Policies on reporting clinical trials and publishing supplements

Clinical research and practice in neurology are rapidly changing with the development and introduction of new therapies. Within the past 3 years, specific treatments have been introduced for disorders that were previously considered untreatable, including ischemic stroke, amyotrophic lateral sclerosis, multiple sclerosis, and Alzheimer's disease. In addition, a flood of new therapeutic agents, some representing novel pharmacologic classes, are being introduced for the treatment of epilepsy and Parkinson's disease. After years of slow discovery of therapeutic agents, this veritable deluge of therapeutic options may be overwhelming for clinicians. Moreover, there is a vast array of other interventions currently in development that will complement the compounds that are already marketed. Neurology has always sought to publish information on new treatments; however, we intend to take additional steps to make experimental therapeutics a major focus of the journal. The development and reporting of new treatments will have a critical impact on the lives of our patients, the practice of neurology, and the planning of future basic and clinical research.

To make Neurology as useful as possible to authors and clinicians, we will expedite publication of important clinical trials (see preceding editorial) and publish well-designed and conducted clinical trials regardless of the outcome of the study. Journals have sometimes had a bias toward publishing "positive" (regarding therapeutic benefit) studies while relegating "negative" studies to appear as abstracts or to less available venues. "Positive" studies with methodologic deficiencies have been prominently published, while "negative" studies with more methodologic rigor have not. We will focus on publishing more rigorously designed studies, specifically those using a control group, randomization, and a method of blinding, instead of publishing studies with weaker designs. Uncontrolled, nonrandomized, unblinded studies will be considered for publicationas Brief Communications or Clinical/Scientific Notes but will rarely be appropriate for publication as full-length articles. Such observations are often the essential stimuli for more rigorous research publication, but the quality of the evidence must be emphasized.

The unevenness in the method of reporting the results of clinical trials has led to difficulty in interpreting the study results. There have been several attempts to standardize the method for reporting clinical trials, but some of these efforts led to cumbersome manuscripts. The recently published CONSORT statement presents a comprehensive yet simple and straightforward method of reporting clinical trials.¹

The CONSORT statement does not prescribe a specific manuscript format but provides a checklist of critical design features that must be reported. Reviewers are provided with the checklist to assess the manuscript for the relevant content. The checklist also serves as a useful guide to readers as a way of assessing the methodologic soundness of the report. In addition, a figure must be included in the manuscript that clearly outlines the flow of participants through the trial. This standardized method of reporting should simplify the evaluation of clinical trial reports and should aid the reader in understanding the implications of such reports for clinical practice. Many journals, including the Journal of the American Medical Association, the British Medical Journal, and The Lancet have adopted these guidelines. We ask that clinical trial reports be submitted in this format beginning immediately, and we will require that clinical trial reports be submitted in this format as of January 1, 1998.

We will have an accompanying editorial with many major clinical trial reports. An editorial provides an unbiased assessment of the trial, as well as a brief review of therapeutic options.

Supplements to Neurology have also provided information on new therapeutic options. In the past 2 years, there have been supplements on the treatment of migraine, Parkinson's disease, epilepsy, amyotrophic lateral sclerosis, and myasthenia gravis. The material in these supplements is not peer reviewed and their publication is usually financially supported by a pharmaceutical manufacturer. There have been a number of objections to accepting any non-peer reviewed material in Neurology. Our response is that supplements are clearly identified as such, including a separate pagination system, they are free of charge, and many clinicians find them useful. Bias, both overt and subtle, may be present in all forms of publication.

Corporate sponsorship and lack of peer review does not necessarily lead to bias in supplements, but there is at least the possibility, and certainly the perception, that bias may be present. To help address this concern, Neurology's existing supplement policy, which has not previously been published, was revised and expanded and is reproduced here. An important part of the policy continues to be our review (including independent external reviewers) of supplements for overt bias, although subtle forms of bias are difficult to identify or change. We require clear disclosure of any financial relationship between the sponsors of the supplement and the participating authors. We hope this policy will help readers in assessing real or apparent conflicts of interest and in deciding for themselves the potential bias in the material presented in supplements.

We will continue to invite and encourage supplements, provided that they comply with the new policy. In the past, readers may have relied on the supplements to help place new treatment options in context. We hope that the new editorial policy regarding the reporting of clinical trials will enable readers to rely more heavily on peer-reviewed publications.

Policy for developing and publishing supplements to Neurology