Maternal deaths on arrival: A 5-year retrospective study in a health district in South Africa

Maternal deaths on arrival remain one of the blind spots in maternal death surveillance and response programs. There are no policy and guideline addressing maternal deaths on arrival in South Africa. Main objective was to analyse maternal deaths on arrival at the health facilities in a South African district. All maternal deaths on arrival at the health facilities at the Ekurhuleni district over a period of five years and nine months (July 2014 to March 2020), were retrospectively reviewed. Analysis included maternal mortality ratio, demography, gestational age at booking, delivery and deaths, newborn outcome, causes and interval between hospital discharge and presenting as dead. Descriptive statistics (proportion, mean, standard deviation, median and interquartile range) and inferential statistics (Chi-square test, Fishers exact test and Students t test) comparing the women with known and unknown causes, are presented. Twenty-eight (5.72%) women out of the 489 maternal deaths were dead on arrival. The maternal mortality ratio was 7.65 maternal deaths per 100000 live births. Causes of deaths were inconclusive in 17 (60.71%). Among the known causes, the commonest (5, 45.45%) cause was haemorrhage. Median interval between discharge from hospital and presenting as dead was 2.5 days. Information on three delay model were unavailable. Delay two was mentioned for only one year where ambulance service was disrupted due to local protests against municipal services. Comparison between the known and unknown causes showed the regional hospitals had significantly higher numbers of known causes.