MPs call for more transparent clinical trials

"Drug companies and medical researchers are putting patients' lives in danger by failing to publish unfavourable results from clinical trials, MPs have warned," The Guardian reports.

The news follows the recent publication of a report (PDF, 2.3Mb) by the influential House of Commons Science and Technology Committee, which called for greater transparency in clinical trials and the data they produce.

What are clinical trials?

There is currently no regulatory definition of exactly what constitutes a clinical trial.

The term is generally understood to mean a study that assesses how effective and safe a new drug or intervention is.

Most clinical trials take the form of a randomised controlled trial, where a new intervention is compared against either an existing intervention or a placebo (dummy treatment).

What are the concerns about clinical trials?

Currently, pharmaceutical companies (or anyone else) have no legal or regulatory obligation to make any data gathered during a clinical trial publically available.

In recent years, this fact has alarmed many healthcare professionals as it may lead to a distorted view of the effectiveness and safety of a particular intervention. This could in turn have an impact on patients' wellbeing.

The problem of publication bias, where studies with positive results are far more likely to get published in peer-reviewed journals than studies with negative results, has long been recognised.

The current best estimate is that trials with positive results are twice as likely to be published than others. Concerns focus not so much on what is being published, but what is being kept out of the public eye.

In his book Bad Pharma , doctor and author Ben Goldacre cites the example of a class of drugs called anti-arrhythmic drugs, which are designed to prevent abnormal heart rhythms. During the 1980s, these were routinely prescribed to people recovering from a heart attack, as developing an abnormal heart rhythm is a common complication.

It actually turned out that these drugs were potentially fatal. It is estimated that around 100,000 people may have died because of this mistake in prescribing policy.

As Goldacre points out, this tragedy may have been prevented if a study carried out in 1980 had been published. The study strongly suggested that anti-arrhythmic drugs are dangerous for heart attack survivors. The researchers involved in the study admitted this failing in a 1993 paper.

The campaigning group All Trials, as well as the BMJ, have also raised concerns that the benefits of Tamiflu, which has been stockpiled globally in case of another flu epidemic, have been "oversold".

They argue that because the full data about the drug's effectiveness has not been made available, it makes it difficult for both the public and policymakers to accurately judge whether it is value for money.

How has the pharmaceutical industry responded to these criticisms?

While there is no single voice for the global pharmacetuical industry, three main points have often been used as justification for not releasing unpublished trial data:

  • it could violate the principles of patient privacy and patient consent
  • it could threaten the confidentiality of intellectual property and damage the commercial prospects of individual companies
  • there is a risk that clinical trial data would be re-analysed in an inexpert or irresponsible way, potentially leading to misleading conclusions becoming widespread among the public

What recommendations do the MPs make?

The report identifies four levels of transparency and makes recommendations for each.

Level one: trial registration

The Science and Technology Committee recommends that all trials conducted in the UK and all trials related to NHS treatments are registered in a WHO-listed primary registry. This is a publically accessible international database of all planned clinical trials. It includes information on the trial sponsors, interventions, participants, sample sizes and the outcomes being tested.

Level two: summary trial results

The MPs' report recommends that summary trial results be made publically available, ideally in peer-reviewed journals. People or organisations funding trials should provide the time and resources to make this possible.

Level three: clinical study reports

The report recognises that it would be too difficult for clinical study reports for non-commercial trials (trials for drugs or interventions that are still at an experimental stage) to be made publically available.

However, once a drug or intervention has been approved for use, the report argues, there is no compelling reason why clinical study reports cannot be made publically available, provided individual patient information has been removed or obscured (redacted).

These reports should offer information on the researchers involved, statistical methods used during analysis, and manufacturing information about the drug in question.

Level four: individual patient-level data

The report does not recommend that individual patient-level data be made publically available, even in a redacted or anonymised form.

However, it does suggest that the data could be accessed via "safe havens", where access is facilitated by a "gatekeeper" who ensures that the data is used in a responsible manner that makes a real contribution to scientific knowledge.

What happens next?

It's hard to say. The Commons Science and Technology Committee is influential, but it does not decide government policy.

The debate about clinical trial data transparency has been raging for decades, but has become more high profile in recent years.

In response to the report, the NHS Health Research Authority has announced that from September 30 2013, all trials will need to be registered with the body before they can receive approval from an ethics committee.

NHS Attribution