Can Drug Clinical Trials Be Trusted?

Drug Safety, Drug Efficacy, and Clinical Trials: Your Frequently Asked Questions

Medically Reviewed by Louise Chang, MD on April 15, 2008
From the WebMD Archives

April 15, 2008 -- Clinical trials are the gold standard for judging the benefits and risks of new drugs. But new revelations suggest this gold standard is sometimes adulterated.

Today's issue of The Journal of the American Medical Association suggests that a major sponsor of clinical trials, the drug company Merck, may have used misleading statistics to downplay the risks Vioxx might pose to people with Alzheimer's disease. Another article in The Journal of the American Medical Association suggests that clinical-trial and review articles about Vioxx often were written by company-hired ghostwriters and later signed by researchers who had little to do with the actual study or review.

Don't scapegoat Merck -- other pharmaceutical companies do the same thing, suggest Catherine D. DeAngelis, MD, MPH, and Phil B. Fontanarosa, MD, MBA. DeAngelis is editor-in-chief of The Journal of the American Medical Association; Fontanarosa is the executive deputy editor.

"Make no mistake -- the manipulation of study results, authors, editors, and reviewers is not the sole purview of one company," they write in a scathing editorial.

Can we still trust clinical trials? Are the drugs we're now taking safe? Here's WebMD's FAQ.

What do clinical trials test?

When a drug finally makes it through clinical trials and FDA approval, here's what we know about it:

  • The new drug, in combination with standard treatment, is more helpful than standard treatment alone.
  • If the trial tested one drug against another, the new drug is at least as good as the other.
  • Benefits of the new drug -- for narrowly defined groups of patients -- outweigh the drug's risks.

But there are a lot of important things clinical trials usually don't tell us:

  • Clinical trials don't tell whether a new drug works better than existing drugs -- unless the trial compared one drug to another.
  • Clinical trials don't show that a drug is risk-free.
  • Clinical trials usually don't show whether a drug is safe to take in combination with other drugs, vitamins, or supplements.
  • Clinical trials don't show whether a drug might have unexpected long-term side effects.
  • Clinical trials don't show how well a drug might work in people, such as pregnant or breastfeeding women, who weren't included in the study.

How are clinical trials supposed to work?

The clinical trial system is designed first to look at whether a drug is safe enough for further testing (phase I clinical trials), then to look at whether a drug seems both safe and effective (phase II clinical trials), and then to compare the drug to an inactive placebo or standard treatment to see whether the drug is both safe and at least minimally effective (phase III clinical trials).

After a drug is shown to be safe and effective in phase III clinical trials, the drugmaker may ask the FDA for approval. The new drug does not have to be more effective than existing drugs, just more effective than placebo in defined patient groups. FDA approval, if granted, sets the conditions under which the drug can be sold.

In 2007, Congress passed a law (the Food and Drug Administration Amendments Act) allowing the FDA to make drug approval contingent on agreement to perform a post-approval clinical trial (sometimes called a phase IV clinical trial). This is done to evaluate possible long-term side effects. And, as before, all drugmakers must collect and send to the FDA patient and doctor reports of adverse events possibly related to taking the drug.

Phase III trials are the crux of the system. These trials can make or break a new drug. Virtually all of these extremely expensive trials are paid for by the companies that make the drug.

"The vast, vast, vast majority of clinical trials in this country are performed by pharmaceutical companies that must do these studies to get FDA approval," DeAngelis tells WebMD. "The government doesn't pay for them. That is not what NIH [National Institutes of Health] money is used for. Until we decide that it is worth it in this country to put many billions of dollars into clinical trials, then the drug companies will have to do it. And there is nothing wrong with that."

Drug companies don't just pay for clinical trials. User fees paid by drug and device manufacturers -- largely dedicated to speeding the approval process -- account for about 25% of the FDA's annual budget.

What roles do drug companies play in clinical trials?

Advocates of this system say it's only fair that companies that stand to benefit from sales of new drugs should underwrite the cost of testing them -- and of approving them in a timely manner.

Many of the most helpful medicines now in routine use emerged from drug-company-sponsored clinical trials. Drug companies "are committed" to making information from these trials available to patients and their doctors, says Ken Johnson, senior vice president of the Pharmaceutical Research and Manufacturers of America (PhRMA), the lobbying group that represents the drug industry.

"It is patently unfair to suggest that a manufacturer-sponsored clinical trial is inherently biased," Johnson says in a news release sent in response to the articles in The Journal of the American Medical Association. "PhRMA and its member companies believe that protecting the integrity of clinical research is paramount to patient safety."

Critics of the system say the drug industry's money has a corrupting influence on the clinical trial process. DeAngelis agrees with Johnson that there's nothing inherently wrong with drug companies paying for clinical trials. But she notes that companies focus more on maximizing profits than on maximizing public health.

"Fifteen years ago or so, pharmaceutical companies transferred most of their resources and emphasis into their marketing divisions rather than their science divisions," DeAngelis says. "There are relatively few new drugs or devices in the pipeline. Most of what they are doing now is testing drugs that are already out there to get approval for people who don't really need them."

Far worse, DeAngelis says, is companies sometimes put marketing ahead of science when reporting clinical trial results. Merck's apparent misrepresentation of data about deaths among Vioxx patients is just one example, she says.

