Hepatitis Central

Clinical Trials:
Difficult, imperfect, expensive--and necessary
September 12 1996

On February 13, 1600, four East India Shipping Company vessels set sail. By chance, only one had lemons aboard. Very few men on the ship with lemons came down with scurvy; most men aboard the other three ships did. Company officials noticed that supplying sailors with lemon juice seemed to make the difference. Still, the practice did not catch on. The British Navy, for example, did not supply citrus juice aboard its ships until nearly 200 years later.

In modern times, discoveries about human health need not rely on such chance observations. Carefully designed clinical trials can lead to better medical treatments and discredit ineffective ones.

Here are some basics about clinical trials:

•Definition: A clinical trial is a carefully designed experiment designed to evaluate the benefits and risks of treatments on humans. Comparable groups of people receive different forms of treatments or a treatment versus a placebo so researchers can compare outcomes.

•Justifying the effort: Before beginning a clinical trial, scientists need to determine whether the problem under consideration justifies the potential risks as well as the costs in time and money. Reviewing relevant past studies is a starting point.

•Arranging the finances: Potential sources of support for clinical trials include private endowments, pharmaceutical companies and government grants or contracts.

•Defining the sample: Generally, a larger sample size lessens the likelihood of reaching inappropriate conclusions. However, when a drug shows effectiveness against a rapidly fatal condition (such as an antibiotic against some forms of bacterial meningitis), researchers may not need a large sample of people to know that the treatment is effective.

•Getting informed consent: Ethical standards require that participants in a clinical trial be fully informed of potential risks and benefits as well as the availability of alternative treatments. There can be no penalties for declining to participate or withdrawing at any time.

•Randomizing: The patient groups should be as alike as possible in age, gender distribution, medical history and other characteristics. Randomizing methods can minimize potential bias in the assignment of individuals to each treatment group.

•Blinding: Keeping patients unaware of which treatment they receive helps prevent bias in reporting or evaluating results. "Double-blinding" means that during treatment, both the study participants and the caregivers do not know who is assigned to each group.

•Controlling with placebos: It's often useful to give a "control" group an inactive treatment to measure against active treatment(s). However, in some circumstances it is unethical to give placebos to people, such as those whose medical condition requires some form of available therapy.

•Setting protocols: Clinical studies must establish and follow specific, painstaking sets of procedures so that the studies can be replicated by others. If, for example, a study were to use two different measures to assess outcomes, it could confuse the results.

•Multiple research centers: Most large clinical trials are conducted simultaneously at several research centers. The benefits of multicenter trials include larger numbers of participants, different geographic locations and ethnic groups, and the ability to compare results between centers.

•Generalizing results: The results and conclusions of a study on men may not apply equally to women. Researchers must take care not to generalize their results too broadly.

•Monitoring ethics: The government and other funding agencies require safety and monitoring boards to oversee clinical trials in progress.

•Checking the numbers: Accepted methods of mathematical analysis can determine whether a difference in treatment outcomes is statistically significant. A statistically significant difference means that the result is very unlikely to be due to chance alone.

•Peer review: Published reports of clinical trials are usually reviewed by scientists uninvolved with the study before findings are made public. When the work is either published or presented at a conference, the researchers open themselves up to critical review by the scientific community at large.

Limits and values of clinical trials

Because we're all human, some study participants may forget to take medications as directed or may report their symptoms inconsistently. Constraints on time, money and patient participation also can limit the conclusions of a particular study. Scientists, meanwhile, debate constantly over whether research methods influence results. Despite imperfections, the ever-evolving database of clinical trials is often the best way for doctors to evaluate many treatments.

Dr. Kenneth G. Berge, a Mayo Health O@sis associate medical editor who has held leadership roles in several major clinical trials in the area of cardiovascular disease and hypertension, says clinical trials are in some cases the only way to quantitatively answer questions such as whether the effectiveness of a drug outweighs its undesired effects.

Dr. Berge says that because clinical trials can influence the health of participating patients, all such research should begin with a fundamental ethical question: "Do the potential benefits outweigh potential harm?"

table Of Contents