[Analysis of risk factors and construction of a predictive model for perioperative urosepsis]

Urosepsis is a common and severe perioperative complication in patients with upper urinary tract calculi. This study aims to analyze the risk factors for perioperative urosepsis in patients with upper urinary tract calculi complicated by pyonephrosis and to construct a predictive model.

Clinical data of patients with upper urinary tract calculi complicated by pyonephrosis who underwent surgical treatment at Xiangya Hospital, Central South University, between January 1, 2019 and September 30, 2023 were retrospectively collected. According to whether patients developed urosepsis, they were divided into a urosepsis group and a non‑urosepsis group. Baseline characteristics, preoperative indicators, and intraoperative and postoperative monitoring parameters were compared between the 2 groups. Least absolute shrinkage and selection operator (LASSO) regression combined with multivariate logistic regression analysis was used to identify risk factors. A nomogram prediction model for perioperative progression from pyonephrosis to urosepsis was established, and its predictive performance was evaluated using receiver operating characteristic (ROC) curve analysis.

A total of 221 patients with upper urinary tract calculi complicated by pyonephrosis who underwent surgery were included, among whom 45 (20.4%) patients were diagnosed with urosepsis and 176 (79.6%) patients were classified as non‑urosepsis. Significant differences were observed between the urosepsis group and the non‑urosepsis group in age, American Society of Anesthesiologists (ASA) classification, highest perioperative body temperature, preoperative laboratory indicators, comorbidities (including diabetes mellitus, cardiovascular and cerebrovascular diseases, and autoimmune diseases), staged surgery, intraoperative fluid infusion volume, operative time, stone size, and length of intensive care unit stay (all P<0.05). Logistic regression analysis showed that elevated highest perioperative body temperature (OR=2.103, 95% CI 1.231 to 3.751, P<0.05) and increased procalcitonin (PCT) levels (OR=1.097, 95% CI 1.058 to 1.152, P<0.05) were independent risk factors for progression from pyonephrosis to urosepsis. Based on eight variables selected by LASSO regression analysis, including sex, age, highest perioperative body temperature, preoperative PCT level, concomitant cardiovascular and cerebrovascular diseases, concomitant hepatic insufficiency, maximum stone diameter >2 cm, and operation time >2 hours, a nomogram prediction model was constructed. The calibration curve demonstrated good agreement between predicted and observed values. The area under the ROC curve was 0.897, indicating that the model had good clinical predictive value.

High body temperature and elevated PCT levels are independent risk factors for perioperative urosepsis in patients with upper urinary tract calculi complicated by pyonephrosis. The predictive model constructed based on these factors may provide a preliminary clinical reference for the prevention and management of perioperative urosepsis.