The three articles on placebo use in clinical care (February) provide a comprehensive and balanced perspective on this important topic. But we believe McCaffery and Arnstein (Pain Control, "The Debate Over Placebos in Pain Management") may have presented a different definition of an N-of-1 trial than the American Pain Society (APS) presents.1 This difference in definitions may have led the authors to perceive a difference in the positions on placebo use in clinical care of the APS and the American Society of Pain Management Nurses (ASPMN).

We initially thought that an N-of-1 trial was simply the use of trial and error in the care of a patient without institutional review board (IRB) oversight or patient consent. Based on the references cited in both the APS position statement and in the McCaffery and Arnstein article, this is not the case. The APS position statement1 cites two studies when it refers to N-of-1 trials.2, 3 Both of those studies are randomized, controlled N-of-1 trials with IRB approval and patient consent. Sheather-Reid and Cohen note that the main difference between a randomized, controlled trial and an N-of-1 randomized, controlled trial is that the former refers to a group while the latter refers to an individual.2 N-of-1 randomized, controlled trials require patient consent and IRB approval, and they involve randomization of treatment, blinding, and the use of predetermined outcome measures. They are especially useful in assessing clinical changes in patients with drugs of short duration of action.

We believe that if the above definition of an N-of-1 randomized, controlled trial were used, the APS, the ASPMN, and the Pain Control authors would agree that the N-of-1 randomized, controlled trial is a circumstance in which the use of a placebo may be ethical and appropriate.

Elizabeth Winslow, PhD, RN, FAAN

Rob Hutchison, PharmD

Dallas

REFERENCES