A test to predict future risk of blood clots in people who have already had one has been developed.
Airplane travel is associated with increased risk of blood clots
A third of people develop a second clot within five to eight years of the first but long-term preventive treatment has dangerous side-effects.
Researchers at Vienna University found that by measuring thrombin, a clotting protein, they could eliminate those who didn't need treatment.
The study is published in the Journal of the American Medical Association.
The incidence of blood clots, technically known as venous thromboembolism, is one in 1000 in the UK.
People most at risk are the elderly, pregnant women, women taking oestrogen hormone treatment such as the pill or HRT, and people with cancer.
Another important risk factor is immobility.
It is estimated that 8,000 to 9,000 British airline passengers suffer from deep vein thrombosis - a type of venous thromboembolism - every year, and 25,000 patients die every year after developing a clot in hospital.
The likelihood that people will develop a second clot varies and is dependent on several complex risk factors but currently it is very difficult for doctors to decide which patients need long-term treatment with anticoagulant drugs.
Treatment with anticoagulant drugs such as Warfarin thins the blood to prevent clotting but may cause bleeding internally and therapy has to be monitored closely.
Doctors can carry out clotting tests, a process called thrombophilia screening, but it is expensive and the results can be inconclusive.
To identify whether thrombin would be a good test to rule out patients at low risk of a second clot, researchers measured generation of thrombin in 914 patients with a first spontaneous venous thromboembolism.
They excluded patients with cancer or who were pregnant or who had suffered trauma.
They found that two-thirds of patients, who had lower thrombin generation (less than 400nM), had a 60% lower risk of blood clot recurrence than those with higher results.
These patients would not benefit from long-term treatment with anticoagulants, the researchers concluded.
A follow-up study looking at the outcome of not treating patients at low risk is being done.
Study leader Dr Paul Kyrle said: "Adequate clinical tests to identify patients who might benefit from long-term (longer than six months) treatment are currently lacking.
"Duration of anticoagulation is decided on an individual basis, taking into account the individual risk profile of the patient.
"Mortality of patients with recurrent venous thrombosis is approximately five out of 100 patients.
"Thus, identification of high risk patients and long-term treatment of these patients with anticoagulants will reduce deaths.
"On the other hand, identification of low-risk patients who should be taken off anticoagulants will also save lives as these patients are no longer at risk of anticoagulation-induced bleeding."
He added that measurement of thrombin generation was "easy to perform and unexpensive".
Dr Patrick Kesteven, Consultant Haematologist at the Freeman Hospital in Newcastle, said there was "no doubt" a test to identify patients who needed to be treated long-term was needed - but added that thrombin would probably end up as one of a few tests used.