12 October 2011

Government evidence based on flawed research, experts say

Research used to compile the Government's controversial Health and Social Care Bill is flawed, according to a Dundee academic who worked with a team analysing the proposed legislation.

Dr Carlo Morelli, an economist from the University's School of Business, undertook economic analysis and interpretation for a report, published in The Lancet, which examined a previous study that was used as evidence to suggest that the introduction of patient choice in the NHS reduced deaths from heart attacks.

The Lancet report, authored by Professor Allyson Pollock of Queen Mary, University of London, pointed out a series of errors in the study, which the Prime Minister has used as justification for the claim that 'competition is one way we can make things work better for patients.'

The original study examined the mortality rates for heart attack patients measured against the number of hospitals within travelling distance of the patient’s GP surgery. It also looked at data on elective surgery for hernia, cataract repair, knee arthroscopy, hip replacement and knee replacement, and claims that introducing greater choice in elective surgery led to lower death rates from heart attacks.

Professor Pollock and her colleagues - including Dr Morelli - say that the study offers no explanation as to why the availability of choice for such elective procedures should have any effect on whether heart attack patients survive.

Their paper, ‘Statistical association is not causation: Claims that patient choice and market competition in the NHS reduce AMI mortality are misleading and false’, was published in the most recent edition of The Lancet. Dr Morelli says the inaccuracies contained in the original research casts severe doubt over the validity of the legislation.

'The link between competition and patient outcomes that the Government claims simply doesn’t exist,' he said. 'The outcome is far from linked to geographical dispersion of hospitals.

'There are problems contained within the report. One I identified was the measure of competition used. The results were obtained by looking at instances where there were a lot of hospitals in an area.

'The better outcomes therefore could be the result of hospitals being closer to heart attack patients, and the availability of specialist treatment. Looking at this, my conclusion would be that competition is used in this sense as a substitute for urbanisation and collaboration.'

Increased choice for patients is a central plank of the Health and Social Care Bill 2011, which currently faces a vote to pass through the House of Lords. Although the legislation relates only to England and Wales, Dr Morelli says there is scope for a study focussing on Scotland which would show the impact of choice for patients.

'My colleagues and I are looking at a possible comparison within Scotland to see how the concentration of hospitals, GPs and patients impacts upon outcomes,' he added.

The Lancet report also concluded the following about the original paper:

• The researchers do not look at whether the availability of choice has any effect on where patients go for treatment.
• They do not look at whether or how GPs’ patterns of referrals changed when choice became available.
• Recent research indicates the majority of patients who have been offered a choice pick their nearest hospital.
• Heart attack is a medical emergency and patients generally have no choice about where they are treated.
• Outcomes for heart attack patients tend to be better when they are treated in specialist centres in urban areas.
• The authors ignore the possible effects of major changes in primary care prevention and secondary care intervention for heart attacks.
• And that there is no evidence that the data on elective operations is in any way a good measure of choice or competition.

Professor Pollock said, 'The Government’s Health Bill has faced enormous opposition from the public and from health professionals. In trying to win over his critics the Prime Minister has used the study to justify competition within the National Health Service.

'Our examination of this research reveals it to be fundamentally flawed, amounting to the conclusion that the paper simply doesn’t prove either cause or effect between patient choice and death rates. This work should not be quoted as scientific evidence to support choice, competition or the new Health and Social Care Bill.'

For media enquiries contact:
Grant Hill
Press Officer
University of Dundee
Nethergate, Dundee, DD1 4HN
TEL: 01382 384768
E-MAIL: g.hill@dundee.ac.uk
MOBILE: 07854 953277