A commonly used exercise test to check whether people need open heart surgery could be sending twice as many people to surgery than is necessary, says a study by the U.K.'s University of Leicester.

Doctors have been using the internationally approved test since 2012 on people with a condition called aortic stenosis (AS) to determine whether they need an operation to save their lives. But the new study found the test is "highly inaccurate."

Aortic stenosis is the narrowing of the aortic heart valve which restricts the slow of blood from the left ventricle to the aorta. It usually affects older people due to the buildup of scarring and calcium buildup in the valve. Up to three percent of people over 75 years of age can have the condition.

According to the Mayo Clinic, the condition can range from mild to severe. Symptoms, such as chest pain, breathlessness and feeling faint, can take years to develop and are often detected during a routine physical. When symptoms develop, it means the person is seriously ill and could die from heart failure or sudden death.

The exercise test, which involves cycling on a stationary bike, is used to determine whether surgery is needed, but the new study found it only has a 60 per cent accuracy rate.

If exercise test participants become breathless, they are recommended to have valve replacement therapy. There are about 200,000 cases of AS in the U.S. every year, and aortic replacement surgery requires a hospital stay between seven and 10 days.

"There is no doubt that valve replacement therapy is highly effective for patients with symptoms, however there are risks involved," said University of Leicester professor Gerry McCann. "It's a major operation and there's a one per cent chance of people dying or having a stroke during or after. There's also the chance they could develop an infection.

"It can often take six months to recover, but if they survive they tend to do very well afterwards," McCann said. "However, if we know a patient has AS and no symptoms and we do nothing there's also a one per cent chance they will die, so there's a fine line between whether we should intervene or not.

"Our findings showed that this exercise test, which has been approved by the American Heart Association/American College of Cardiology and the European Society of Cardiology, was highly inaccurate as almost twice the number of people who became breathless during the test did not develop symptoms within a year."

The findings have been published in the European Heart Journal.