What constitutes high risk for venous thromboembolism? Determining an appropriate threshold to initiate prophylaxis

Guidelines recommend pharmacological venous thromboembolism (VTE) prophylaxis only for high-risk patients, but the probability of VTE that is considered "high risk" is not specified.

We compared cost-effectiveness analysis (CEA) with alternative approaches to define an appropriate probability threshold for initiating prophylaxis in medical inpatients.

Our analytic sample included 47 889 adults admitted to 10 Cleveland Clinic hospitals from October 2017 to January 2020, as well as 214 hospitalists and internal medicine residents who were surveyed from June to November 2023. We compared decision analysis with 4 alternative approaches to determine a probability threshold: the Youden Index, deriving a probability from a widely used point-based model, asking physicians' opinions, and observing physician behavior. To assess each unique approach's threshold, we applied the Cleveland Clinic VTE Model to calculate the percentage of patients whose predicted risk exceeded that threshold, and we modeled hypothetical adverse events (VTE plus major bleeding) if that threshold were used to guide prescribing.

Most thresholds ranged from 0.9% to 1.5%, inversely corresponding to 19% to 35% of patients being high risk. The CEA threshold of 1% would result in 15.9 adverse events per 1000 individuals. The lowest threshold, 0.3%, produced the fewest adverse events but was highly inefficient. Compared with CEA, the 0.3% threshold required treating more than 1000 patients to prevent 1 adverse event, and quadrupled the prophylaxis rate (100% vs 24.9%).

To improve the efficiency of prophylaxis and standardize care, guidelines should explicitly quantify the high risk of VTE using a standardized probability threshold range.