Revealed: death rates in your local hospital

One hundred and fifty years after Florence Nightingale identified the importance of measuring survival rates, NHS trusts throughout England are publishing details of deaths following four common operations.

The move signals a shift from the delivery of care – measured in waiting times – to its outcome – measured in death rates – as called for in last week's NHS Review by Lord Darzi as a means of driving up standards.

It is the first time any government has exposed the mortality rates of surgical teams in different hospitals to public scrutiny. But, while Britain's most famous nurse marked her patients "dead", "relieved" or "unrelieved", her descendants are still struggling to find an accurate way of measuring the effects of treatment.

Click on the links below to compare death rates for different surgeries at your hospital

Repair of abdominal aortic aneurysm (planned and emergency)

Hip replacements

Knee replacements

The findings show all hospitals are performing to acceptable standards in the four operations, with a handful performing better than expected. No hospital recorded a higher-than-anticipated death rate.

However, the crude figures show wide variations. Some hospitals have three times the death rate expected while others have less than half that expected for the same operation. The four procedures are hip replacements, knee replacements, planned repair of abdominal aortic aneurysm and emergency repair of aortic aneurysm.

Sir Bruce Keogh, the NHS medical director, said it would be unfair to rank hospitals by the crude figures as the numbers of operations and deaths were too small in most cases to allow valid comparisons. "They are not statistically significant," he said.

Sir Bruce said he had feared the new data for the four operations would show major differences in death rates between trusts. "I was warned there might be significant variations in outcomes and I was expecting to find them. I was pleasantly surprised to see there was so little. You have got to believe the data."

He denied that the lack of variation meant the death rates did not help patients to choose a hospital. "I utterly refute that. People want a hospital close to their home and to know that it is safe. That is what this does."

Asked if the bar for achieving an "acceptable" death rate had been set too low, so no NHS trust had failed, he said it was the "industry standard". But he acknowledged there was a long way to go to improve ways of measuring quality of care. He said: "I think mortality is a pretty blunt measure. The reason I wanted to publish [the death rates] was to send a very clear message to the health service that we are now in the business of measuring quality in the NHS. Mortality is one measure but real quality is different. I am keen to establish a debate about the methodology."

How the figures add up

Death rates for four operations are listed for around 100 NHS trusts in England on the NHS website ( www.nhs.uk). Standardised Mortality Rates (SMRs) are presented in two bands – "better than expected" and "within the expected range".

Just five NHS trusts are described as having "better than expected" mortality rates. Department of Health officials said no hospitals were ranked worse than expected because those with high SMRs (above 100) had carried out too few operations or had too few deaths for the figures to be meaningful.

The trusts ranked better than expected are Newcastle upon Tyne NHS Trust, University Hospitals of Leicester NHS Trust, Worcestershire Hospitals NHS Trust, Norfolk and Norwich University Hospital NHS Trust and East Kent Hospitals NHS Trust.