AFJOG

African Journal of Obstetrics and Gynaecology | Volume 3 | Issue 2 | 2025 | 37 ORIGINAL RESEARCH African Journal of Obstetrics and Gynaecology | Volume 3 | Issue 2 | 2025 | Comparing two maternal near miss criteria in an academic hospital in Johannesburg, South Africa: A retrospective cross-sectional study Table 3: Incidence of maternal near misses (MNM), MNM ratio and maternal near miss to maternal deaths ratio Description First period Second period P-value* Absolute numbers Number of maternal near misses and deaths 38 58 0.06 Number of maternal deaths 8 4 Number of live births during the period 9034 9074 Ratios Maternal near miss ratio 421 per 100 000 639 per 100 000 Maternal mortality ratio 88 per 100 000 44 per 100 000 Maternal near miss to maternal 5:1 14:1 0.89 mortality ratio 7 (18.4) 8 (13.8) Mortality index 17% 7% The maternal near miss to maternal mortality ratio (MNM to MMR in this study is like a study by Iwuh et al, [7] Oppong [2] and Soma Pillay et al [8] but less to that of Roopa et al. [9] We observed a decrease in the mortality index (MI) from 17% to 7%, after the updated criteria was implemented. More near misses were identified and managed using the new criteria which possibly prevented more maternal deaths by enabling early intervention. In Table 4, the timing of the near misses was assessed. There were no significant differences between the two groups. However, most of the near misses occurred when women had a hospital delivery compared to a clinic delivery (89.4 vs 57.6%). While there were no differences between the groups in terms of pre-existing medical conditions, hypertension (89.5 vs 84.5%) and diabetes mellitus (97.4 vs 98.2%) were the most common pre-existing medical conditions observed among the near misses (Table 5). This is comparable to the studies by Oppong [2] , Souza et al [10] and Heemelaar. [11] Similarly Habtei et al found an increased incidence of MNM with diabetes and hypertension. [12] Despite the second near miss period coinciding with Covid 19 pandemic, the pandemic did not affect the profile of the near misses between the two study periods. The leading causes of near misses were obstetric haemorrhage (19/96; 19.8%), complications of caesarean sections (17/96; 17.7%), followed by hypertension and medical disorders (7/96 cases each; 7.3%) (Table 5). During the first period, 34.8% of the women had either an ante- or post-partum haemorrhage, compared to 39.3% during the second period. Despite the second period coinciding with the Covid 19 pandemic, there was a slight decline in maternal near misses related to hypertensive disorders in pregnancy, while near misses related to obstetric haemorrhage and caesarean section related complications were in the opposite direction. Table 4: Timing of the maternal near miss over the two periods Description First period N=38 n (%) or median (IQR) Second period N=58 n (%) or median (IQR) P-value* Delivered 37 (97.8) 57 (98.3) 0.76 EGA at delivery 38 (31-41) 38 (32-42) 0.49 Type of delivery NVD 9 (24.3) 9 (15.9) 0.18 CS 24 (64.9) 28 (49.1) Miscarriage 2 (5.4) 11 (19.3) Ectopic 1 (2.7) 6 (10.5) Illegal TOP 1 (2.7) 2 (3.5) BBA 0 (0.0) 1 (1.8) When near miss happened 0.21 (n=52) 17 (89.4) 19 (57.6) Hosp or clinic delivery 0 (0.0) 1 (3.0) Admission after being discharged 0 (0.0) 6 (18.2) Post miscarriage 0 (0.0) 1(3.0) Post ectopic 0 (0.0) 1 (3.0) Unknown Table 5: Medical conditions associated with near misses to that of maternal deaths during the study period Description First period N=38 Second period N=58 P-value* Chronic hypertension 34 (89.5) 49 (84.5) 0.49 Diabetes 37 (97.4) 57 (98.2) 0.76 Hypothyroidism 0 (0.0) 1 (1.7) 0.42 Valvular heart diseases 0 (0.0) 1 (1.7) 0.42 Other chronic 15 (39.5) 20 (34.5) 0.62 Covid 19 positive during pregnancy or puerperium 0 (0.0) 4 (6.9) 0.09 Asymptomatic Covid 19 0 (0.0) 3 (5.5) 0.09 When changing to the WHO criteria, low platelets of less than 50 x10 9/l were included. INR was also statistically significant when comparing the two groups. However, this was not part of the changes for the new near miss criteria, but an incidental finding for patients. The rest of the blood results reviewed were similar for both periods. (Table 6). More women were admitted to ICU in the first group than the second group (42.1% vs 36.2%), however this was not statistically significant. More maternal deaths occurred in ICU than in the ward, in the first group (75%) (Table 7). This is somewhat expected as women who are admitted to