Evaluating the Ability of Caprini and Padua Scores to Predict Venous Thromboembolism in a Nationwide Study

Venous thromboembolism (VTE) is a preventable complication of hospitalization. VTE risk-assessment models (RAMs) including the Caprini and Padua RAMs quantify VTE risk based on demographic and clinical characteristics. Both RAMs have performed well in selected high-risk cohorts with relatively small sample sizes but few studies have evaluated the RAMs in large, unselected cohorts. We assessed the ability of both RAMs to predict VTE in a large, nationwide, diverse cohort of surgical and nonsurgical patients. We analyzed consecutive first hospital admissions of 1,252,460 unique patients to 1283 VA facilities nationwide between January 2016 and December 2021. Caprini and Padua scores were calculated using data from the VA Informatics and Computing Infrastructure, the electronic medical record repository of the entire VA system. We first assessed the ability of the two RAMs to predict VTE within 30, 60, and 90 days after admission. We then compared the predictive ability of the two RAMs in surgical and nonsurgical patients by comparing the areas under their respective receiver operating characteristic curves. We analyzed 1,252,460 hospitalized patients; 26.6% (n = 333,286) where surgical patients and 73.4% (n = 948,728) were nonsurgical. Caprini scores ranged from 0 to 28 (median, 4; interquartile range, 3-6); Padua scores ranged from 0 to 13 (median, 1; interquartile range, 1-3); higher scores were associated with higher VTE rates. 35,557 patients developed VTE within 90 days of admission (2.88%) (Table). The ability of both RAMs to predict VTE within 90 days of admission was low (area under the curve, 0.56 and 0.59, respectively). The predictive ability of both RAMs was similar within 30, 60, and 90 days of admission. Both RAMs performed marginally better in nonsurgical compared with surgical patients (Caprini areas under the curve: 0.59 vs 0.54; P = .001; Padua areas under the curve: 0.60 vs 0.56; P = .001) (Figure). In the largest validation study performed to date, increasing Caprini and Padua RAM scores are associated with increasing VTE rates. Both RAMs have low ability to predict VTE at 30, 60, or 90 days after admission in an unselected population of surgical and nonsurgical patients. Both RAMs are marginally better at predicting VTE risk in nonsurgical populations. Studies are needed to improve the predictive ability of existing VTE RAMs before they can be applied to the general hospital population.TableDistribution of VTE events within 30, 60, and 90 days after admission of 1,252,460 unique consecutive surgical and nonsurgical patients nationwide between 2016 and 2021Time from hospital admission, daysVTE typeTotal VTE eventsPE onlyPE with DVTDVT only0-308092314414,37225,6080-609801376417,99231,5570-9010,940411620,50135,557DVT, Deep vein thrombosis; PE, pulmonary embolism; VTE, venous thromboembolism. Open table in a new tab