Clinical and cost benefit of rapid molecular screening for carbapenem-resistant organisms: a seven-year analysis in a tertiary hospital

Abstract Background The endemic spread of carbapenem-resistant organisms (CROs) is often driven by inadequate prevention of patient-to-patient transmission. This study evaluates the impact of rapid molecular surveillance on in-hospital CRO spread, invasive infections, and costs. Methods In this retrospective, single-center study (Jan 2017–Dec 2023), adult patients at IRCCS Humanitas Research Hospital underwent CRO simultaneous screening using PCR (Xpert™ Carba-R) and culture (Brilliance™ CRE Agar) on admission, with weekly follow-ups in ICUs and onco-hematology wards. Upon PCR result availability, patients were isolated/cohorted. Bloodstream infections (BSIs) were defined by positive blood cultures. Cox regression estimated the risk of incident BSIs. In parallel an economic analysis of the two methods was conducted to asses their impact on resources. Results Among 21,525 samples, PCR and culture showed high concordance (Cohen’s Kappa: 0.909, 95% CI: 0.89–0.93). PCR excluded CRO colonization with 99.7% specificity (95% CI: 99.61–99.76) and a negative likelihood ratio of 0.02. Negative PCR results were available in a median of 2.6 h versus 50 h for culture ( p < 0.001), reducing isolation time and adding ~ 10 bed-days annually in our two-bed setup. PCR sensitivity for culture-proven colonization was 97.7% (95% CI: 96.2–99.18). K. pneumoniae accounted for 88% of isolates, with blaKPC as the predominant gene, followed by blaVIM (32.1% culture-negative). CRO acquisition rates were low in ICUs (1.3/1000 person-days) and onco-hematology wards (0.32/1000 person-days). CRO-BSIs occurred in 75 patients after a median of 12 days, with incidences of 2.35/1000 person-days in colonized versus 0.17/1000 in non-colonized patients. Cox regression identified acquired CRO colonization (aHR = 8.11, 95% CI: 2.38–27.61) and pre-existing blaNDM colonization (aHR = 6.23, 95% CI: 1.05–36.99) as independent predictors of CRO-BSI. The short turnaround times of molecular approach allows to free up beds, increasing the number of patient treatable each years. Conclusions PCR screening is highly concordant with culture, while delivering faster results with positive impacts on isolation time, bed availability, and rapid identification of patients at high risk for CRO-BSIs. From an economic perspective, the molecular approach leads to an optimization of scarce and precious resources.