Development and Validation of a Nomogram for Predicting High-Risk Hidden Blood Loss After Total Knee Arthroplasty

Hidden blood loss (HBL) accounts for 49% of total postoperative blood loss after total knee arthroplasty (TKA), predisposing patients to anemia, infection, and delayed functional recovery. Despite partial elucidation of its pathophysiology, standardized predictive models for HBL remain lacking, rendering current clinical management reliant on subjective judgment and precluding individualized risk stratification. This study aimed to identify predictors of high-risk HBL after TKA and develop a validated visual prediction model integrating multiple clinical parameters.

Clinical data from 400 TKA patients were analyzed. The dataset was randomly divided into a training group (70%) and a validation group (30%). Univariate analysis initially screened potential predictors, followed by least absolute shrinkage and selection operator (LASSO) regression for dimensionality reduction and key feature extraction. Multivariate logistic regression identified independent predictors of HBL. A nomogram was constructed and validated through receiver operating characteristic curves, calibration curves, and decision curve analysis to assess discrimination, calibration, and clinical utility.

Univariate analyses identified 18 significant variables associated with high-risk HBL (P < 0.05). A least absolute shrinkage and selection operator regression refined these to eight candidate predictors. Multivariate analyses confirmed weight, preoperative hemoglobin, postoperative white blood cell count, and postoperative albumin as independent predictors of high-risk HBL after TKA. The nomogram demonstrated the area under the curve of 0.858 (95% confidence interval = 0.812 to 0.904) and 0.818 (95% confidence interval = 0.719 to 0.917) in the training and validation groups, respectively, indicating good discrimination. Calibration curves demonstrated high agreement between predicted and observed probabilities, supported by a nonsignificant Hosmer-Lemeshow test (P > 0.05). Decision curve analysis confirmed net clinical benefit across low-to-moderate risk thresholds.

The nomogram incorporating weight, preoperative hemoglobin, postoperative white blood cell count, and postoperative albumin demonstrates excellent discrimination, calibration, and clinical utility, enabling early identification of high-risk HBL patients for personalized postoperative management of TKA.