Background: In hospitalized medical patients, the venous thromboembolism (VTE) risk is notable. Nevertheless, the available assessment model (TPF) is generally underused. In this work, we propose an ex novo risk assessment model based on the elaboration of the clinical data exhibited by the VET patients. Differently from previous studies, the proposed approach does not exploit pre-established models, resulting in a more valid and easy-to-use score.
Methods: We performed a double case-control observational study. For each case of VTE, we enrolled two consecutive patients without VTE of equal sex and age group (18-50, 50-55, 55-60, 60-65, 65-70, 70-75, 75-80, >80 years). The study involved both the EM and the IM Departments of 23 hospitals and universities in Lazio and Umbria (Italy).
Results: We analyzed the data of 1215 patients, 409 with VTE (50% - deep venous thrombosis [DVT], 9.9% - pulmonary embolism [PE], 40.1% - PE+DVT) and 806 case-control. 365 patients (30%) were in charge to the EM department, while 850 patients (70%) to the IM one. The VET risk factors with more statistical significance (P<0.01) are: previous VTE, active cancer, known thrombophilic condition, immobilization, chronic venous insufficiency, hyperhomocysteinemia, central venous catheter, recent hospitalization. Obesity, recent surgery, family history of VTE, hormone therapy and treatment with drugs that stimulate hematopoiesis are resulted at intermediate statistical significance (P<0.05 but >0.01). A multiple logistic regression was used with robust standard errors and forward selection of the candidate variables using the Bayesian information criterion. A new score is developed, the "TEVere Score", which shows a higher specificity and sensitivity (respectively 43.3 and 87.5, with accuracy 72.1) compared with the Padua, the Kuscer and the Chopard Score. TEVere Score also exhibits a greater predictive validity for thromboembolism risk (AUROC 0.7266; 95% CI: 0.71 to 0.73) than the Kuscer Score (AUROC 0.6891; 95% CI: 0.67 to 0.70) (P=0.0093).
Conclusions: The TEVere Score has proven to exhibit a higher accuracy than the other scores commonly used in clinical practice to stratify the thromboembolism risk.