MRSA hospital league tables are meaningless because of the volatile nature of the superbug, an expert says.
MRSA rates have been recorded for the past few years
Medical Research Council statistics expert David Spielgelhalter said the infection occurred in clusters which could distort the true picture.
The trend meant even if hospitals reduced the underlying risk it would not always be reflected in the regular infection rates published since 2002.
But the government said the recording system was getting more sophisticated.
Ministers introduced a mandatory surveillance scheme for MRSA four years ago, and last year set a target of a 50% reduction in rates by 2008.
Other infection surveillance schemes for hospital bugs are also up-and-running.
The government acted after increasing public concern over the hospital superbug, which is linked to nearly 1,000 deaths a year.
The data, which is produced six monthly, initially showed a rise in MRSA infection rates, but latest results have shown they are now declining nationally.
But the Department of Health also publishes individual rates for each trust, which means they are used by journalists and the public to rank hospital performance.
Dr Spielgelhalter said using the data as league tables of hospital performance was "entirely spurious".
"The basic problem is that it is unclear whether the targets refer to an observed rate reduction or a true reduction in underlying risk.
"This ambiguity is unimportant at the national level but, for individual hospitals, chance variation can make the observed rates extremely volatile and make simplistic notions of hitting targets unreliable."
Using data from 2001 to 2004 and analysing the long-term trends, he found the figures varied too much for individual trusts for them to be meaningful.
The British Medical Journal study cited the case of Aintree Hospitals NHS Trust which had 34 cases in 2001-2, rising to 66 in 2002-3 and falling to 48 the following year.
And he concluded even if a hospital reduced the real risk by 20% this only had a 50:50 chance of being reflected in the infection rates.
Georgia Duckworth, of the Health Protection Agency, which collates the data, accepted there were limitations to the figures.
But she added: "Mandatory surveillance of MRSA infection rates has raised the profile of infection control."
And a Department of Health spokesman said a more "sophisticated" reporting scheme had been introduced this year to reflect things such as the patients having the infection when they entered hospitals.