Experimental Design

An experiment is a study designed to compare benefits of an intervention with standard treatments, or no treatment, such as a new drug therapy or prevention program, or to show cause and effect (see ). This type of study is performed prospectively. Subjects are selected from a study population, assigned to the various study groups, and monitored over time to determine the outcomes that occur and are produced by the new drug therapy, treatment, or intervention.

Experimental designs have numerous advantages compared with other epidemiological methods. Randomization, when used, tends to balance confounding variables across the various study groups, especially variables that might be associated with changes in the disease state or the outcome of the intervention under study. Detailed information and data are collected at the beginning of an experimental study to develop a baseline; this same type of information also is collected at specified follow-up periods throughout the study. The investigators have control over variables such as the dose or degree of intervention. The blinding process reduces distortion in assessment. And, of great value, and not possible with other methods, is the testing of hypotheses. Most important, this design is the only real test of cause–effect relationships.

The disadvantages of experimental design involve subject participation criteria that may limit generalizability of findings. Restrictive criteria for inclusion or exclusion of subjects may produce a very homogeneous study population that restricts application of the results to patients with other characteristics. Clinical trials, especially those focused on chronic diseases, may require years of follow-up and prolonged observation to determine treatment outcomes. The result is often higher costs, increased likelihood that patients will be lost to follow-up, and delayed treatment recommendations. Large sample sizes are typically required to demonstrate differences among study groups, especially if there is wide variability in responses to treatment. Increasing the size of the study population also raises the cost of the trial and may make it difficult to locate a sufficiently large pool of eligible patients. Ethical concerns also arise in clinical trials, and subjects may not comply with the treatment and assignment.

Clinical trials can be divided into three types: (1) therapeutic trials in which therapeutic agents or procedures are given to patients in an attempt to cure the disease, relieve symptoms, or prolong survival; (2) intervention trials in which the investigator intervenes before the disease has developed in individuals with certain characteristics that increase their risk of developing the disease; and (3) prevention trials in which an attempt is made to determine the efficacy of a preventive agent or procedure among people who do not have the disease but may be at greater risk for developing it. The randomized, controlled clinical trial is the most widely accepted approach for comparing the benefits of treatments. The basic design of a randomized, controlled clinical trial is outlined in . Community intervention trials have the same basic design; the difference is that groups of people are assigned to the various study groups.