Every maternal near-miss counts: Need for a national audit in South Africa? A mixed-methods study

Background. To improve maternal health, studies of maternal morbidity are increasingly being used to evaluate the quality of maternity care, in addition to studies of mortality. While South Africa (SA) has a well-established confidential enquiry into maternal deaths, there is currently no structure in place to systematically collect and analyse maternal near-misses (MNMs) at national level.Objectives. To synthesise MNM indicators and causes in SA by performing a systematic literature search, and to investigate perceived needs for data collection related to MNMs and determine whether the MNM tool from the World Health Organization (WHO-MNM) would require adaptations in order to be implemented. Methods. The study used a mixed-methods approach. A systematic literature search was conducted to find all published data on MNM audits in SA. Semi-structured interviews were conducted virtually with maternal health experts throughout the country who had been involved in studies of MNMs, and main themes arising in the interviews were synthesised. A method for MNM data collection for SA use was discussed with these experts.Results. The literature search yielded 797 articles, 15 of which met the WHO-MNM or Mantel et al. severe acute maternal morbidity criteria. The median (interquartile range) MNM incidence ratio in SA was 8.4/1 000 (5.6 -8.7) live births, the median maternal mortality ratio was 130/100 000 (71.4 -226) live births, and the median mortality index was 16.6% (11.7 -18.8). The main causes of MNMs were hypertensive disorders of pregnancy and obstetric haemorrhage. Eight maternal health experts were interviewed from May 2020 to February 2021. All participants focused on the challenges of implementing a national MNM audit, yet noted the urgent need for one. Recognition of MNMs as an indicator of quality of maternity care was considered to lead to improved management earlier in the chain of events, thereby possibly preventing mortality. Obtaining qualitative information from women with MNMs was perceived as an important opportunity to improve the maternity care system. Participants suggested that the WHO-MNM tool would have to be adapted into a simplified tool with more clearly defined criteria and a number of specific diagnoses relevant to the SA setting. This 'Maternal near-miss: Inclusion criteria and data collection form' is provided as a supplementary file. Conclusion. Adding MNMs to the existing confidential maternal death enquiry could potentially contribute to a more robust audit with data that may inform health systems planning. This was perceived by SA experts to be valuable, but would require context-specific adaptations to the WHO-MNM tool. The available body of evidence is sufficient to justify moving to implementation.