Background: Many maternal deaths (as well as related severe morbidity) are of women who do not attend antenatal care in a given health unit but are referred there when they develop life-threatening obstetric complications.
Objective: To determine the reproductive characteristics of emergency obstetric referrals, and determine the contribution of emergency obstetric referrals to severe acute maternal morbidity (near-misses) and maternal mortality.
Study design: Descriptive cross-sectional study.
Setting: Mulago hospital, the National Referral hospital, Kampala, Uganda, from 1st March to August 30th 2000.
Subjects: Nine hundred and eighty three consecutive women admitted as emergency obstetric referrals in labour or puerperium.
Interventions: Subjects were followed from time of admission to discharge (or death). They were interviewed (or examined) to obtain data on socio-demographic characteristics, reproductive history, obstetric outcome of the index pregnancy, obstetric complications and cause of death. Their records were reviewed to determine evidence of severe acute morbidity from acute organ/system dysfunction, using the definition by Mantel et al. These data were analysed using the Epilnfo computer programme in terms of means, frequencies and percentages.
Main outcome measures: Socio-demographic characteristics, obstetric complications, cause of deaths, cause and type of near miss mortality and case fatality rates.
Results: Of the 983 referrals, over 100 were near-misses and 17 died. Using the definition of Mantel et al of near-misses enabled identification of six times as many near-misses as maternal deaths. The commonest causes of death were postpartum haemorrhage and eclampsia. Low status was highly associated with both maternal deaths and near misses.
Conclusion: In developing countries, with poor obstetric services, emergency transfers in labour are very common. These women, who are of low status, contribute significantly to maternal mortality and morbidity.