Model for Predicting Risk of Postpartum Purulent-Inflammatory Complications after Cesarean Section: Cohort Retrospective Study

Background. Relevance of the study is related to the lack of a unified strategy for determining the risk of infectious complications of puerperium. Currently, the specialists use the collected data from studies documenting the main factors of infection in a population or cohorts of postpartum women. However, accuracy of these factors is not established and prognosis could not be personalized. The paper analyzes significant risk factors for the infectious complications associated with cesarean section (C-Section) and develops a model for their prognosis and an individual assessment of the risk of postpartum infection in order to take timely preventive measures.

Objective. To identify predictors of postpartum purulent-inflammatory complications after cesarean section, to evaluate their predictive value, and to develop a statistical model for determining the risk of their occurrence.

Methods. The cohort retrospective study (January 2019 to January 2022) was conducted in four obstetric health facilities of Krasnodar and was focused on analysis of medical records. Anamnestic, clinical and laboratory data of all women after cesarean section delivery were collected. Model: a patient diagnosed with any infection associated with cesarean section within 42 days after delivery — postoperative suture infection, endometritis, peritonitis, thrombophlebitis, sepsis. Infections were grouped to carry out a single risk assessment with an internal validation test and to develop a multifactor logistic regression model. All analyses were conducted using version R 3.2.3 (SPSS Inc., Chicago, IL) and Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD).

Results. Infection associated with cesarean section occurred in 2.50% of women (n = 548; 95% CI 2.6–3.5), suture disruption — in 0.59% (n = 129; 95%, CI 0.41–0.81), endometritis — 1.46% (n = 321; 95%, CI 1.15–1.94), peritonitis — 0.16% (n = 35; 95%, CI 0.11–0.20), thrombophlebitis 0.04% (n = 9; 95%, CI 0.01–0.05), sepsis 0.25% (n = 54; 95%, CI 0.15–0.35). The authors have identified 14 major variables with a high prognostic risk for the development of infectious complications of puerperium. The model differentiated women with and without purulent inflammatory complications of puerperium by internal validation (concordance index = 0.712, 95%, CI 0.672–0.755).

Conclusion. The developed model can be used to accurately predict the risk of infectious complications after abdominal delivery and to identify high-risk patients. It ensures a differentiated approach with a possibility to expand studies and strengthen antibiotic therapy, which promotes reduction of surgical site puerperal infection.