Search tips
Search criteria 


Logo of bmjThis ArticleThe BMJ
BMJ. 2007 June 30; 334(7608): 1341.
PMCID: PMC1906627

Clinical trial results often overstate benefits of treatment

Failings in the way that clinical trials are designed and presented may lead doctors to overstate the benefit of treatments, experts warned last week.

The conference on clinical trials, organised by the James Lind Alliance and the Lancet and held at the Royal Society of Medicine in London, also heard that key groups of participants were often excluded from clinical studies and as a result were denied the benefits of evidence based medicine. Stephen Holgate, professor of immunopharmacology at Southampton University, said that children and elderly people were “especially neglected” in this area.

As another example he noted that the routine exclusion of smokers from asthma studies meant that it has only recently been discovered that inhaled steroids do not work in this group—decades after millions of smokers began taking these drugs for their asthma.

Professor Holgate said, “In order to redress the balance, more real world ‘effectiveness' studies are needed, recruiting all comers and using more patient centred outcome measure and global assessments.” He added that such studies should cover a “wide range of age and ethnic groups to take account of adherence and cultural factors.”

Another speaker, Peter Rothwell, professor of clinical neurology at Oxford University, said that many patients were being denied the best treatment because researchers were failing to analyse their findings with sufficient depth.

He said that by providing more detailed subgroup analysis, which identifies exactly which types of patients respond well to drugs or show serious side effects, doctors would be able to maximise the benefits and minimise the risks for their patients, by targeting drugs at the right people.

He noted that many researchers and statisticians disregarded subgroup analysis. But “there is nothing intrinsically misleading about it,” he said. By adopting this type of inquiry more often, doctors might find the answers to important and as yet unanswered clinical questions, including which groups taking warfarin were particularly at risk from brain haemorrhages.

He said that subgroup analysis had already dispelled many pervasive medical myths, including the beliefs that aspirin did not protect women from heart disease and that tamoxifen did not work in women under 50 years of age.

Doug Altman, professor of statistics in medicine at Oxford University, said that the presentation of statistical results of clinical trials “lacked transparency and precluded any further analysis.”

Finally, Paul Glasziou, director of the Centre for Evidence Based Medicine at Oxford University, warned that many clinical trials described treatments that were difficult to replicate in normal clinical settings—often because the researchers provided too little information about their study. Research of his that is due to appear in the BMJ indicates that 20% of clinical studies published in medical journals fell into this category.

He said, “One of my messages would be: ‘Please give sufficient details of treatments to allow other clinicians to replicate them.'”


More information is at

Articles from The BMJ are provided here courtesy of BMJ Group