Creating league tables for heart surgeons is fraught with risk, researchers have warned.
Patients will have information about their surgeon's performance
They say only performance in low-risk cases should be assessed.
The researchers warn including high-risk cases would give an unfair picture of surgeon's mortality rates because they are performed on the sickest patients, who are therefore the most likely to die.
The government has called for all operations to be included in mortality tables.
It wanted league tables to be introduced after the Bristol hearts scandal, where surgeons at the Bristol Royal Infirmary were found to have continued carrying out heart operations on children even though they had higher than average death rates.
It helps to know what the mortality rate is so if someone's out of line, you can look and see why
Mr Colin Hilton, Society of Cardiothoracic Surgeons
But experts had warned rates can appear high if surgeons operate on a large number of very sick patients.
The Society of Cardiothoracic Surgeons has announced it will give star ratings to surgeons depending on death rates in patients. It has said the data will exclude very high-risk patients.
Researchers from the South Manchester University Hospital assessed 8,572 patients who were having heart bypass surgery for the first time between April 1999 and March 2002 in north west England.
The patients were operated on by 23 different surgeons across the area.
They calculated mortality rates for each surgeon and predicted mortality for each patient using a recognised scoring system, defining mortality as any death which took place in hospital.
Overall the risk of death for patients undergoing this kind of operation was
But they found mortality rates between different surgeons ranged from 0% to
The majority of patients - 85% - were defined as being at low risk of dying, and there was relatively little variation between surgeons.
But it was found that the proportion of high-risk patients that surgeons operated on varied from 5.6% to 23.9% which meant there was a much greater variation in death rates for this group of patients.
The researchers led by consultant cardiac surgeon Ben Bridgewater, warned publishing "crude" rates, could mean surgeons were less prepared to take on high-risk cases.
Patients 'turned down'
Writing in the British Medical Journal, they said: "Cardiac surgeons already work in a stressful environment, and the perception that a 'bad run' might jeopardise their career may lead to a tendency to turn down high risk patients.
"The easiest way to obtain low mortality is to only do straightforward operations - so-called risk averse behaviour."
In order to prevent this trend developing, the researchers say surgeons' performance should be compared on the basis of death rates among low risk patients.
Mr Colin Hilton, president of the Society of Cardiothoracic Surgeons, told BBC News Online: "The society will not be including very high-risk patients in its data."
But he added: "It helps to know what the mortality rate is so if someone's out of line, you can look and see why.
"Most of the time, it will be because they have been operating on high-risk patients."