Analysis of influencing factors for postoperative hypoalbuminemia in colorectal cancer patients and construction of a prediction model

[Objectives] To analyze the influencing factors for postoperative hypoalbuminemia in colorectal cancer patients, construct a nomogram prediction model based on these factors, and validate the model’s performance. [Methods] A total of 200 colorectal cancer patients admitted to Nanjing Hospital of Traditional Chinese Medicine affiliated to Nanjing University of Chinese Medicine between January 2023 and March 2024 were randomly selected and divided into a modeling group and a validation group at a ratio of 7∶3. Patients in the modeling group were further categorized into non-hypoalbuminemia and hypoalbuminemia groups based on postoperative occurrence of hypoalbuminemia. Clinical data were collected to identify influencing factors. A nomogram model for predicting postoperative hypoalbuminemia was developed. The model’s predictive performance was assessed using the receiver operating characteristic (ROC) curve and the Hosmer-Lemeshow goodness-of-fit test. Clinical utility was evaluated using decision curve analysis (DCA) and clinical impact curve (CIC) analysis. [Results] Among the 200 patients, 38 (19%) developed postoperative hypoalbuminemia. No significant differences were found in clinical data between the modeling and validation groups (P>0.05). Univariate analysis showed significant differences (P<0.05) between the non-hypoalbuminemia and hypoalbuminemia groups regarding smoking history, tumor diameter, preoperative neoadjuvant chemoradiotherapy, preoperative skinfold thickness, preoperative prealbumin level, preoperative albumin (ALB) level, preoperative total protein (TP) level, preoperative hemoglobin level, and intraoperative blood loss. Multivariate Logistic regression analysis identified smoking history, tumor diameter, preoperative ALB level, preoperative TP level, and intraoperative blood loss as influencing factors for postoperative hypoalbuminemia in colorectal cancer patients (P<0.05). The area under the ROC curve (AUC) for predicting hypoalbuminemia in the modeling group was 0.958 (95% CI: 0.920-0.996). For the validation group, the AUC was 0.903 (95% CI: 0.806-0.999). The Hosmer-Lemeshow test yielded a χ2 value of 1.664 (P=0.990). DCA and CIC results indicated that clinical intervention based on the model would yield net benefit. [Conclusion] The occurrence of postoperative hypoalbuminemia in colorectal cancer patients is associated with smoking history, tumor diameter, intraoperative blood loss, and preoperative serum protein levels. The constructed nomogram prediction model accurately predicts the risk of postoperative hypoalbuminemia, providing a reference for its early prevention and management.