Rate and predictors of thromboprophylaxis in internal medicine wards: Results from the AURELIO study

Abstract

Background

Randomized controlled trials suggest that prophylactic doses of anticoagulants effectively prevent venous thromboembolism (VTE) in hospitalized medical patients with high thromboembolic risk. However, no prospective studies exist regarding the real-world prevalence of prophylactic anticoagulant use. This prospective study aimed to determine the rate and predictors of thromboprophylaxis in an unselected population of patients hospitalized in medical departments.

Methods

We conducted a multicenter prospective observational study (AURELIO – rAte of venous thrombosis in acutely iLl patIents hOspitalized) to assess the rate of deep vein thrombosis (DVT) in unselected acutely ill patients hospitalized in medical wards using compression ultrasound (CUS) at admission and discharge. Additionally, we evaluated the rate of pharmacological thromboprophylaxis administration in this population and analyzed the thrombotic risk by assessing RAMs (Risk Assessment Models) such as the IMPROVE-VTE and PADUA scores following the clinician's decision to administer thromboprophylaxis. Patients with IMPROVE-VTE scores ≥3 and/or PADUA scores ≥4 were classified as high thrombotic risk; those with IMPROVE-VTE scores <3 and/or PADUA scores <4 were classified as low risk.

Results

We recruited 2371 patients (1233 males [52 %] and 1138 females [48 %]; mean age 72 ± 16 years). The median length of hospitalization was 13 ± 12 days. Overall, 442/2371 (18.6 %) patients received prophylactic parenteral anticoagulants (subcutaneous low weight molecular heparin or fondaparinux once daily) at admission. Assessing the thrombotic risk of the population recruited 1016 (42.9 %) patients were classified as high risk and 1354 (57.1 %) were low risk. Among high-risk patients, 339/1016 (33.4 %) received anticoagulant prophylaxis compared to 103/1354 (7.6 %) low-risk patients. During hospitalization, 9 patients developed DVT, comprising 7 asymptomatic and 2 symptomatic cases of proximal DVT. Of these, 3 patients were on anticoagulant prophylaxis, while 6 were not. Among the high-risk population, 7 out of 1016 patients (0.7 %) experienced proximal DVT during hospitalization, with 2 out of these 7 (28 %) receiving anticoagulant thromboprophylaxis. In the low-risk population, 2 out of 1354 patients (0.2 %) developed DVT, with 1 out of these 2 (50 %) receiving anticoagulant thromboprophylaxis. Age, heart or respiratory failure, pneumonia, active neoplasia, previous VTE, reduced mobility, and absence of kidney failure were more frequent in patients receiving prophylaxis. Multivariable logistic regression identified age (RR 1.010; CI 95 % 1002–1019; p = 0.015), heart/respiratory failure (RR 1.609; CI 95 % 1248–2075; p < 0.0001), active neoplasia (RR 2.041; CI 95 % 1222–2141; p < 0.0001), pneumonia (RR 1.618; CI 95 % 1557–2676; p < 0.0001), previous VTE (RR 1.954; CI 95 % 1222–3125; p < 0.0001), and reduced mobility (RR 4.674; CI 95 % 3700–5905; p < 0.0001) as independent predictors of thromboprophylaxis.

Conclusions

This study, conducted without pre-established thromboembolic risk scores, offers a comprehensive view of venous thromboembolism prophylaxis in medical patients with acute conditions hospitalized in internal medicine departments. It reveals that advanced age, heart or respiratory failure, active cancer, pneumonia, previous VTE, and reduced mobility are predictors that may influence the decision to administer thromboprophylaxis in these patients.