Pharmaceutical companies "are doing their darndest to hide the side effects of the drugs they are developing," DeAngelis says. Everybody who uses a drug has to judge its cost -- not just the financial cost but cost in terms of side effects -- to understand its benefits. But you can't do that unless you have the data. Some of them are trying to suppress the data on side effects."

What roles do independent scientists play in clinical trials?

Drug companies employ many eminent scientists. But the companies usually ask outside scientists to serve as "principal investigators" for clinical trials. These scientists are responsible for conducting the trial in a scientific manner and for guaranteeing that the sponsor has not exerted undue influence either in conducting the trial or in reporting its results.

These outside scientists usually are prominent experts in their fields. For performing this work, sponsoring drug companies pay the university or institution with which the scientist is affiliated, says Margaret Dale, JD, dean for faculty and research integrity at Harvard Medical School.

"We do not accept contract language that restricts publication of study findings," Dale tells WebMD. "We have policies that relate to whether that investigator can have any financial interest in the sponsoring company, and their obligation to disclose it."

In addition, these experts often are paid to discuss the study findings at scientific meetings, symposia, advisory-board meetings, and drug-company-sponsored events. Dale says Harvard restricts investigators to $20,000 in such consulting fees -- and requires them to disclose this support in publications and scientific presentations.

Most journals require study authors to declare that they have received such payments. Dale says that journals are becoming more and more insistent that scientists disclose all consulting fees.

Troubled by the appearance of conflict of interest, a few prominent researchers have recently announced they will no longer accept payment for these services.

When clinical trials are over, the results appear as official reports in peer-reviewed medical journals. These journals have their own rules to ensure that the studies are scientific. The Journal of the American Medical Association even requires the statistical analysis of the study findings to be done by someone who does not work for the sponsor and who has an academic appointment.

When clinical trial studies are submitted to a journal, they undergo a process called peer review. During this process, draft versions of the manuscripts are sent -- anonymously -- to experts who were not involved in the study. These experts tell the journal whether, in their opinion, the study meets acceptable scientific standards. They often suggest changes or ask for more information and do not deem the manuscript acceptable until their questions are answered.

The peer review process seems to have failed to detect the statistical sleight of hand chronicled by The Journal of the American Medical Association. And, as the journal's studies suggest, journal editors permitted publication of study reports and review articles written by drug company employees and contractors that were signed by independent scientists who actually had little to do with the study's research or writing. Moreover, some of the company hires who worked on the study were not listed as investigators or authors.

"Drug companies shouldn't be paying people to ghostwrite a paper and then find an author who would prostitute him or herself -- that is the proper word, I believe -- to put their name and reputation on a paper for which they had very little to do," DeAngelis says.

Dale notes that Harvard forbids its scientists from being listed as study investigators unless they made substantial contributions to the work. And she says the school also insists that everyone who made substantial contributions to a study must be listed as authors.

Richard Bookman, PhD, executive dean for research at the University of Miami, says many medical schools and academic research institutions are tightening their policies so that potential conflicts of interest always are declared.

What's wrong with the clinical trial process?

Today's studies in The Journal of the American Medical Association suggest -- not for the first time -- that not all clinical trials are designed, conducted, analyzed, or reported ethically.

DeAngelis says it's wrong to blame the drug companies. The misdeeds suggested by the report "could not occur without the cooperation (active and tacit) of clinical researchers, other authors, journal editors, peer reviewers, and the FDA," DeAngelis and Fontanarosa write.

Directly asked about this allegation by WebMD, the FDA chose not to respond.

How can the process be fixed? DeAngelis and Fontanarosa detail an 11-point program they say is a starting point. Their recommendations include:

  • There should be a registry for all clinical trials that clearly shows -- before patients enter the study -- the names of the principal investigators.
  • The statistician who analyzes study results should not be employed by a for-profit company.
  • All study investigators should fully disclose all pertinent relationships with for-profit companies.
  • All of the investigators listed as study authors should make substantial contributions to the work -- and all investigators who make substantial contributions should be acknowledged.

The medical research community already is taking action, Bookman says.

"It is absolutely clear that in academic medicine, awareness of these issues has mushroomed over the last five years," Bookman tells WebMD. "The pendulum has swung too far one way, but it is already returning. It is clear that the academic medical community has waked to the fact they have to look at this more carefully. Academic medicine in five years will be transformed with much more complete conflict-of-interest disclosure policies and much more management capability to avoid conflict of interest."

Show Sources


DeAngelis, C.D. The Journal of the American Medical Association, April 16, 2008; vol 299: pp 1833-1835.

Ross, J.S. TheJournal of the American Medical Association, April 16, 2008; vol 299: pp 1800-1813.

Psaty, B.M. and Kronmal, R.A. The Journal of the American Medical Association, April 16, 2008; vol 299: pp 1813-1817.

Memorandum, "Food and Drug Administration Act Amendments Act of 2007," Hyman, Phelps & McNamara, P.C., Oct. 11, 2007; accessed April 14, 2008.

News release, Pharmaceutical Research and Manufacturers of America, April 14, 2008.

Catherine D. DeAngelis, MD, MPH, editor-in-chief, The Journal of the American Medical Association.

Margaret Dale, JD, dean for faculty and research integrity, Harvard Medical School.

Richard Bookman, PhD, executive dean for research, University of Miami.

© 2008 WebMD, LLC. All rights reserved. View privacy policy and trust info