The Use of Placebo in Clinical Trials for Mood Disorders: 
A Consumer Perspective
Yale Medical School Department of Psychiatry - Grand Rounds
February, 2001
Lydia Lewis, Executive Director of DBSA

• I’m very honored to be here today representing the patient point of view. It’s very special to be asked to speak to doctors. We often feel invisible, powerless and unimportant.

• Last spring, I was asked by Steve Hyman, the director of the National Institute of Mental Health, to speak about the recruitment of volunteers into clinical trials. Specifically, I was asked to outline what advocacy groups can do to promote participation. As I was thinking about what I was going to say, I started feeling real uneasy. Being asked to promote clinical trials when we know protections aren’t always adequate is tough stuff for us.

• On the one hand, mood disorder research is desperately needed. In fact, advocating for research is one element of our mission. But on the other hand, trial participation can leave a person quite vulnerable.

• We know more research is needed. And we also know that research can’t be conducted without human subjects. So here’s the challenge for DBSA: How can we help protect our constituents?

• We all know the issues:

• Exposure to placebo

• Informed consent that doesn’t inform

• The push to enroll "subjects" at the expense of patient protection

• Poor safety monitoring procedures

• Challenge studies

• Drug washouts

• Poorly trained IRBs or IRBs with inadequate consumer representation
  

• So what can we, as advocates, do to achieve this balance? Research or protection? Both are needed. How can we insure both happen? We realize there’s no way risk can be completely eliminated in clinical research. However, we all need to fight the fight for patient protection while promoting patient participation in clinical trials.

• When media coverage of psychiatric clinical trials became increasingly negative, DBSA decided to convene a consensus conference on the use of placebo in clinical trials for mood disorders. The 37-member Consensus Development Panel consisted of experts in psychiatry, psychology, bioethics, biostatistics, and patient advocacy. In 1 ½ days, this group tackled some really tough questions.

• Our panel began its discussion with the question: Is the use of placebo ethical?

• While the recently revised Declaration of Helsinki argues that placebo-controlled trials are unethical when effective treatments exist and stipulates that new treatments should always be tested against the best current treatment, our panel believes that a blanket policy of disallowing placebo-controlled trials because standard, effective treatments are available could have far-reaching negative implications. It concluded that yes, the use of placebo in clinical trials for mood disorders can be ethical, even when treatment already exists.

• In addition to the minimum ethical standards that apply for other clinical research, the risks of administering a placebo rather than an "active" treatment must be weighed against anticipated benefits, including the knowledge gained that can improve future treatment.

• The use of placebo must be scientifically necessary to test the hypotheses or to enable substantially greater efficiency than alternative research designs. Our panel also concluded that all protocols reviewed by IRBs should justify placebo use. The use of placebo is ethical when the need for information justifies the risk.

• While the use of placebo is ethically problematic when effective treatments exist, several considerations significantly reduce ethical concerns when studying mood disorders. These include:

• The fact that mood disorders are chronic, episodic conditions. There is no cure.

• That people are at increased risk of suicide when incorrectly diagnosed or inadequately treated

• The high rate of placebo response

• Existing medications provide only partial symptom relief for so many of us and have lousy side effects that can significantly affect quality of life

• The FDA database of nearly 20,000 patients didn’t find the rates of completed and attempted suicide significantly different for placebo, standard antidepressants or antidepressants under investigation.

• The use of placebo is not ethical when there’s inadequate protection from serious risk or when treatments with a high degree of proven efficacy and tolerable side effect profiles become available. As the possibility of harm or serious discomfort increases, additional protections must be provided in order to ensure that a placebo-controlled trial is conducted ethically.

• However, as I mentioned earlier, our panel felt a policy of not allowing placebo-controlled trials could have widespread implications. In the absence of placebo controls, active-controlled trials that are not properly designed to demonstrate between-group differences may produce unreliable results or may falsely lead to claims of equivalence between experimental and standard drugs. Also, clinical trials without placebo controls probably would require recruitment of more people, exposing them unnecessarily to experimental drugs that may prove ineffective or have difficult side effects or toxicity.

• There are special informed consent considerations for placebo-controlled trials. So many people don’t understand that research is not intended to provide individualized care. In addition, there are some important informed consent requirements specific to placebo controlled trials. The informed consent document must be written so that people are advised not to participate if they are unwilling to accept the risks associated with not receiving an active treatment.

• What a placebo is must be explained and it should be made clear that they might be randomly assigned to placebo. They should also be informed that the researcher, as well as themselves, will probably not know if they are receiving an active treatment or placebo. The objective of the study and its design should be clearly explained, especially the inclusion/exclusion criteria, masking procedures, and the use of medication washout, placebo, medication discontinuation and specific interventions. All this must be communicated in language we can understand.

• Currently, information about the extent of disability associated with placebo compared with standard treatment isn’t made available after a study has concluded. Our panel is recommending that future studies include outcome analyses for both placebo and active treatment.
  
• Monitoring clinical status is essential, even if the patient is no longer following the study medication regimen. Without post-adherence monitoring, it’s not possible to measure the consequences of placebo use.

• Research protocols must include frequent symptom assessments with clear mechanisms for identifying the need for rescue, hospitalization, medications or other interventions.

• In studies where an acute trial is followed by a medication discontinuation phase, informed consent should be re-administered when the study calls for patients to be randomized to either continued treatment or placebo. The likelihood, severity, and consequences of a relapse should be clearly explained. And alternatives to entering a discontinuation phase, such as continuing on a treatment that appears to be effective, must be made clear.

• It’s debatable how long free or transitional care should be available. Indefinite or permanent care isn’t appropriate possibly on ethical grounds because it may serve as a coercive incentive. There is one exception -- the case of an unexpected devastating outcome. In this situation, appropriate care must be given as long as necessary.

• Our panel agreed that delaying active treatment during a placebo-controlled trial is acceptable, assuming reasonable clinical safeguards are in place. However, there are ethical concerns if someone who is resistant to standard therapy never gets access to a drug that turns out to be superior, or if people are required to wait until the end of study before the blind is broken.
  
• Our panel recognizes the legitimate scientific concern in breaking the blind as patients leave the trial. One possible alternative might be to use a nonresearch clinician acting as a liaison with patients in this situation.

• It’s important we address issues voiced by critics of psychiatric research. We can do this by including certain details in published materials.

• The use of placebo controls should be justified in writing.

• Details about the safety monitoring process should be included.

• Our panel felt researchers should always document ethics safeguards. Relevant information such as what was the inclusion/exclusion criteria and how many people decided not to participate should be included in all published articles of randomized, controlled trials.

• As a patient advocacy organization, the ethical issues raised by placebo controlled trials are among DBSA’s greatest concerns and we hope our paper will help increase the protection of human subjects. Please don’t forget we are desperate for better treatments but also don’t forget to make patient protection of paramount importance.