Ted P. Szatrowski, MD
To the Editor: Recently, Wilt and colleagues (1) concluded that fish oil supplementation fails to lower plasma cholesterol in hypercholesterolemic men. In a crossover study of 19 patients, no statistically significant (α = 0.10) difference in cholesterol levels was found in 38 treated subjects and 38 placebo controls. The authors implied that the study had greater than 99.9% power to detect a 5% drop (compared with placebo) in cholesterol level, suggesting a false-negative rate of less than 0.1%. The notion of type II error, or false-negative rate, in the setting of clinical trials has been reviewed (2).
However, power calculations
Szatrowski TP. Fish Oil Supplementation and Plasma Cholesterol: The Power of a Clinical Trial. Ann Intern Med. 1990;112:633–634. doi: 10.7326/0003-4819-112-8-633
Download citation file:
Published: Ann Intern Med. 1990;112(8):633-634.
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use