Infectious and inflammatory diseases of the postpartum period: new approaches to diagnosis and treatment, predicting the risk of septic complications

Introduction. Infectious and inflammatory diseases (IIDs) represent a serious problem in modern obstetrics. Aim: to improve methods for postpartum IIDs diagnosis, prognosis and treatment. Materials and Methods. The study was conducted with 362 puerperants. At stage 1, there were retrospectively examined 199 patients with postpartum endometritis stratified as follows: IA (n = 113) – delivered by caesarean section (CS), IB (n = 86) – delivered by natural childbirth. At stage 2, a prospective, controlled study was conducted enrolling 163 puerperants. Group IIA (n = 124) consisted of patients with endometritis after CS, divided into 2 subgroups: subgroup IIA1 (n = 63) received antibacterial therapy and intrauterine sorbent VNIITU-1PVP, subgroup IIA2 (n = 61) – antibacterial treatment alone. Group IIB consisted of 39 puerperants who had a critical obstetric condition (COC), divided into 2 subgroups: IIB1 (n = 18) – patients with obstetric sepsis, IIB2 (n = 21) – without septic complications. Anamnestic data (complaints, disease history, characteristics of delivery, course of the postpartum period, timing of endometritis diagnosis), laboratory assay data (complete blood count, biochemical blood test, microbiological examination of uterine cultures, immunological study of IL-1β, IL-10, TNF-α cytokines in endometrial aspirate), and instrumental methods (ultrasound examination of pelvic organs, hysteroscopy, measurement of central venous pressure, infrared spectrometry of crushed carbon sorbent VNIITU-1PVP) were assessed. The APACHE, SOFA, NEWS2, AVPU scales were used to assess condition of post-COC puerperant women; χ 2 -test and Mann–Whitney test were applied to qualitative and quantitative variables to determine p-value. To develop a prognostic model for assessing obstetric sepsis in post-COC patients, the method of multiple logistic regression was used with step-by-step exclusion of variables until the minimum value of the Akaike criterion was reached. The obtained predictive formula was subjected to ROC analysis. Data calculations and graphical visualization were carried out using special libraries of the R language. Results. In group IA vs. group IB patients, anemia (p = 0.004), leukocytosis (p < 0.001), a left shift in the leukocyte formula via leukocytosis (p < 0.001), hypoproteinemia (p< 0.001) were significantly more common, with Enterococcus faecalis (p = 0.02) and Enterococcus faecium (p = 0.02) more often cultured from the uterine cavity. The risk ratio of performing endometritis-related uterine extirpation in group IA was 5.33 (95 % CI = 1.43–19.78) compared with group IB. The use of the molded sorbent VNIITU-1PVP in subgroup IIA1 allowed 87.3 % to avoid microbial growth in the uterine cavity. The concentration of pro-inflammatory cytokines such as interleukin-1β (IL-1β) and tumor necrosis factor-alpha (TNF-α) in the uterine cavity in subgroup IIA1 was also significantly lower than that of in subgroup IIA2 – by 4.0 and 3.2 times, respectively (p < 0.05). In subgroup IIA1, organ-preserving surgery was performed in 23 cases related to failure of uterine sutures. While analyzing the data of group IIB patients, the following cut-off points were found: international normalized ratio – 1.13, central venous pressure – 6 mm Hg, aspartate aminotransferase level – 45 IU/L, and a mathematical model for post-COC sepsis development was constructed. The effectiveness of the developed prognostic model for obstetric sepsis in post-COC puerperant patients had 94.5 % sensitivity and 90.5 % specificity. Conclusion. The use of the molded sorbent VNIITU-1PVP reduces a progression risk for uterine inflammatory process. Using a prognostic risk model for obstetric sepsis allows to timely identify this complication.