Prediction models show poor accuracy in stratifying VTE risk in patients

Predicting the risk of venous thromboembolism (VTE) among hospitalized medical patients remains a challenge, what with the poor prognostic performance of Padua, IMPROVE*, and NICE** risk assessment models (RAMs) in a recent study.

A total of 1,273 patients were included in this multicentre prospective observational study conducted in 15 Spanish hospitals. They were monitored for symptomatic VTE over a 90-day follow-up period.

The investigators assessed the discriminative performance using time-to-event analyses and competing risk models accounting for non-VTE–related mortality. They also assessed predictive accuracy by calculating sensitivity, specificity, and area under the receiver operating characteristic curve (AUROC).

The 90-day cumulative VTE incidence among patients was 1.0 percent. High-risk patients showed no significantly increased VTE risk relative to low-risk patients according to Padua (adjusted subhazard ratio [aSHR], 5.71, 95 percent confidence interval [CI], 0.70–46.86), IMPROVE (aSHR, 3.72, 95 percent CI, 1.00–13.87), and NICE RAM (aSHR, 0.97, 95 percent CI, 0.30–3.18).

Although Padua exhibited the highest sensitivity (92.3 percent, 95 percent CI, 62.1–99.6), its specificity was the lowest (32.3 percent, 95 percent CI, 29.7–35.0). On the other hand, IMPROVE showed the highest specificity (52.9 percent, 95 percent CI, 50.1–55.7) but had moderate sensitivity (76.9 percent, 95 percent CI, 46.0–93.8).

All RAMs demonstrated suboptimal discriminative performance (Padua: AUROC, 62.3 percent; IMPROVE: AUROC, 64.9 percent; NICE: AUROC, 50.1 percent).

“Padua, IMPROVE, and NICE RAMs demonstrated poor predictive accuracy in stratifying venous thromboembolism risk among hospitalized medical patients,” the investigators said. “These findings underscore the need for more precise, dynamic RAMs tailored to real-world clinical settings to enhance thromboprophylaxis and patient outcomes.”

*International Medical Prevention Registry on Venous Thromboembolism 

**National Institute for Health and Care Excellence