Objective: Early onset group B streptococcal (EOGBS) infection, the major neonatal infection in industrialized countries, can be prevented by intrapartum antibiotics, but population studies are lacking. This study aimed to determine the incidence of early onset infections caused by group B Streptococcus (GBS) and other organisms in Australia and to assess intrapartum antibiotic use.
Design: Longitudinal, prospective surveillance of neonatal infections in Australian neonatal units from 1991 to 1997. Early onset infection defined as clinical sepsis in first 48 h after birth, with positive cultures of blood or cerebrospinal fluid or positive urine GBS antigen detection.
Results: The incidence of EOGBS sepsis fell from 2.0 per 1000 live births (95% confidence interval, 1.4, 2.5) in 1991 to 1993, to 1.3 (1.2, 1.4) in 1993 to 1995, to 0.5 (0.4, 0.7) in 1995 to 1997 (P < 0.0001). The incidence in Aboriginal babies was 5.2 (1.8, 8.6) in 1991 to 1993, 5.1 (3.0, 7.2) in 1993 to 1995 and 1.8 (1.1, 2.5) in 1995 to 1997 (P < 0.05). The incidence of early onset infections caused by organisms other than GBS also fell, from 1.2 per 1000 live births (0.8, 1.7) in 1991 to 1993, to 0.8 (0.7, 0.9) in 1993 to 1995 and 0.5 (0.3, 0.7) in 1995 to 1997 (P < 0.0001). In 1991, 3 of 9 study hospitals had a formal policy on intrapartum antibiotic use, whereas in 1997 all 11 hospitals had a formal policy (P=0.002).
Conclusions: A steady fall in EOGBS infections in Australia from 1991 to 1997 has been associated with increasing use of intrapartum antibiotics. Increased antibiotic use is probably causal in the fall in GBS, because the incidence of early onset infections caused by other organisms has also fallen.