SADA-16 LIMITED UTILITY OF CAPRINI SCORE FOR STRATIFYING VENOUS THROMBOEMBOLISM RISK IN SPINE METASTASIS PATIENTS
Seeley Yoo, Tara Dalton, Ramzy Ahmed, Ezinne Oguguo, Nicole Rivera, Brandon Lee, Nina Min, Kerri-Anne Crowell, Toyosi Onwuemene, Trupti Trivedi, Melissa Erickson, C Rory GoodwinAbstract
Background
Metastatic spine disease (MSD) patients are at elevated risk for perioperative venous thromboembolism (VTE), yet no standardized risk stratification framework exists to guide prophylaxis in this population. The Caprini score can effectively predict VTE risk in other surgical cohorts; however, its applicability to MSD patients remains uncertain. Therefore, this study assessed the performance of the Caprini score in predicting postoperative VTE among MSD patients.
Methods
A retrospective cohort study was conducted including adult MSD patients undergoing surgery from September 2017 to December 2024. Caprini scores were calculated using documented clinical characteristics. Postoperative deep vein thrombosis (DVT) and pulmonary embolism (PE) occurring within 90 days of surgery were identified through imaging reports and chart review. Comparative analyses evaluated the relationship between Caprini score and VTE outcomes, and receiver operating characteristic (ROC) analyses assessed predictive performance.
Results
A total of 336 patients were included. Overall VTE incidence was 8.0% (2.1% DVT, 6.5% PE). The cohort’s average Caprini score was 9.0±2.9, with 98% of patients categorized as high or highest risk. Caprini score was not significantly associated with DVT (p = 0.33), PE (p = 0.45), or overall VTE (p = 0.32) occurrence and demonstrated poor discriminatory ability for predicting VTE on ROC analysis (AUC 0.56, 95% CI [0.44, 0.58]). Using a threshold score of 10, Caprini score predicted VTE risk with a sensitivity of 44.4%, specificity of 66.3%, low positive predictive value (10.3%), and high negative predictive value (93.2%). VTE events occurred throughout the postoperative period, with a notable proportion diagnosed after discharge (∼50% of PEs and 72% of DVTs).
Conclusions
The Caprini score showed limited ability to stratify postoperative VTE risk in MSD patients, likely reflecting the uniformly elevated baseline risk within this population. Thus, improved predictive models incorporating disease-specific variables and biomarkers may be necessary to guide thromboprophylaxis strategies in this high-risk population.