Influence of Perceived Risk Factors for Bleeding on Venous Thromboembolism Prophylaxis Practices in Acutely Ill Medical Patients: Findings from IMPROVE.

Abstract

Background Acutely ill medical patients at risk for venous thromboembolism (VTE) should receive VTE prophylaxis. However, factors perceived by physicians to increase patients' risk of bleeding may influence VTE prophylaxis practices. In this analysis from The International Medical Prevention Registry on Venous Thromboembolism (IMPROVE), we examined whether perceived risk factors for bleeding had a significant influence on physicians' prescribing of in-hospital prophylaxis in acutely ill medical patients.Methods Patients aged ≥18 years and hospitalized ≥3 days with an acute medical illness have been enrolled consecutively since July 2002. Risk factors considered were: severe renal failure, known bleeding disorder, hemorrhagic stroke, thrombocytopenia, bacterial endocarditis, active gastroduodenal ulcer, NSAID use, hepatic failure, age, immobility and alcoholism. Factors associated with different prescription rates of prophylaxis compared with rates in patients without these factors were identified and included in a multiple logistic regression model (significance at p<0.05).Results Up to 31 March 2005, 6946 patients were enrolled in 49 hospitals in 12 countries. Pharmacologic prophylaxis was received by 42%, 25%, 16%, and 14% of patients with a platelet count at admission >100, 50–100, 20–50 and <20x109/L, respectively (p<0.0001), and 43%, 39%, 30% and 32% of patients with none, 1, 2 and 3 risk factors for bleeding (p<0.0001). Factors independently associated with a lower/higher prescription rate of heparin-based prophylaxis compared with the risk in patients without these factors are shown in the Table.Conclusions The likelihood that hospitalized acutely ill medical patients receive in-hospital pharmacologic prophylaxis decreases as their platelet count at admission decreases, or their cumulative number of perceived risk factors for bleeding increases. Further studies are needed to determine whether the changes in prophylaxis practices observed in this study are justified.TableFactors Independently Associated with a Higher/lower Rate of Heparin-based VTE ProphylaxisFactorOdds Ratio95% Confidence IntervalAge (per 10-year increase)1.191.66–1.22Immobility (per 10-day increase)1.031.02–1.05Alcoholism0.620.46–0.83Thrombocytopenia0.600.48–0.74Active duodenal ulcer0.360.26–0.52Hepatic failure0.340.21–0.54