AUTHOR=Ruvinsky Silvina , Voto Carla , Roel Macarena , Deschutter Verónica , Ferraro Daiana , Aquino Norma , Reijtman Vanesa , Galvan María Eugenia , Motto Eduardo , García Mauro , Sarkis Claudia , Bologna Rosa TITLE=Carbapenem-resistant Enterobacteriaceae bloodstream infections: A case-control study from a pediatric referral hospital in Argentina JOURNAL=Frontiers in Public Health VOLUME=10 YEAR=2022 URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2022.983174 DOI=10.3389/fpubh.2022.983174 ISSN=2296-2565 ABSTRACT=Background

Antibiotic-resistant gram-negative bloodstream infections (BSI) remain a leading cause morbidity and mortality in pediatric patients with a high impact on the public health system. Data in resource-limited countries, including those in Latin America and the Caribbean region, are scarce. The aim of the study was to identify risk factors for acquiring carbapenem-resistant Enterobacteriaceae (CRE) bacteremia in children and to assess the use of resources.

Methods

A retrospective case-control study was conducted to analyze demographic, epidemiological, clinical, microbiological, and outcome data as well as the use of resources between 2014 and 2019. Univariate and logistic regression analysis was performed in order to identify risk factors associated with CRE-BSI. The R software version 4.1.2 was used.

Results

A total of 46 cases with CRE-BSI and 92 controls with gram-negative non-CRE-BSI were included. No statistical difference was observed regarding: median age (36 months; IQR, 11.2–117 vs. 48 months, IQR 13–119), male sex (50 vs. 60%), and underlying disease (98 vs. 91%) in cases vs. controls, respectively. The most frequent mechanism of CRE bacteremia were: KPC in 74%, OXA in 15%, and NDM in 6.5%. A total of 54.3% of cases vs. 32.6 % (p = 0.016) of controls were admitted to the pediatric intensive care unit (PICU), and 48 vs. 21% (p = 0.001) required mechanical ventilation. Bacteremia secondary to intra-abdominal infection was observed in 56.5% of cases vs. 35% of controls (p = 0.032). Previous colonization with CRE was detected in 76% of cases vs. 8% of controls. Combination antimicrobial treatment was most frequent in cases vs. control (100 vs. 56.5%). No difference was observed in median length of hospital stay (22 days; IQR, 19–31 in cases vs. 17.5 days; IQR, 10–31 in controls; p = 0.8). Overall case fatality ratio was 13 vs. 5.5%, respectively. The most statistically significant risk factors included previous PICU stay (OR, 4; 95%CI, 2–8), invasive procedures/surgery (OR, 3; 95%CI, 1–7), central venous catheter placement (OR, 6.5; 95%CI, 2–19), urinary catheter placement (OR, 9; 95%CI 4–20), mechanical ventilation (OR, 4; 95%CI, 2–10), liver transplantation (OR, 8; 95%CI, 2–26), meropenem treatment (OR, 8.4; 3.5–22.6) in univariate analysis. The logistic regression model used for multivariate analysis yielded significant differences for previous meropenem treatment (OR, 13; 95%CI, 3-77; p = 0.001), liver transplantation (OR, 13; 95%CI, 2.5–100; p = 0.006), and urinary catheter placement (OR, 9; 95%CI, 1.4–94; p = 0.03).

Conclusion

CRE-BSI affects hospitalized children with underlying disease, mainly after liver transplantation, with previous urinary catheter use and receiving broad-spectrum antibiotics, leading to high PICU requirement and mortality. These risk factors will have to be taken into account in our region in order to establish adequate health policies and programs to improve antimicrobial stewardship.