Top health economists have given validity to Department of Health policies on centralisation of specialist services and reducing MRSA rates.

University of Arizona health economics professor Gautam Gowrisankaran claims he has proved bigger hospitals have better health outcomes for some surgery patients because clinicians "learn by doing".

Although this has been assumed for some time and is crucial to moves to centralise specialist services, Professor Gowrisankaran told a health policy conference last week that until now no one had examined whether bigger hospitals provided better services or better services attracted more patients.

He said: "If learning by doing is the right explanation then consolidation of procedures may be in the public interest."

He examined the details of 4,000 pancreatic cancer surgery patients in the US and found for every extra patient operated on the hospital's mortality rate for that procedure lowered by 0.6 percentage points. There was no evidence selective patient referral affected mortality rates.

At the same conference, academics from Nottingham, Oxford and Surrey universities said although the number of MRSA infections was rising, this was because of the huge increase in patients being treated in hospital. From 2001-06 the rate of MRSA per stay in hospital had fallen by 25 per cent.

The economists found the biggest contributor to MRSA infection rates was length of stay of stay not high bed usage rates.