ABSTRACT
Randomised clinical trials (RCTs) are the standard method to assess clinical effectiveness [1] and are, with meta-analyses, the standard of excellence for comparisons of treatment effects over time. But clinical questions are most easily answered when a disease is fairly common and the outcome of interest has a high risk of occurring. On the contrary, when diseases are rare, or recruitment of patients is difficult, with modest benefits, clinical trials may not provide a “definite answer”. Indeed, they cannot be expected to reach conventional levels of power and statistical precision. Thus, investigators who wish to test new treatments for rare diseases, tend to conduct either single arm studies or comparative studies using historical controls. Alternatively, investigators may attempt to conduct small, underpowered RCTs. These give rise to estimates of outcome that have unacceptably large confidence intervals and this fails to provide clear answers.
