Skip to main content
. Author manuscript; available in PMC: 2019 Nov 27.
Published in final edited form as: J Biopharm Stat. 2015;25(4):812–829. doi: 10.1080/10543406.2014.920858

Table 4.

Probability of a meaningful decrease, power to detect a treatment effect based on this dichotomization with n=64 in each treatment group in a two-arm parallel study, and the sample size in each group to have 80% power to detect a such treatment effect, when the correlation between baseline and follow-up measurements is 0.6, by mean baseline CVL HIV-1 and mean treatment effect.

Mean treatment effect Baseline mean
2.0 2.5 3.0
Pr(decrease) Power (%) N per arm Pr(decrease) Power (%) N per arm Pr(decrease) Power (%) N per arm
 
0.50 0.367 28 217 0.446 44 135 0.486 56 104
0.75 0.421 51 115 0.525 77 67 0.583 90 50
1.00 0.464 70 79 0.585 94 44 0.668 99 31

A meaningful decrease is a decrease of at least 0.5 if both measurements are quantifiable, or a decrease from quantifiable to non-quantifiable.

Quantification limit is 1.9; standard deviation of both baseline and follow-up values is 1.

The sample sizes (N per arm) were obtained from the web site http://www.swogstat.org/stat/public/binomial_twoarm.htm