more_vert
REPORTING of complete enrolment information in randomised clinical trials is improving and is a step in the right direction, say Australian experts.
Professor Warwick Anderson, CEO of the NHMRC, said transparency was the key in research and particularly in the development of clinical guidelines. Exclusion criteria and how they were applied needed to be disclosed so a study’s limitations could be properly considered.
Professor Anderson was commenting on a research letter published in JAMA Internal Medicine that found reporting of complete enrolment information in influential randomised clinical trials (RCTs) had increased from just 45% of studies in 2002 to 75% in 2010. (1)
The researchers examined data from the 20 most influential English-language RCTs in each of 14 prevalent chronic conditions, including cancer, depression, diabetes and hypertension.
Professor Anderson said this positive trend was attributable to the release of the Consolidated Standards of Reporting Trials (CONSORT) statement in 2001. However, he said, a little more improvement was needed.
“I would hope that what articles like this can do is make researchers … be completely transparent in what they do”, he said, adding that it might also prompt researchers to think about how tightly exclusion criteria really needed to be drawn.
The researchers analysed 145 RCTs published from 2002 to 2010 and found an average of 40.1% of patients identified with the disorder being studied were not enrolled in the trials.
“The larger the proportion of patients not enrolled, the more likely it is that the results of the study will not reflect what the intervention would produce in front-line clinical practice”, the researchers wrote. “Although exclusion criteria are sometimes essential in trials, including to protect patient safety, we add our voices to those of others who have suggested that treatment researchers use them as minimally as possible and only with good justification.”
Professor Anderson said the call by the researchers to minimise exclusion criteria was too blunt, noting that there was a trade-off between precision and universality in research.
“As a general principle, broad is good, but it’s not always what you want because sometimes the question the trial is addressing is really relatively narrow”, Professor Anderson told MJA InSight.
Exclusion criteria were drawn firmly for good reason in many cases, he said. “There are downsides to that, [but] I think these are really issues that have to be judged on a case-by-case basis.”
Professor Paul Glasziou, professor of evidence-based medicine at Queensland’s Bond University, agreed it was “good news” to see that the rates of complete enrolment information on the increase.
He said the rates of non-enrolled patients were not as high as many would expect. “If these figures are correct, where you have 60% of [patients] they’re screening as potentially eligible going into the studies, that’s actually pretty good”, Professor Glasziou told MJA InSight.
He said a more important problem than patient non-enrolment was the restriction of eligibility criteria. Broadening the eligibility criteria for RCTs would help to ensure results were more generalisable, he said.
“The biggest problem is knowing whether that relative risk holds across all risk groups. If you do the trial only in the high-risk patients, does it really apply to the lower risk groups? So it’s not so much a problem of getting everybody who’s eligible in, it’s a matter of what you define as your eligibility criteria in the first place.”
Professor Glasziou said broader eligibility criteria would not only make it easier to generalise the trial results, but it would also make it easier to recruit patients.
Professor Anderson said the NHMRC was seeking to not only fund, but also to build clinical trial activity in Australia. Last year, the NHMRC, in collaboration with the Department of Innovation, Industry, Science and Research, launched a new website — www.australianclinicaltrials.gov.au — to inform patients about clinical trials and how they could volunteer to participate in them.
1. JAMA Int Med 2013; Online 22 April