Prognostic model of postpartum endometritis development

Abstract

We aimed to build a prognostic model to forecast development of postpartum endometritis based on the use of standard diagnostic algorithm. Additionally, we implied the use of fluorescent spectroscopy, logistic regression and ROC-analysis. The main cohort consisted of 75 women with postpartum endometritis and control group consisted of 40 women with uncomplicated course of the postpartum period. As a part of this study we have analysed the course of pregnancy, labour and postpartum period of both main and control cohorts. Moreover, we carried out an extended analysis of the clinical and laboratory data, including information obtained during instrumental examination of the patients. Furthermore, we determined 5 factors whose cumulative effect has the biggest impact on occurrence of postpartum endometritis among the main cohort of patients. Logistic regression enabled identification of these factors as such: extragenital pathology, TORCH-infections, colpitis, fluorescence intensity ≤ 0.845, a number of bed days.