Instead, researchers found that for certain new rheumatoid arthritis medications, subjects in the comparison groups were often assigned to continue taking a drug that didn't help them or to take a fake drug called a placebo -- in both cases effectively depriving those patients of treatment for their disease.

"Of course it's easier to compare to a placebo than an active treatment but in no way can it justify exposing patients to irreversible morbidity," said Dr. Candice Estellat at the French national medical research institute, INSERM, who led the study.

Her concern is that if studies don't compare a new drug to other effective ones, people's conditions will persist or even worsen throughout the study.

Additionally, without so-called head-to-head trials, doctors will have little evidence to go on to determine whether one treatment is better than another, Estellat said.

Rheumatoid arthritis is an autoimmune disease that causes inflammation and damage to joints.

About 1.5 million adults have the painful disorder, and they typically require lifelong treatment with physical therapy or medications, such as methotrexate.

The newest forms of rheumatoid arthritis medications to become available are called biologic disease-modifying antirheumatic drugs (DMARDs).

These include the brand-name medications Enbrel and Humira. They come in the form of an injection, and cost around $15,000 per year.

Estellat and a colleague gathered information on all clinical trials of biologic DMARDs registered with the U.S. government's clinicaltrials.gov website and that were ongoing between 2002 and 2009.

She said they decided to do the study after hearing from specialists about the lack studies comparing these new drugs against other biologic DMARDs.

Of the 91 trials they identified, just five studies compared one biologic medication to another.

"Unfortunately we were not surprised as it confirms rheumatologists' feeling(s)," Estellat said.

The remaining trials will not provide doctors and patients with information for making evidence-based decisions, she and her coauthor write in the Archives of Internal Medicine.

"There (are) plenty of studies to show that A is better than placebo, B is better than placebo, C is better than placebo but so few to know which one is the best between A, B or C," Estellat told Reuters Health by email.

Not only are placebo-compared experiments less useful in understanding how well a drug works compared to others, but, the report points out, they may violate international ethical research standards and specific guidelines from the American College of Rheumatology if people are not given treatment for a serious condition like rheumatoid arthritis.

The 91 trials included 102 comparisons of a biologic medication against something else -- in most cases that something else was a placebo or a treatment previously shown not to work for them.

"Despite recommendations to give biologic treatments to patients with an inadequate response to conventional treatment, 9,879 patients were or will be randomized to control arms to receive no treatment or their previous ineffective treatment," Estellat said.

Studies on humans have to go through ethical scrutiny by independent bodies called institutional review boards, and in the U.S. they must also be approved by the Food and Drug Administration.

A spokesperson for the pharmaceutical trade group, PhRMA, wrote in an email to Reuters Health that "much of that decision-making is done under the guidance of FDA, whose experts are able to work with companies to evaluate the pros and cons of different types of trials."

Dr. Steven Pearson, the president of the Institute for Clinical and Economic Review in Boston, explained the possible reasons for not using head-to-head trials for certain drugs in an editorial accompanying Estellat's study.

He agreed that using placebos is a less desirable trial design to ultimately help physicians determine which medication they should prescribe for their patients, but said it's a trade-off for making an experiment less difficult.

"For one, having an active agent against a comparator would require many more patients," he told Reuters Health.

"In order to be able to get clear signals of the safety and effectiveness of new agents it's going to be easiest to compare it to a placebo, in terms of costs and duration," Pearson added.

To be most useful to patients and physicians, clinical trials need to compare one drug to another, and Pearson said that companies, regulators and researchers should think of creative ways to do so without having the studies become prohibitively expensive or unwieldy.

"I think the study is helpful to (the Food and Drug Administration) and others to take stock and see if there are other innovative study designs and approaches that allow more head-to-head trials," Pearson said.

Estellat said that people involved in clinical studies "have to imagine original designs to conciliate ethics and scientific requirements."

SOURCE: bit.ly/yh1orf Archives of Internal Medicine, February 13, 2012.

• A Permanent Economic Emergency