Study design for the evaluation of treatment.

Study design is a critical step in clinical research. Such studies may be either randomized or nonrandomized; the latter may be retrospective or prospective. Nonrandomized designs include surveys and observational studies. The randomized clinical trial (RCT) is an experimental design, with patients randomly allocated to treatment arms. A number of designs are possible. These include the simple experiment, repeated measure design, crossover study, and factorial design. Issues in clinical trials include subject inclusion/exclusion criteria, control group treatment, blinding, compliance, sample size, treatment assignment, and subject withdrawal. Disadvantages of the RCT include cost and the lack of true clinical equipoise. It is also worth noting that the RCT is not infallible. Patients should be analyzed based not on their compliance but rather on an intent-to-treat basis.