Predicting severe maternal outcomes in a network of sentinel sites in Latin-American countries

Abstract

Objective: This study aimed to determine incidences of potentially life-threatening conditions (PLTC), maternal near misses (MNM), and maternal deaths (MD) in women who gave birth in participating facilities, and to determine the probability that a pregnancy involving a PLTC would evolve into an MNM and/or an MD. Methods: This was a multicentric observational study implemented on a maternal network from August 2018 to May 2019 in five Latin-American countries. We summarized categorical variables as frequencies and continuous variables with median, interquartile range, and standard deviations. Positive and negative likelihood ratios were calculated and multivariate predictive models were built. Results: There were 33 901 deliveries and miscarriages, of which 8.0% had at least one PLTC and 0.6% had an MNM. Hypertensive disorder was the most frequent condition to evolve into a severe maternal outcome. Conclusion: Identifying PLTC can help to prevent MNM and MD. The inclusion of these predictors in a real-time data registration system like the Perinatal Informatic System could work as a surveillance tool for early detection, leading to a reduction in the rate of worsening conditions.