Overuse and underuse of thromboprophylaxis in medical inpatients.

Background Thromboprophylaxis (TPX) prescription is recommended in medical inpatients categorized as high risk of venous thromboembolism (VTE) by validated risk assessment models (RAMs), but how various RAMs differ in categorizing patients in risk groups, and whether the choice of RAM influences estimates of appropriate TPX use is unknown. Objectives To determine the proportion of medical inpatients categorized as high or low risk according to validated RAMs, and to investigate the appropriateness of TPX prescription. Methods This is a prospective cohort study of acutely ill medical inpatients from three Swiss university hospitals. Participants were categorized as high or low risk of VTE by validated RAMs (i.e., the Padua, IMPROVE, simplified, and original Geneva score). We assessed prescription of any TPX at baseline. We considered TPX prescription in high-risk and no TPX prescription in low-risk patients as appropriate. Results Among 1352 medical inpatients, the proportion categorized as high risk ranged from 29.8% with the IMPROVE to 66.1% with the original Geneva score. Overall, 24.6% were consistently categorized as high risk, and 26.3% as low risk by all four RAMs. Depending on the RAM used, TPX prescription was appropriate in 58.7-63.3% of high-risk (i.e., 36.7-41.3% underuse) and 52.4-62.8% of low-risk patients (i.e., 37.2-47.6% overuse). Conclusion The proportion of medical inpatients considered as high or low VTE risk varied widely according to different RAMs. Only half of patients were consistently categorized in the same risk group by all RAMs. While TPX remains underused in high-risk patients, overuse in low-risk patients is even more pronounced