Development of clinical risk-prediction models for uterine atony following vaginal and cesarean delivery. (August 2022)

• Record Type: Journal Article
• Authors: Ende, H.B.; French, B.; Shi, Y.; Kertai, M.D.; Osmundson, S.S.; Bauchat, J.R.
• Is Part Of: International journal of obstetric anesthesia. Volume 51(2022)
• Publication Date: 2022-08
• Subjects: Logistic regression -- Maternal morbidity -- Postpartum hemorrhage -- Risk prediction -- Uterine atony; Obstetrics -- Periodicals; Anesthesia -- Periodicals; Anesthésie en obstétrique -- Périodiques; Anesthesia; Obstetrics; Electronic journals; Periodicals; 617.9682
• DOI: 10.1016/j.ijoa.2022.103550
• Languages: English
• ISSNs: 0959-289X

Abstract

Highlights: Uterine atony is the most common cause of postpartum hemorrhage. Accurately predicting uterine atony may decrease associated morbidity. We developed risk prediction models to quantify patients' risk of atony. Antepartum model had moderate discriminatory ability, C-statistic 0.61 (0.60–0.62) Intrapartum model had improved discriminatory ability, C-statistic 0.68 (0.67–0.69) Abstract: Background: Uterine atony is the most common cause of postpartum hemorrhage and is associated with substantial morbidity. Prospectively identifying women at increased risk of atony may reduce the incidence of subsequent adverse events. We sought to develop and evaluate clinical risk-prediction models for uterine atony following vaginal and cesarean delivery, using prespecified risk factors identified from systematic review. Methods: Using retrospective data from vaginal and cesarean deliveries occurring at a single institution between 2010 and 2019, antepartum and intrapartum risk-prediction models for uterine atony, defined by supplementary uterotonic administration in addition to prophylactic oxytocin infusion, were developed using logistic regression. The C-statistic quantified the ability of the model to discriminate between cases and controls. Results: Data were available for 4773 atony cases and 23 933 controls. The antepartum model included 20 risk factors and exhibited moderate discriminatory ability (C-statistic 0.61, 95% confidence interval 0.60 to 0.62). The intrapartum model included 27 risk factors and showed improved discriminatory ability (C-statistic 0.68, 95% confidence interval 0.67 to 0.69). Conclusions: We identified antepartum and intrapartum risk-prediction models to quantify patients' risk of uterine atony. Models performed similarly for all delivery modes, races, and ethnic groups. Future work should further improve these models through inclusion of more comprehensive prediction data.