Predictive Model for Systemic Inflammatory Response Syndrome after Percutaneous Nephrolithotomy

Abstract

Objective: To investigate the factors associated with systemic inflammatory response syndrome (SIRS) after percutaneous nephrolithotomy (PCNL) for upper urinary tract stones and establish a predictive model to prevent postoperative SIRS that correlates with PCNL. Methods: Patients who had undergone PCNL between June 2012 and December 2019 were retrospectively enrolled and categorized into groups according to their urosepsis status. Univariable and multivariate logistic regression analyses were performed to determine the predictive factors associated with SIRS after PCNL. Nomograms were generated using the predictors, and the discriminative ability was assessed using receiver operating characteristic (ROC) curve analyses. Results: A total of 262 PCNL patients were enrolled in this study, and SIRS occurred in 117 (44%) patients after PCNL. Multivariable logistic regression analysis suggested that there were three factors related to postoperative SIRS, namely stone size ≥ 3 cm (odds ratio [OR], 3.61; 95% confidence interval [CI], 2.08–6.26; p = 0.001), positive preoperative urine white blood cells (WBC) (OR, 3.1; 95% CI, 1.65–5.81; p = 0.028), and positive urine culture (OR, 2.89, 95% CI, 1.63–5.13; p = 0.017). According to the results of the logistic regression analysis, stone size ≥ 3 cm, positive preoperative urine WBC, and positive urine culture were used to develop the nomogram. A nomogram prediction model was established to calculate the cumulative probability of urosepsis after PCNL and displayed favorable fitting in the Hosmer–Lemeshow test (p=0.938). Internal validation of the nomogram showed that the area under the ROC curve was 0.702. Conclusion: Positive preoperative urine WBC, positive urine culture, and stone size ≥ 3 cm are the most significant predictors of SIRS after PCNL. The nomogram, which is based on independent risk factors, helps identify high-risk individuals and facilitates the early detection of SIRS after PCNL.

Diagnostic test accuracy in longitudinal study settings: Theoretical approaches with use cases from clinical practice

“…As shown in the epilepsy example, the sensitivity and specificity estimates of all three estimation levels using 'days' as time unit differed barely. Other diseases which are characterized by medium to long episode periods and medium to long disease-free intervals between episodes, such as SIRS 39 or depression [40][41][42] , could theoretically be assessed using any of the three time units. However, using 'minutes' as time unit significantly increases the number of observations; thus, the evaluation using an estimation tool could be slower due to the large number of observations, while also being more precise.…”