'Normal but dead': perinatal mortality in non-malformed babies of birthweight 2.5 kg and over in the northern region in 1983

The case notes relating to 75 of the 91 perinatal deaths of nonmalformed babies of birthweight greater than or equal to 2.5 kg born in the Northern Region in 1983 were examined. The major groups involved antepartum deaths of unknown cause (40%), and deaths due to intrapartum anoxia or trauma (35%). A case-control study compared each of the 75 cases with two controls matched for place of birth, obtained by taking the next two babies born in the same maternity unit (excluding perinatal deaths, birthweight less than 2.5 kg, and malformations). Four factors were found to be significantly associated with risk of perinatal death in this group: primigravidity, parity greater than or equal to 3, not booked for antenatal care by 20 weeks, and corrected birthweight less than 3.2 kg (adjusted for gestation). Two further factors were related only to the risk of perinatal death consequent upon intrapartum events: labour post-term and malpresentation in labour. All four factors relevant to the whole group remained independently associated with risk of perinatal death after multivariate analysis by two techniques. Adjusted odds ratios (95% CI) were estimated as: primigravidity 2.1 (1.1 to 4.1); parity three or more 5.7 (1.9 to 17); not booked for antenatal care by 20 weeks 15.7 (3.0 to 81); and corrected birthweight less than 3.25 kg 2.5 (1.3 to 4.6). An avoidable factor, as defined, was detected in 50% of deaths. In 30% of deaths there was an avoidable factor (grade 2) such that absence may have been expected to lead to a different outcome had all other factors remained equal. Of the avoidable factors detected, 61% related to intrapartum management, as did 76% of the grade 2 factors. Most of these involved failure to respond to evidence of fetal distress in labour. The defined group constituted 21% of all perinatal deaths, suggesting that this is an important category, particularly as their potential for normal survival should otherwise have been high.