Agreement evaluation in prophylaxis against venous thromboembolism (VTE) in medical patients at a University Hospital, Bogotá, D.C. 2007-2008

Abstract

Introduction: some strategies or algorithms have been proposed in order to improve VTE risk evaluation and prophylaxis. Our main purpose was to evaluate agreement between VTE pharmacological prophylaxis indicated by physician in a real and hypothetic scenario with prophylaxis suggested by a computational program (Decision Matrix). Additionally, we explore agreement between three algorithms for VTE risk evaluation and the decision matrix program (DM).

Patients and methods: concordance study. Consecutive patients older than 18 years old, hospitalized for = 24 hours in a Universitary hospital with acute medical illnesses, were included. Exclusion criteria included therapeutic anticoagulation, active bleeding, indication for critical care unit and suspicious of VTE.

Results: in all, 1703 patients were eligible, out of which 896 (52,6%) were included. The average age was 60,6 years. Decision Matrix program indicated prophylaxis in 95,8%. Patient received unnecessary prophylaxis in 52% and 22% did not receive any at all when it was indicated. Fifty seven patients without prophylaxis in a hypothetic scenario received it in the real scenario Kappa index between clinical decision and decision matrix was deficient (0,08). In low risk patients, there was disagreement between Menajobsky, Cohen and Milani algorithm"(TM)s and the decision matrix program.

Conclusions: efforts are required to increase prophylaxis when it is indicated but simultaneously decreasing inappropriate prophylaxis.

Menajobsky, Milani and Cohen algorithms are extremely sensible and little specifics.