Development and validation of predictive models for SIRS and severe hemorrhage following percutaneous nephrolithotomy: the role of hydronephrosis and inter-correlation.

objectivPostoperative systemic inflammatory response syndrome (SIRS) and severe hemorrhage are two major complications threatening patient safety after percutaneous nephrolithotomy (PCNL). This study aimed to investigate their risk factors, develop and validate predictive models, and analyze the impact of hydronephrosis severity on prognosis, as well as the potential association between these two complications.

methodsClinical data from 708 patients with kidney stones who underwent PCNL were retrospectively analyzed. An additional 104 patients were selected as an external validation cohort. Independent risk factors for postoperative SIRS and severe hemorrhage were identified using univariate and multivariate logistic regression analyses, and clinical prediction models were constructed. Model performance was evaluated using the Hosmer-Lemeshow goodness-of-fit test, receiver operating characteristic (ROC) curve analysis, and decision curve analysis (DCA). Spearman’s correlation analysis was used to examine the correlation between SIRS and severe hemorrhage. Postoperative stone-free status was defined as residual fragments ≤ 4 mm, and severe hemorrhage was defined as the need for blood transfusion and/or interventional embolization.

resultsMultivariate analysis showed that prolonged operative time (OR = 1.016), positive preoperative urine culture (OR = 3.651), and residual stones (OR = 2.680) were independent risk factors for postoperative SIRS. For postoperative severe hemorrhage, the independent factors were prolonged operative time (OR = 1.014), positive preoperative urine culture (OR = 2.763), and moderate hydronephrosis (OR = 0.113). The predictive models for SIRS and severe hemorrhage, constructed based on these variables, both demonstrated good discrimination and calibration. In the development cohort, the area under the ROC curve (AUC) for the SIRS model and the hemorrhage model were 0.761 (95% CI: 0.659–0.863) and 0.807 (95% CI: 0.719–0.895), respectively. In the external validation cohort, the corresponding AUC values were 0.773 (95% CI: 0.586–0.961) and 0.800 (95% CI: 0.618–0.982). Subgroup analysis revealed that patients with moderate hydronephrosis had the lowest risk of severe postoperative hemorrhage (0.9%) and the highest stone clearance rate (92.1%). Furthermore, a significant positive correlation was observed between the occurrence of postoperative SIRS and severe hemorrhage (r = 0.176, P < 0.001).

conclusionThis study successfully developed and validated predictive models for SIRS and severe hemorrhage after PCNL, demonstrating moderate-to-good clinical discriminatory ability. Moderate hydronephrosis may be a protective factor against hemorrhage after PCNL. The occurrence of postoperative SIRS and severe hemorrhage are not independent events; they may share interrelated pathophysiological mechanisms, warranting clinical attention and further investigation.