Risk prediction model for pegaspargase-associated severe coagulopathy and blood product costs: a single-center retrospective cohort study.

INTRODUCTION

Pegaspargase-induced severe coagulopathy (SC) is a common complication in pediatric acute lymphoblastic leukemia (ALL) treatment, yet reliable predictive tools are limited and its economic burden remains unclear. This study developed and internally validated a clinical prediction nomogram for SC, concurrently created a web-based calculator, and quantified the impact of coagulopathy on the cost of blood products.

METHODS

Using retrospective data from 356 pediatric patients treated according to the Chinese Children's Leukemia Group Acute Lymphoblastic Leukemia 2018 protocol at Hebei Children's Hospital between January 2020 and August 2025, SC was defined as grade 4 fibrinogen reduction or grade ≥3 prolongation of the activated partial thromboplastin time. A nomogram was developed via multivariate logistic regression, and cost drivers were analyzed by multiple linear regression.

RESULTS

Hypoalbuminemia, elevated bilirubin, and elevated procalcitonin were independent predictors of SC. The nomogram achieved an area under the curve of 0.819 (95% CI: 0.773-0.864), and decision curve analysis indicated favorable clinical utility. Blood product cost was primarily driven by patient age, product type, and transfusion frequency (adjusted R 2 = 0.677).

DISCUSSION: This study developed an accurate predictive model for SC and clarified the associated economic burden, providing evidence to support clinical vigilance and inform healthcare resource planning.