Rare diseases present a number of problems for clinicians, patients and biostatisticians. They challenge clinicians with the lack of treatments and awkwardness of placebos. They challenge patients with the slow pace of research and they challenge statisticians with the need to use small samples and under powered studies. For many statisticians working in more prevalent diseases, the designs and issues in rare diseases and small samples were uninteresting as you could simply count the results and you had little power in the classical sense. However as the mantra of personalized medicine moves forward, all of us will be working with rare diseases!
The example presented here addresses a design for a placebo controlled trial with 3 to 1 randomization for ethical reasons; a truncation of the follow-up period to limit exposure to a placebo and a sample size readjustment and futility analysis all designed to assuage the clinicians concerns, the need to be ethical to the patient's perspective while addressing the safety and effectiveness of a new drug in a serious neurological disease while minimizing the occurrence of events from a population perspective.
|