Releasing Study Data before Journal Publication Could Result in Incorrect Clinical Applications, Yale Researchers Find
When results of clinical trials were revealed before publication in a peer-reviewed journal, Yale researchers saw changes in medical practice that were not necessarily applied to the kind of patients and settings described in the trial results.
Research results on a surgical procedure designed to reduce the risk of stroke were released in two “clinical alerts” several months prior to publication in medical journals. After the clinical alerts were released, there were immediate and substantial changes in medical practice, although the new medical evidence was applied to some patients and in some clinical settings not directly supported by the study results.
“Both of the studies were performed in highly specialized centers, and elderly patients were excluded-allowing the investigators to minimize the surgical complication rate,” said Cary P. Gross, M.D., assistant professor of internal and general medicine at Yale School of Medicine and principal investigator on the study. “However, we found that after the results were disseminated, many procedures were performed in hospitals without a strong track record for success, and in patients who would have been too old to participate in the trials.”
Published in the December 13 issue of Journal of the American Medical Association (JAMA), Gross’ research highlights an ongoing debate about access to study results among journal editors, scientists and the media. Medical journal editors often prohibit widespread dissemination of clinical trial results until a manuscript is published, ensuring that clinicians, patients and researchers have nearly simultaneous access to study data. But members of the media have argued that the public has a right to know medically urgent study results as soon as they are available, especially if the trials are funded by a government agency such as the National Institutes of Health (NIH).
Gross and colleagues analyzed data on the use of carotid endarterectomy (CEA), a surgical procedure to remove blockages of the carotid artery when it becomes narrowed by altherosclerosis. They examined hospital discharge data on CEA rates before and after release of clinical alerts on two studies of the procedure.
The first study was the North American Symptomatic Carotid Endarterectomy Trial (NASCET), which looked at patients under age 80 with symptomatic carotid artery stenosis (mini-stroke or non-disabling stroke 90 days prior to entry). The second study, the Asymptomatic Carotid Atherosclerosis Study (ACAS), enrolled patients with asymptomatic carotid artery stenosis of greater than 60 percent who were between 40 and 79-years-old, and randomized them to receive either CEA or medical care. Both studies ended early because preliminary analysis showed that CEA was effective in reducing the risk of ipsilateral stroke. Clinical alerts were issued six months and seven months, respectively, before the full articles were published.
After results of the NASCET study were released in a clinical alert, the number of CEAs performed in acute care hospitals increased by 3.4 percent per month during the following six months and then increased 0.5 percent per month after journal publication of the study. After results of the ACAS trial were released, the CEA rate increased by 7.3 percent per month during the following seven months and then decreased by 0.44 percent per month after the trial was published.
Authors of both the NASCET and ACAS trials cautioned in their published articles that CEAs should be performed in hospitals with low surgical mortality rates. Gross found that at least with regard to hospital expertise and patient information, these warnings apparently went unheeded.
“After the ACAS clinical alert, the CEA rate increased more in patients aged 80 years or older than in younger patients, but after journal publication of ACAS, the CEA rate decreased more rapidly in the older population,” said Gross. “Although it is well-known that elderly patients have increased surgical risk, there was a large increase in CEA rates in this age group.”
Gross and colleagues point out that in the ACAS clinical alert and subsequent article, the investigators were clearly concerned about extrapolating their results to settings in which the operative risk was much higher. The NASCET article, but not the clinical alert, also contained a warning cautioning readers not to apply their conclusions too broadly.
“Policymakers should consider that the prepublication release of clinical trial results may have unintended consequences,” said Gross. “Because publication of the full article may help promote optimal use of the new information, future work should explore the impact of other mechanisms of prepublication dissemination, such as Internet publication of full-length, peer-reviewed articles.”
Other researchers on the study included Claudia A Steiner, M.D., Eric B. Bass, M.D., and Neil R. Powe, M.D. The study was funded by the Robert Wood Johnson Clinical Scholars Program and the Agency for Healthcare Research and Quality.