Risks of Participating in a Clinical Trial

	New treatments may have unknown risks and serious side effects There may be no benefit, There may be extra doctor visits, and Extra costs.

 

Clinical Trial Result Analysis, Reporting and Publication

The design of all clinical trials should include plans and methods for analysis and reporting of the data collected.

Publication (and non-publication) of trial results can be motivated by several factors. Research reports may:

• Affect clinical practice
• Inform patients
• Influence future research
• Prevent duplication of effort.

Many believe there is a moral obligation to report results of all trials that involve patients.

Publication also has important secondary effects. These may explain existing patterns of publication and resistance to change-despite evidence that current systems are not the best. Authorship of papers in journals can establish reputations and enhance career prospects. Institutional or corporate reputations may also benefit.

The productivity of academic departments is judged on their publication output, which also affects their chances of obtaining future funding. The peer-reviewed (and indexed) journal has thus become part of the process of academic appointments and promotions, which, in turn, has spawned the discipline of bibliometrics.

And, of course, there are financial interests: drug companies use publications to increase sales while publishers make money both directly from journal sales and from spin-offs such as reprints and advertising.

Formats for how publications of results of clinical trials have been created and major medical journals expect these formats to be followed in papers submitted for publication.

 

Authors should include the following elements:

	Describe the patient sample List the data collected Explain the methods of statistical analysis used Describe quality control used to insure completeness and accuracy List endpoints and their interpretations Conclusions

 

Problems

While there have been some major changes such as the development of "open access" journals, information remains hard (or expensive) to access, scattered across many places, and difficult to synthesize. Producing and disseminating printed material can be slow, and journal peer-review often contributes further delays.

Another problem with the existing system is that non-publication of negative trials and non-reporting of negative outcomes, coupled with redundant publication of positive findings, has led to systematic publication bias, which can undermine the reliability of medical evidence.

 

Solutions

One of the most important initiatives aimed at reducing publication bias is trial registration (i.e., making details of study designs publicly available at the start of a study). One potential advantage of electronic publication coupled with trial registration is that study results can be linked to the original protocol, or at least to a summary of its main features.

Clear study identification, e.g., by including a trial register number, should highlight redundant publications. Secondary analyses will be easy to identify, and the risk of inadvertently including the same results more than once (which can bias systematic reviews and meta-analyses) will thus be reduced. Reviewer (and reader) access to information about trial designs and planned outcomes, which was recorded at the start of each trial, will permit detection of selective reporting since reports of results can be compared with the original protocol. Suppression of unfavourable findings will be apparent if all studies are registered at inception. But trial registration can prevent publication bias only if it is coupled with a commitment to make results from all studies available and a mechanism for implementing and policing this.

Guidelines for pharmaceutical companies and investigators have long emphasized the importance (and some would argue moral obligation) of publishing results from all studies, regardless of their outcome. But if such guidelines are to be enforced, we need a clear definition of what is meant by publication.

 

"Summary: Reporting Results from Clinical Trials

What's the problem?

• Clinical trial results are scattered and often difficult to access.
• Non-publication of unfavourable trials or negative results causes systematic publication bias.
• Standard journal articles often omit important details.
• Traditional peer-review and journal publishing delays the publication of findings.
• Enforcing commitments to publish all results are hampered by a lack of a clear definition of publication.

 

What are the opportunities?

• Trial registers/linking protocols with publications.
• Online publication on Web sites with minimal running costs.
• Data banks are a possibility for the future.

 

What's the solution?

• If all trials were registered at inception, redundant publication would be easy to spot and non-publication could be challenged.
• Links between trial protocols (or registers) and results would highlight selective reporting and post hoc analyses.
• Electronic templates for results reporting should raise standards and improve searchability.
• Results could be posted on publicly accessible Web sites at minimal cost. Peerreviewed journals should concentrate on interpretation, synthesis, commentary, and discussion."

 

Quote excerpted from: Wager E (2006) Publishing Clinical Trial Results: The Future Beckons. PLOS Clin Trial 1(6): e31. doi:10.1371/journal.pctr.0010031