• Shinichiro Ueda.
• Department of Clinical Pharmacology & Therapeutics, University of the Ryukyus, Okinawa 903-0215.
• Masui. 2011 Mar 1; 60 (3): 277-84.

AbstractAppropriate and careful interpretation of results from clinical trials is essential for evidence-based medicine. Clinical trials for approval of new drugs and new indications are rigidly designed and conducted in highly selected patients for the safe and strict assessment of efficacy of drugs, and is associated with high internal validity but low generalizability of results. On the other hand, pragmatic trials for the assessment of effectiveness of treatment with approved drugs use less rigid design, for example, unmasked trial treatment and less strict selection of patients, and may be associated with high external but low internal validity. Recently several problems associated with pharmaceutical company led trials have been claimed including early termination of trials based upon interim analyses, leading to overestimation of results. Composite endpoints are used to potentiate statistical power and capture the overall impact of intervention in recent trials. However, they may be misled if component end points are of widely differing objectivity and importance to patients. Less objective end points in unmasked trials may also lead to overestimation of effects of intervention. There are many pitfalls in terms of the interpretation of the results from clinical trials, and they must be recognized for evidence-based medicine.

4 keywords...

hide…