None of those involved were told that people were engaging in what is known as "intercessory prayer" on their behalf.

Just over half of these, picked at random, were made the subject of intercessory prayer.

However, on average, the 500 patients prayed for had 11% less complications during their stay in hospital.

In all, the researchers looked at 990 people admitted to the coronary care unit over the course of a year.

Using a standard coronary care scoring system, patients from prayer and non-prayer groups were assessed. Those who were the subject of prayer did better than their fellow patients, although in general, they did not get out of hospital any earlier.

'Could be chance'

The research team, based at a university hospital in Kansas City, US, admitted that no rational explanation could be found to explain away the difference.

But their report said: "We have not proven that God answers prayer or that God even exists. It was intercessory prayer, not the existence of God, that was tested here.

"Chance still remains a possible explanation of our results."

The prayers were said by volunteers from a local church. They were simply sent the patient's first name on a piece of paper, and told to pray for "a speedy recovery with no complications".

Studies inconclusive

Those being prayed for did not even know that a clinical trial was going on, let alone that they were the target of prayer. The scientists believed that knowing that someone was praying for you could conceivably have an effect on outcomes.

Other studies into the effects of intercessory prayer have been inconclusive - it has been found in earlier studies to have no significant effect on leukaemia patients, or those suffering from anxiety or depression.

However, one study into AIDS patients found significant improvements in the number of infections acquired when intercessory prayer was given.