the Illinois XDRO [extensively drug-resistant organism] Registry. In a 2014-2015 patient cohort, the study team defined cases as index adult patient hospital encounters with a positive CRE culture collected within the first 3 days of hospitalization, as reported to the Illinois XDRO registry.

Controls were all patient admissions in the same month at the same hospital. The data was split into training and validation sets, with approximately 60% and 40% of data dedicated to each. Investigators then developed a logistic regression model to estimate coefficients for predictors of interest.

The study team identified 486 index cases alongside 340,005 controls. They found several independent risk factors for CRE at time of admission: age, number of short term acute care hospitalizations in the previous year, mean short term acute care hospital stay, number of long-term acute care hospital hospitalizations in the past year, mean length of stay in long-term acute care hospitals, current admission to a long-term acute care hospital, as well as prior admission with an infection diagnosis code. When the model was applied to the validation dataset, the area under the receiver operating characteristic curve was 0.84.

Past frequency and duration of health care exposure and prior infection treatment were found to be particularly useful independent predictors for CRE carriage.

The investigators believe that the prediction model they developed could help to stratify patients by CRE risk. They recommended incorporating their model into an automated alerting system in order to prevention efforts like active surveillance and preemptive isolation precautions.

The study authors emphasized the benefits of incorporating a model of the sort they developed, writing that “current strategies that are used to identify patients at high risk of CRE or other multidrug-resistant organism carriage are not efficient.”

The Illinois XDRO Registry used in this study was previously discussed by one of the study authors, Michael Y. Lin, MD, MPH, of Rush University Medical Center, in a presentation at IDWeek 2019. Lin highlighted the usefulness of automatic alerts, and the presentation revealed that 49% of patients who received an alert had an unknown status in the hospital, meaning that if the alert had not been issued the health care providers may not have known about the XDRO colonization.

The Illinois XDRO Registry was implemented to address CRE, but has been applied to other resistant infections, such as Candida auris. Augmenting alert systems of this nature with the predictive risk models discussed in the Open Forum Infectious Diseases article could further provide prompt and actionable information to better the care of patients at risk for CRE infection.