To investigate the factors associated with acute kidney injury (AKI) in postoperative colorectal cancer (CRC) patients and develop a risk prediction model. The clinical data of 389 CRC patients were retrospectively analyzed. The patients were divided into AKI (n = 30) and non-AKI groups (n = 359) according to KDIGO diagnostic criteria. Demographic data, the presence of underlying diseases, perioperative conditions and related examination results were compared between the two groups. Binary logistic regression was used to analyze the independent risk factors for postoperative AKI, and a risk prediction model was established. And a verification group (94 patients) was used to verify the model. 30 patients (7.71%) with CRC had postoperative AKI. Binary logistic regression analysis showed that preoperative combined hypertension, preoperative anemia, inadequate intraoperative crystalloid infusion, low intraoperative minimum mean arterial pressure (MAP) and moderate to severe postoperative decline in hemoglobin (Hb) levels were independent risk factors. The risk prediction model developed was expressed as Logit P = -0.853 + 1.228 * preoperative combined hypertension + 1.275 *preoperative anemia -0.002 * intraoperative crystalloid infusion (ml) -0.091 * intraoperative minimum MAP (mmHg) + 1.482 * moderate to severe postoperative decline in Hb levels. In Hosmer-Lemeshow test, χ2 = 8.157, P = 0.718 showed that the fitting effect was good. The area under ROC curve was 0.776 (95% CI 0.682-0.871, P < 0.001), with a prediction threshold of 1.570, a sensitivity of 63.3% and a specificity of 88.9%. The sensitivity and specificity of the verification group were 65.8% and 86.1%. Preoperative combined hypertension, preoperative anemia, inadequate intraoperative crystalloid infusion, low intraoperative minimum MAP, and moderate to severe postoperative decline in Hb levels were independent risk factors for AKI development in CRC patients. The prediction model can effectively predict the occurrence of postoperative AKI in patients with CRC.
Read full abstract