Scientists Call for Increased Transparency in Research

Scientists at the Yale Collaboration for Research Integrity and Transparency (CRIT) published a new policy paper this month criticizing the current state of biomedical research and calling for improved transparency in research methods. The researchers lead with examples of instances where a lack of transparency caused undue harm to the public. They also report on the improvements in transparency so far, and how to continue to enhance transparency.

Transparency, broadly defined, is achieved when research data of all kinds are made available to the public, to the government, and to other researchers. When data is not available, “Investigators can conduct research in the dark and deviate from good research practices.” However, when data is available, it allows other researchers to check the work, it allows government regulators to ensure that medical products and drugs are safe and effective, and it allows those in the medical profession to help the public to make an informed decision about their treatment.

The writers of the policy paper enumerate recent examples of research misconduct that was obscured by lack of transparency:

• GlaxoSmithKline’s Study 329.
  • The drug company’s study of the antidepressant paroxetine (Paxil) is one of the most infamous examples of outcome switching. By 2002, over 2 million children received prescriptions for Paxil, which the company claimed was safe and effective.
  • However, in 2015, researchers re-analyzed the actual original data from the study and found that Paxil was not effective for children, increased children’s risk of suicide, and had many more negative side effects than were initially reported.
  • Investigators discovered that the publication of Study 329 did not report a single one of the original tests on Paxil—because all of those tests showed that it was neither safe nor effective. Instead, in one of the most infamous examples of the deceptive practice known as “outcome switching,” the publication included all new outcome measures, making the drug appear more effective and less dangerous.
  • GlaxoSmithKline received a criminal conviction in 2012 for its deceptive practices.
• Pfizer’s drug trials on fluconazole as an antifungal medication for cancer patients.
  • Throughout the late 1990s, Pfizer sponsored numerous trials of fluconazole. Researchers wanted to re-analyze the data, but Pfizer refused to share it.
  • In 1999, researchers were able to re-analyze some of the data and found that the trials were incredibly biased.
  • The drug trials compared the new drug, fluconazole, to a drug that was currently in use, amphotericin b. However, when publishing the results, the researchers combined the amphotericin b group with a group taking nystatin, a drug known to be ineffective. Thus, the amphotericin b group appeared to do worse than they actually did. Other studies used an ineffective version of the amphotericin b medication to ensure that that group did more poorly.
• GlaxoSmithKline’s rosiglitazone (Avandia) for diabetes.
  • In 2004, after patients harmed by their medications sued the company, GlaxoSmithKline was forced to release data from its drug trials.
  • Researchers re-analyzed the data on Avandia and found that the drug was linked to a vast increase in cardiovascular events such as stroke, heart attack, and death.
  • In Europe, Avandia was taken off the market. It is still available in the United States.
• Pfizer’s antidepressant reboxetine (Edronax).
  • Edronax is a prime example of publication bias, in which negative results are buried, while positive results are published and promoted.
  • Based on the published studies, Edronax appears safe and effective.
  • However, when researchers analyzed the unpublished studies, they found that Edronax was neither safe nor effective.
  • Edronax is now available only in some countries, based on their own analyses of the published and unpublished studies of its efficacy. It did not receive FDA approval in the United States.
• Stockpiles of ineffective antiviral medicines for influenza treatment.
  • According to CRIT, “Across the globe, over $20 billion in public money has been spent on stockpiling Tamiflu and Relenza.”
  • In 2014, researchers with the Cochrane Group re-analyzed published and unpublished data on these two flu vaccines and found that they were both ineffective.
  • Both drugs “failed to prevent the spread of the flu, reduce admissions to the hospital, or minimize complications associated with the flu.”
  • The drugs were heavily promoted by pharmaceutical companies Roche and GlaxoSmithKline.
• Genentech’s cancer drug erlotinib (Tarceva).
  • According to CRIT, “In 2004, Genentech researchers had data indicating that the drug only worked for patients who had a particular gene mutation or who had never smoked. However, Genentech downplayed the mutation’s importance and discouraged patients from testing for the mutation.”
  • When a study in 2016 revealed that the drug was ineffective for up to 90% of the cancer patients who had taken it, Genentech paid $67 million to settle a civil suit and the FDA restricted the drug to the appropriate patients.

All these examples are cases in which more transparency might have reduced the ability of the pharmaceutical industry to mislead researchers, governments, and the public.

Current improvements to transparency, according to CRIT, include the World Health Organization (and many national organizations, such as the FDA), mandating that researchers register their trial before conducting it, on a database such as clinicaltrials.gov, which allows others to determine if outcome switching or other deceptive research methods were employed.

Journal editors have begun to request that researchers adhere to this requirement in order to be published. However, CRIT suggests that editors need to be firmer about this requirement and that governmental agencies penalize those who do not adhere—such as by removing grant funding if deceptive practices are found.

Beyond the lack of transparency, however, there is one factor in common to all these cases. Each was supported by the pharmaceutical industry. According to CRIT:

“Pharmaceutical companies are often reluctant to reveal negative information that could hurt sales. This can lead to uninformed decisions by patients, clinicians, and payers.”

Indeed, transparency may be the first step toward altering the system that incentivizes ineffective and risky drugs being passed off by the pharmaceutical industry as safe and effective. However, it seems a much larger task than transparency alone can accommodate.

 

****

Yale Collaboration for Research Integrity and Transparency (CRIT). (2017). Promoting transparency in clinical research: Why and how. New Haven, CT: Yale. (Link)