Were the patients randomized into treatment and non-treatment groups?
Studies in which treatment is allocated by any method other than
randomization tend to show larger (and frequently false-positive)
treatment effects than do randomized trials. The beauty of
randomization is that it assures, if sample size is sufficiently
large, that both known and unknown determinants of outcome are
evenly distributed between treatment and control groups.
(Guyatt, Sackett, et al)
Application to Pollock et al's
thyroxine paper
In this study, randomization of patients to treatment group or placebo
was by "toss of a coin in batches of four".
Is this appropriate with such small patient numbers? What if a whole
family or group of close friends were placed into the same patient group?
The "crossover" feature of this study should reduce the effects of some
of this potential bias.