Abstract

Background: Hospitalized medical patients are at significant risk of venous thromboembolism (VTE). Although evidence-based guidelines exist which provide recommendations for thromboprophylaxis in hospitalized medical patients, the optimum regimen for prophylaxis is not clear. We have therefore created a model, based on established literature, which examines the 2-year clinical outcomes following no prophylaxis, thromboprophylaxis with unfractionated heparin (UFH), or thromboprophylaxis with low-molecular-weight heparin (LMWH) in medical patients at risk of VTE.

Methods: A decision-analytic model was developed that replicates and extends an existing, published VTE model (

McGarry et al.
Am J Manag Care.
2004
;
10
:
632
–42
) from 30 days to 2 years. Hypothetical cohorts of 10,000 medically ill patients at risk of VTE (according to MEDENOX criteria) were assembled, using a Markov chain model with resampling, to compare the rates of primary VTE events and related outcomes at 90 days and VTE complications and recurrent events at 2 years. Clinical outcomes were estimated from published clinical trial and observational data, and compared between three interventions, namely no prophylaxis, UFH, and the LMWH enoxaparin. Outcomes included in the analysis were clinical or venographically detected primary VTE, major and minor bleeds, asymptomatic or symptomatic heparin-induced thrombocytopenia, and death within the first 90 days, as well as VTE recurrence, post-thrombotic syndrome, pulmonary hypertension, and death within 2 years. Sensitivity and threshold analyses were performed to test the general applicability of the model. The simulation model was run using TreeAge software (Williamstown, USA).

Results: VTE rates and death were the lowest in the enoxaparin prophylaxis cohort, followed by the UFH and no prophylaxis cohorts respectively (Table 1). Adverse events were lowest in the no prophylaxis group, followed by the enoxaparin group and the UFH group (Table 1).

Conclusion: In this Markov model, based on robust data from clinical trials and observational studies, prophylaxis with enoxaparin reduced VTE occurrence and mortality over two years when compared with no prophylaxis or UFH prophylaxis in hospitalized medical patients at risk of VTE. Enoxaparin was also associated with a reduced incidence of adverse events when compared with UFH.

Table 1.

Two-year outcomes in simulated cohorts of hospitalized medical patients at risk of VTE

Outcome (n)Enoxaparin (n=10,000)Unfractionated heparin (n=10,000)No prophylaxis (n=10,000)
VTE at 2 years 683 791 1787 
DVT 545 633 1426 
PE 138 158 361 
Death 1573 1600 1745 
Adverse events at 90 days 364 725 314 
Minor bleed 285 510 244 
Major bleed 65 116 55 
Asymptomatic HIT 45 
Symptomatic HIT 54 
Outcome (n)Enoxaparin (n=10,000)Unfractionated heparin (n=10,000)No prophylaxis (n=10,000)
VTE at 2 years 683 791 1787 
DVT 545 633 1426 
PE 138 158 361 
Death 1573 1600 1745 
Adverse events at 90 days 364 725 314 
Minor bleed 285 510 244 
Major bleed 65 116 55 
Asymptomatic HIT 45 
Symptomatic HIT 54 

Author notes

Disclosure:Employment: Jay Lin-Employee of sanofi aventis. Research Funding: Steven Deitelzweig-sanofi-aventis, Bristol-Myers Squibb, Scios; Josh Benner, Russ Becker-employees of IMS Health which has received funding to perform this work from sanofi-aventis. Honoraria Information: Steven Deitelzweig-sanofi-aventis, Bristol-Myers Squibb, Scios. Membership Information: Steven Deitelzweig-speakers bureau/advisory board for sanofi-aventis, Bristol-Myers Squibb, Scios, Pfizer. Financial Information: Financial and editorial support for this publication was provided by sanofi-aventis US, Inc.