PT - JOURNAL ARTICLE AU - G. Guyatt AU - D. Sackett AU - J. Adachi AU - R. Roberts AU - J. Chong AU - D. Rosenbloom AU - J. Keller TI - A clinician's guide for conducting randomized trials in individual patients DP - 1988 Sep 15 TA - Canadian Medical Association Journal PG - 497--503 VI - 139 IP - 6 4099 - http://www.cmaj.ca/content/139/6/497.short 4100 - http://www.cmaj.ca/content/139/6/497.full SO - CMAJ1988 Sep 15; 139 AB - In determining optimal treatment for a patient conventional trials of therapy are susceptible to bias. Large-scale randomized trials can provide only a partial guide and have not been or cannot be carried out for most clinical disorders. However, randomized controlled trials (RCTs) in individual patients (N of 1 RCTs) may in some circumstances provide a solution to this dilemma. In an N of 1 RCT a patient undergoes pairs of treatment periods (one period of each pair with the active drug and one with matched placebo, assigned at random); both the patient and the clinician are blind to allocation, and treatment targets are monitored. N of 1 RCTs are useful for chronic, stable conditions for which the proposed treatment, which has a rapid onset of action and ceases to act soon after it is discontinued, has shown promise in an open trial of therapy. The monitoring of treatment targets usually includes quantitative measurement of the patient's symptoms with the use of simple patient diaries or questionnaires. Pairs of treatment periods are continued until effectiveness is proved or refuted. The cooperation of a pharmacy is required for the preparation of matching placebos and conduct of the trial. Formal statistical analysis may be helpful for interpreting the results. The practical approach presented in this paper allows clinicians to conduct their own N of 1 RCTs.