P-24 Characteristics of pregnant women identified as nearmiss and factors related to maternal deaths in Saint Lucia: a review of case records

Methods: This descriptive study included all maternal deaths from 2006-2015 (n=23), near-miss (n=111) and potential near-miss (n=121) from 2011-2015. Uncomplicated pregnancies for 2015 were used for comparison. Data was analyzed using Chi square, ANOVA, T-tests and Logistic Regression.

Results: 47.8% of maternal deaths were misclassified (absent from the national maternal mortality database). The main causes for misclassification were lack of pregnancy information on death certificates and coding errors. Misclassified deaths occurred at home or non-maternity hospital wards; were postpartum, in early pregnancy or died undelivered. Advanced maternal age and being unemployed, were significantly associated with mortality, p<0.05. Direct maternal deaths predominated, 65%. Embolic events, hypertensive disorders, postpartum haemorrhage (PPH) and ectopic pregnancy, were the main causes of maternal deaths. The 3 main causes of near-miss were hypertensive conditions, PPH and ectopic pregnancies. Cesarean section was 3 times and preterm births 6 times as likely for pregnancies with Severe Maternal Outcomes (p<0.001). Newborn outcomes were best for uncomplicated pregnancies and worsened as the severity of maternal complications increased (p<0.001).

Conclusion: Multi-source methods will improve identification of maternal deaths, especially in settings of low incidence. The source of maternal death data should encompass mortuary records, non-obstetric hospital wards, hospital death registers, along with the Civil Status Registry and Health Department. Practitioners should be trained on maternal death certification and coding. Implementing nearmiss surveillance, and health promotion re warning signs in pregnancy would positively impact maternal deaths.