Government plans to publish the death rates at hospitals will give the public and the health profession little useful information, according to research.

The tables were promised by the Health Secretary Frank Dobson following the deaths of 29 babies and young children at the Bristol Royal Infirmary.

The doctors who carried out the operations, Mr James Wisheart and Dr Janardhan Dhasmana, had success rates well below the national average.

But a study in the British Medial Journal (BMJ) concludes that crude league tables will not give a reliable picture of performance or best practice.

Hospital comparisons

Researchers from Sheffield and Scotland looked at the death rates at nine neonatal intensive care units in the UK and constructed crude performance tables based on the data.

Simple mortality rates could be damaging

Over a period of six years, they found the hospitals that they studied fluctuated wildly in the rankings.

Dr William Tarnow-Mordi, one of the authors of the study, says it is difficult to see how such information could be of value.

"It's such a blunt approach, we may not learn anything from these random order rankings. If you look at the league tables in the paper, you'll see that in five out of six years a different hospital finished last.

"But we did discover this unintended possible negative effect that the hospital that happened to be last would have a harder job to do because of the damage the tables would do to their credibility," he says.

Early warning system

Dr Tarnow-Mordi, a reader in neonatal medicine and perinatal epidemiology at the University of Dundee, says mortality league tables would eventually show up under-performing hospitals, such as Bristol, if the data was collected for long enough. But they could not be used as an early warning system.

"If you wait ten years before you get enough numbers in the sample to be able to say this hospital has been under-performing, how do you know it is still under-performing? It's like the stock market: past performance is no guide to the future."

The information would also not tell you what needed to be put right, he adds.

The research team suggested more refined techniques be used to judge the performance of hospitals. They said studies that asked specific questions about the different procedures and policies used across large groups of hospitals would be more likely to produce useful data. The more relevant information could be available in a matter of months, not years, Dr Tarnow-Mordi says.

"We are completely behind openness and accountability," he says. "All the health service staff I know are very keen they get feedback as to whether they are doing a good job, and if they are not, they want to know what to put right. But we think we need to look for more rapid and reliable early warning systems that relying on mortality. That takes too long."

Bristol repercussions

The BMJ's editor uses the current edition of the journal to reflect on the lessons of the Bristol heart scandal. Dr Richard Smith says the case will probably prove much more important to the future of health care in Britain than the reforms suggested in the White Papers.

"...the Bristol case is a once in a lifetime drama that has held the attention of doctors and patients in a way that a White Paper can never hope to match," he said.

He says Frank Dobson was wrong to say that all three of the doctors involved in the case should have been struck off and wonders how the Secretary of State for Health will now proceed with the concept of self-regulation within the medical profession.

Dr Smith believes that the Government will not attempt an overhaul of the General Medical Council as this would "miss the point".

He says that the Royal Colleges and postgraduate deans have a much stronger everyday influence on the practice of doctors and that they must recognise their role in self regulation.

Dr Smith warns that the failure of doctors' organisations to implement much better mechanisms for ensuring high quality of care might lead to the "micromanagement" of doctors that is routine in the United States.