Abstract

We report final results from the first, large, multicenter, prospective cohort study designed to define the incidence of symptomatic venous thromboembolism (VTE) in patients with isolated leg fractures below the knee, and determine if routine thromboprophylaxis will be cost-effective. The reported incidence of deep vein thrombosis (DVT) using screening venography in such patients is as high as 40%. This has lead to the use of routine anticoagulant prophylaxis for several weeks in many such patients. However the vast majority of venographically-detected DVTs are asymptomatic, distal thrombi, whose clinical relevance is uncertain. Therefore venography may not be the best outcome measure to assess the clinical burden of clinically-important VTE, or as an endpoint for trials of thromboprophylaxis.

Methods: From August 2002 to June 2005, 2446 consecutive patients with fractures of the patella, fibula and foot (treated operatively or conservatively) and tibial fractures (treated non operatively) were assessed at 5 hospitals in Ontario, Canada. Patients were enrolled after informed consent within 96 hours of the injury. Patients with major trauma, active cancer, and previous VTE were excluded. Thromboprophylaxis was not allowed. Patients were followed prospectively for 3 months, with telephone calls at 14 days, 6 weeks and 3 months. Education regarding symptoms of DVT and pulmonary embolism (PE) was provided at study entry and patients were asked about VTE symptoms at follow up. Suspected DVT and PE were investigated in a standardized manner.

Results: 1200 patients have been enrolled and 1174 have completed 3-month follow up at this time. The last patient will complete follow up by the end of September 2005. The mean age of this cohort was 45 years (range 16 to 93) and 60% were female. Most injuries were caused by falls (56 %), followed by vehicular accidents (25 %), sports injuries (16 %), and occupational injuries (3 %). 18 % of patients had other minor injuries. 99 % of these fractures were unilateral. Fractures of the fibula were the commonest (39 %), followed by metatarsal (30 %), phalanges (12 %), calcaneus, talus or tarsal (10 %), tibia (10 %) and patella (5 %). 7 % of fractures were surgically treated. 82% of patients received a cast or splint for a mean duration of 42 ± 35 days. Six patients (0.5%) received anticoagulant prophylaxis (outside study protocol) at some time within the study period. Complete follow up was available for 98 % of this cohort. 21 patients either withdrew consent for follow up, or could not be contacted despite numerous attempts. By 3 months, only 7 of 1174 patients had sustained a symptomatic, objectively-confirmed VTE (2 proximal DVT, 3 calf DVT, 2 PE) with no fatal PE - an incidence of 0.6 % (95% C.I. 0.2 to 1.2).

Conclusions: Based on a large prospective cohort of patients with leg fractures below the knee with no prophylaxis use, and 98% follow-up, we observed no fatal PE, and only 0.6% symptomatic VTE, confirming that routine thromboprophylaxis is not warranted, and not likely to be cost-effective in these patients. Furthermore symptomatic VTE is so uncommon, that it is unlikely that any high-risk subgroups can be identified for targeted prophylaxis. This study also highlights the significant discrepancy in the incidence of VTE between studies using venography and those utilizing clinical endpoints.

Author notes

Corresponding author