Introduction: The study aimed to construct and validate a risk prediction model for incidence of postoperative renal failure (PORF) following radical nephrectomy and nephroureterectomy.
Methods: The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database years 2005-2014 were used for the derivation cohort. A stepwise multivariate logistic regression analysis was conducted, and the final model was validated with an independent cohort from the ACS-NSQIP database years 2015-2017.
Results: In cohort of 14,519 patients, 296 (2.0%) developed PORF. The final 9-factor model included age, gender, diabetes, hypertension, BMI, preoperative creatinine, hematocrit, platelet count, and surgical approach. Model receiver-operator curve analysis provided a C-statistic of 0.79 (0.77, 0.82; p < 0.001), and overall calibration testing R2 was 0.99. Model performance in the validation cohort provided a C-statistic of 0.79 (0.76, 0.81; p < 0.001).
Conclusion: PORF is a known risk factor for chronic kidney disease and cardiovascular morbidity, and is a common occurrence after unilateral kidney removal. The authors propose a robust and validated risk prediction model to aid in identification of high-risk patients and optimization of perioperative care.
Keywords: Acute kidney injury; Kidney neoplasms; Logistic models; Minimally invasive surgical procedures; Nephrectomy.
© 2021 S. Karger AG, Basel.