Abstract
IMPORTANCE: The national impact of replacing systematic microbiological sampling with clinical surveillance of asymptomatic newborns at risk of early-onset neonatal bacterial infection (EONI) has not been assessed in France. Specifically, it is not known whether the new guidelines could lead to delayed or missed diagnoses, potentially increasing rates of severe or late-onset infections. OBJECTIVE: To assess changes in hospitalization rates for neonatal bacterial infections (NBIs) and their associated costs following the introduction of the 2017 EONI management guidelines in France. DESIGN, SETTING, AND PARTICIPANTS: This nationwide, retrospective cohort study was conducted from January 1, 2014, to December 31, 2023. The analysis was based on the French National Hospital Discharge Database, which covers the activity of all public and private health care facilities in France. Participants were newborns of at least 34 weeks' gestational age (GA) hospitalized for an NBI in the first 28 days of life in metropolitan France. EXPOSURE: The 2017 French EONI management guidelines. MAIN OUTCOMES AND MEASURES: The primary outcomes were monthly incidence rates of NBI hospitalizations per 1000 live births of at least 34 weeks' GA between 2014 and 2023. Changes in NBI hospitalization rates following the 2017 management guidelines for EONI were assessed using monthly time series of incidence rates and segmented regression models with autocorrelated errors. RESULTS: A total of 68 107 EONI hospitalizations were recorded (37 316 male newborns [54.59%]; 61 264 newborns [89.95%] of ≥37 weeks' GA). The expected incidence rate for nonsevere EONI in March 2018 (start of the postrecommendation period) was 9.13 hospitalizations per 1000 live births (95% CI, 8.27 to 9.98 hospitalizations per 1000 live births), which was significantly higher than the observed incidence rate of 6.17 hospitalizations per 1000 live births (95% CI, 5.74 to 6.61 hospitalizations per 1000 live births) at the same time, corresponding to a relative difference of -32.36% (95% CI, -41.54% to -22.18%). A similar trend was observed for EONI without neonatal intensive care unit admission; the expected incidence rate was 9.46 hospitalizations per 1000 live births (95% CI, 8.56 to 10.36 hospitalizations per 1000 live births), which was significantly higher than the observed incidence rate of 6.68 hospitalizations per 1000 live births (95% CI, 6.22 to 7.13 hospitalizations per 1000 live births), resulting in a relative difference of -29.42% (95% CI, -40.16% to -18.69%). No significant differences were found for severe EONI or EONI requiring neonatal intensive care unit admission. The incidence of late-onset NBI remained stable throughout the study period. Between 2014 and 2023, the total cost for all EONI hospitalizations decreased by half, from €45 975 289 to €23 459 606. CONCLUSIONS AND RELEVANCE: In this nationwide cohort study, implementation of the 2017 French EONI guidelines was likely associated with improvements in the management of NBI, contributing to a reduction in nonsevere hospitalizations with no change in the rates of severe cases or late-onset infections.