Controlled clinical trials

Alcohol Health & Research World,  Fall, 1990  by Richard K. Fuller

• E-mail
• Print
• Link

Clinical trials are scientific experiments, conducted by observing or treating actual patients in realistic conditions to determine the effectiveness of a treatment or combination of treatments for preventing or ameliorating disease. The term "clinical" distinguishes such experiments from those conducted in laboratories; the term "trial" indicates an attempt to determine a treatment's effects--or range of effects--in a manner that is equitable, or unbiased by external factors.

Lillienfield (1982) identified three types of clinical trials:

* Therapeutic trials, which test the effectiveness of a measure intended to relieve symptoms of a disease or to improve survivorship, or both--trials of new medications for treating human immunodeficiency virus (HIV) infection, for example.

* Intervention trials, which test the effectiveness of a measure intended to reduce the sequelae, or consequences, of a disease or condition before they occur--trials of antihypertensive medications intended to counteract the likelihood of occurrence of heart attack or stroke, for example.

* Prophylactic trials, which assess the effectiveness of a strategy intended to prevent a disease in persons who are at risk for developing the disease--trials of polio vaccine, for example.

Although derived primarily from the field of medicine, clinical trials are used to evaluate the effects of psychological and behavioral, as well as pharmacological, treatments.

Researchers use a number of standard methodological approaches to ensure that a clinical trial tests the treatment intended. Although none of these approaches is new to scientific evaluation or unique to clinical trials, several are considered critical to well-designed clinical trials. The most important feature of clinical-trial methodology is the use of control groups. As evidenced by the title of this article, the terms "clinical trial" and "controlled clinical trial" are used interchangeably in today's vernacular.

Clinical trials differ in one important respect from laboratory experiments: Because clinical trials are conducted in clinics or inpatient settings, rather than in laboratory settings, clinical investigators may have difficulty controlling all the factors that may bias the trial results. To reduce the possibility that factors other than the experimental treatment are responsible for the trial results, clinical investigators sometimes employ study designs that vary from customary clinical practice. For example, in an "explanatory" or "fastidious" trial, an investigator tests the effects of a new medication only in patients who are known to take medication as prescribed (compliant patients), excluding patients who may not take the medication as prescribed (noncompliant patients); the investigator thus can avoid attributing to the medication the effects observed in noncompliant patients. However, in a "management" or "pragmatic" trial, the investigator enrolls patients regardless of whether they are known to be compliant; the investigator thus can test the medication's general efficacy for a typical office or clinic practice.

Researchers who conduct clinical trials have been trained in a discipline that involves treating patients, such as medicine or psychology, and in experimental methodology. The researchers work closely with statisticians, who assist in designing the study and analyzing the study results. Participants usually are recruited from the hospitals or clinics where the researchers work; however, if necessary, additional participants may be recruited through advertising.

Although clinical-trial methodology continues to evolve, the trials themselves have become an essential tool of researchers in multiple disciplines, including alcohol research. This article reviews the methodological approaches that today are considered essential to designing and conducting clinical trials (DerSimonian et al. 1982) and discusses the importance of these trials in alcohol treatment research.

USING CONTROLS

Clinical trials are conducted using an experimental group, comprising persons who are administered the treatment being tested, and a comparison, or control, group, comprising persons who do not receive the treatment. Members of a control group receive either no treatment or the treatment that is considered the best available treatment at the time the study is conducted. When the members of a control group receive no treatment, the experimental therapy is tested against the natural course of a disease. However, because it is considered unethical in cases of serious illness to withhold the best proven therapy, clinical trials that use seriously ill patients usually test a new treatment against the best available treatment.

Investigators sometimes use "historical" controls, that is, persons studied at some time before the experimental therapy was developed, to reduce the time and expense involved in a clinical trial. Although this practice simplifies the process of conducting a clinical trial, it renders questionable the conclusions drawn from the results (Byar et al. 1976).