Complement Ther Med. 2001 Dec; 9(4): 234-6.

How to reduce the number of patients needed for randomized trials: a basic introduction.

Vickers A.

Integrative Medicine Service, Memorial Sloan-Kettering Cancer Center, New York, New York 10021, USA. vickersa@mskcc.org

Complementary therapies are often used to treat patients with persisting conditions. In a typical randomized trial, patients are assessed for symptom severity at baseline, randomized to treatment or control and then reassessed after a suitable follow-up period. It can be shown that the number of patients required for such trials can be reduced by the use of two methods: repeat measurement and analysis of co-variance (ANCOVA). An example is given of a trial of acupuncture for back pain. Administering the pain questionnaire twice at baseline and four times at follow-up reduces the number of patients by approximately 20% compared to single questionnaire administration. Use of ANCOVA decreases the number of patients by approximately a further 10%. There are considerable economic and ethical advantages to reducing sample size by a third; however, ANCOVA is statistically complex and expert statistical help is required.