A register-based study of the impact of obstetric and neonatal care on stillbirth and neonatal death rate was performed on all births in Sweden in 1983–1995. Each birth was assigned to a primary delivery hospital where the mother with a term singleton pregnancy was most likely to have been delivered (not possible for 25% of the deliveries), and the catchment areas of each hospital were classified according to the level of care of that hospital. Only small differences in total mortality existed between the different levels of care of the primary hospital: areas served by primary hospitals with obstetric service and resources for neonatal intensive care including continuous positive airway pressure but without facilities for ventilator treatment for prolonged periods showed a 7% excess risk of stillbirth or neonatal death.
Conclusion: In areas with the lowest level of care of the primary delivery hospitals (with no or only basic neonatal care) the total mortality was not increased, indicating that the referral system works well. When the analysis was repeated for specific causes of death, more marked differences were noted, especially for death due to obstetric complications where the death risk increased with decreasing level of care of the primary delivery hospital. Even though no marked differences in total mortality were seen, a further reduction can be obtained by increasing referral for some specific conditions.