Trends in sepsis-related neonatal mortality in the United States, 1985-1998

Background: In the United States, bacterial sepsis affects up to 32,000 live births annually. In the 1990s, intrapartum antibiotic prophylaxis (IAP) was recommended to prevent maternal-infant transmission of group B Streptococcus (GBS), a leading cause of sepsis occurring in the first week of life (early onset sepsis). Since IAP has been used, early onset GBS disease declined 70%; however, increased antibiotic use associated with IAP might lead to more severe or antimicrobial resistant etiologies of sepsis. To understand the influence of IAP on neonatal sepsis, in general, we evaluated neonatal mortality from sepsis before and after IAP recommendations were issued. Methods: Using the National Center for Health Statistics Linked Birth/Infant Death Datasets, we compared trends in sepsis-related early neonatal mortality (<7 days) and late neonatal mortality (7-27 days) among singleton United States births from 1985 through 1991 to 1995 through 1998 [data beyond 1998 not included because of International Classification of Diseases (ICD)-10/ICD-9 coding differences]. We compared trends in mortality between the 2 time periods by estimating the average annual percent change in mortality using log linear regression and stratified by gestational age. Results: Combined early and late neonatal mortality from sepsis averaged 39.6/100,000 live births from 1985 through 1991 and 31.8/100,000 live births from 1995 through 1998. Early neonatal mortality from sepsis averaged 24.9/100,000 live births from 1985 through 1991 and 15.6 from 1995 through 1998; late neonatal mortality averaged 14.8/100,000 live births from 1985 through 1991 and 16.2 from 1995 through 1998. Early neonatal mortality declined more steeply after IAP recommendations were issued, 5.0% annually from 1995 through 1998 versus 3.0% annually from 1985 through 1991. Late neonatal mortality increased more from 1995 through 1998, 5.0% annually compared with 0.5% from 1985 through 1991. Conclusions: Lower mortality rates and greater declines in early neonatal mortality from sepsis during 1995-1998 indicate greater survival of infants beyond 7 days of life and suggest an association with GBS disease prevention efforts. Thus these findings provide some evidence for continuing IAP for GBS-colonized women. Our findings of apparent increasing trends in late neonatal mortality from sepsis necessitate follow-up with clinical studies.