Etiology and clinical characteristics of neonatal sepsis in different medical setting models: A retrospective multi-center study.

Objective: General hospitals admit lower gestational age neonates than maternal and child health care centers, therefore associated with a higher morbidity and mortality. This study aimed to assess the etiology and clinical characteristics of neonatal sepsis in different medical setting models.

Methods: Neonates admitted to 5 tertiary medical centers, including one national general hospital, two maternal and child health care hospitals and two regional general hospitals, in central-south China with culture-proven sepsis between January 2010 and December 2019 were included in the study. We compared maternal and neonatal characteristics, pathogen distribution, treatment and neonatal outcomes among 3 different medical setting models in this retrospective cohort.

Results: We identified 757 episodes of culture-proven sepsis in 757 neonates. The predominant pathogens were coagulase-negative staphylococci, , and Group B streptococci. A total of 683 neonates with detailed information were involved in further comparison; 54.6% were from the national general hospital, 35.9% were from the maternal and child health care hospital, and 9.5% were from the regional general hospital. Neonates in national and regional general hospitals had significantly lower gestational age and birthweight ( < 0.001). Patterns of pathogen distribution were different among these medical setting models. Early-onset sepsis was more common in maternal and child health care hospitals (61.4% vs. 42.1% vs. 46.7%,  < 0.001), while hospital-acquired late-onset sepsis was more common in national and regional general hospitals (32.7% vs. 33.3% vs. 11.4%,  < 0.001). The proportion of complications or comorbidities of neonates in maternal and child health care hospitals were significantly lower than neonates in national and regional general hospitals ( < 0.001). The case fatality rate was significantly higher in regional general hospitals (10.8% vs. 3.2% vs. 0.8%,  = 0.001).

Conclusion: We report distinct patterns of clinical characteristics, pathogens and outcomes in patient subgroups with neonatal sepsis from national general hospital, maternal and child health care hospital and regional general hospital. It might have some implications for improvement of prevention, management and empirical antibiotic use in neonatal sepsis in different setting models, especially in resource-limited settings from middle and low-income countries.