- James W T Chalmers, Consultant in public health medicine,
- Etta Shanks, Senior health information scientist,
- Samantha Paterson, Administrative and technical officer,
- Kevin McInneny, Health information scientist,
- David Baird, Statistical support manager,
- Gillian Penney, Programme coordinator
- Information and Statistics Division, National Health Service in Scotland, Edinburgh EH5 3SQ
- Scottish Programme for Clinical Effectiveness in Reproductive Health, Department of Obstetrics and Gynaecology, University of Edinburgh, Edinburgh EH3 9EW
EDITOR—Stewart et al studied deaths related to intrapartum asphyxia in Wales and found that mortality was higher in babies born at night; during the holiday months of July and August; during February and August, when junior staff rotate to new posts; and towards the end of the week (although the latter two measures were not significant).1 We have attempted to replicate their study, using records relating to births in Scotland during the same period (1993-5). Because of differences between Scottish and Welsh schools' academic terms, the peak holiday period in Scotland is a few weeks earlier than that in Wales. We therefore hypothesised that the effect of annual leave would arise in July alone.
Scotland does not use the confidential inquiry into stillbirths and deaths in infancy; instead, it uses a slightly different system, the Scottish stillbirth and infant death inquiry. This derives detailed information concerning all stillbirths and deaths of infants from the relevant healthcare establishments. This information is sufficient to define intrapartum related deaths, as in the confidential inquiry into stillbirths and deaths in infancy system. For the denominator we used the Scottish morbidity record (maternity) (known as SMR2). This system does not record the time of birth, but we were able to perform a record linkage using probability matching with birth registrations from the registrar general for Scotland, which allowed these data to be appended to each record. In all there were 132 intrapartum related deaths among 184 306 births. For 6427 birth records no time of birth was appended, usually because a matching record could not be found in the registrar general's birth registrations.
The table shows our findings. Altogether 125 deaths were of singleton infants and seven of a twin. Forty one deaths (22 at night) were associated with emergency caesarean section and 11 with elective caesarean section. There were 72 deaths during labour, 57 early neonatal deaths, and three late neonatal deaths. The relative risk of death at night compared with during the day was 1.33 (95% confidence interval 0.95 to 1.88); of death in July 1.06 (0.59 to 1.92); of death in February and August (changeover time) 1.27 (0.83 to 1.95); and of death at weekends (Saturday and Sunday) 1.22 (0.83 to 1.77).
Although these results are in the same broad direction as those of the Welsh study, none are as great and none are significant, despite the larger population. A more detailed comparison of data and of obstetric practices between these countries would be valuable.