Association between Timing of Colonization and Risk of Developing Klebsiella pneumoniae Carbapenemase-Producing K. pneumoniae Infection in Hospitalized Patients

Abstract

Colonization by KPC-producing Klebsiella pneumoniae (KPC-Kp) is associ ated with the risk of developing KPC-Kp infection. The impact of the time elapsed since a patient becomes colonized on this risk is not well known. An observational, prospec tive, longitudinal cohort study of colonized patients undergoing active rectal culture screening to rule out KPC-Kp colonization (July 2012 to November 2017). Patients with a positive culture at inclusion (colonized at start of follow-up) and those with a negative culture at inclusion who became colonized within 90 days (colonized during follow-up) were included in the analysis. CART analysis was used to dichotomize variables accord ing to their association with infection. Kaplan–Meier infection-free survival curves and the log-rank test were used for group comparisons. Logistic regression was used to identify variables associated with KPC-Kp infection. Among 1310 patients included, 166 were colonized at the end of follow-up. Forty-seven out of 118 patients colonized at start of follow-up developed infection (39.8%) versus 31 out of 48 patients colonized during follow-up (64.6%; P = 0.006). Variables associated with KPC-Kp infection in the logistic regression analysis were: colonization detection during follow-up (OR, 2.74; 95% CI, 1.07 to 7.04; P = 0.03), Giannella risk score (OR, 1.51; 95% CI, 1.32 to 1.73; P , 0.001), high-risk ward (OR, 4.77; 95% CI, 1.61 to 14.10; P = 0.005) and urological manipu lation after admission (OR, 3.69; 95% CI, 1.08 to 12.60; P = 0.04). In 25 out of 31 patients (80.6%) colonized during follow-up who developed KPC-Kp infection, infection appeared within 15 days after colonization. The risk of KPC-Kp infection was higher when coloni zation is recently acquired during hospitalization. In this prospective study, we con cluded that the timing of colonization was a factor to assess when considering empiri cal treatment for suspected KPC-Kp infection and prophylaxis or infection control.