Clinical impact of application of risk assessment models (Padua Prediction Score and Improve Bleeding Score) on venous thromboembolism, major hemorrhage and health expenditure associated with pharmacologic VTE prophylaxis: a "real life" prospective and retrospective observational study on patients hospitalized in a Single Internal Medicine Unit (the STIME study).

International guidelines recommend the use of pharmacological prophylaxis in hospitalized medical patients at high risk of venous thromboembolism (VTE). The same international guidelines suggest the employment of standardized risk assessment models (RAMs) when evaluating the administration of pharmacological prophylaxis in acutely ill medical patients. The Padua Prediction Score and the Improve Bleeding Score have been indicated as the best available RAMs to predict thrombotic and haemorrhagic risk in hospitalized medical patients, but it is still unknown whether their combined use may lead to a significant reduction in thrombotic and haemorrhagic events. It is also unclear whether their extensive use can affect to some extent health expenditure associated with pharmacological VTE prophylaxis. The purpose of this single-centre, prospective and retrospective observational study is to investigate these unanswered questions. All patients admitted to our Internal Medicine Department between May 2015 and August 2015, i.e., before the introduction and extensive use of RAMs, were consecutively enrolled (retrospective group). Similarly, all patients admitted between November 2016 and February 2017-once RAMs clinical use became a consolidated practice-have also been consecutively recruited (prospective group). Consecutively, 203 patients were enrolled in the retrospective group and 210 patients were enrolled in the prospective group. Three events of major bleeding and one event of pulmonary embolism were observed in the prospective group; three events of major hemorrhage and two events of pulmonary embolism were observed in the retrospective group (p = not significant). A statistically significant decrease in pharmacological VTE prophylaxis among study groups was detected: 43.3% of prospective group patients and 56.7% of retrospective group patients received pharmacological prophylaxis (p = .028). Overall, 299 drug doses for VTE prophylaxis have been spared after RAMs introduction (p = .0001) and health expenditure decreased by 27.2% (i.e., 1.67 € saved for each single patient). In conclusion, the extensive use of RAMs in our population of hospitalized medical patients did not statistically affect VTE rate or incidence of major bleeding, but it resulted in a significant drop in health expenditure related with pharmacological prophylaxis. Awaiting new clinical trials, a broad use of RAMs may be a safe strategy for reducing health expenditure associated with VTE prophylaxis in hospitalized medical patients.