The Best Laid Plans of Mice and Men

No abstract available. Article truncated after 150 words. When writing a grant proposal, many of us do a power analysis to ensure that we will have a sufficient number or “n” to detect a statistically significant difference between two populations. We estimate the number needed in each group by considering the likely intergroup difference and then add additional subjects depending on the number who will not give informed consent, refuse, die, are lost to follow up, etc. Often the number of nonparticipants is estimated based on previous experience, but sometimes a small study is done first called a feasibility study which tests the assumptions about recruitment. For both clinical trials and epidemiologic studies, a pilot or feasibility study also helps assure that participants will be representative of the relevant population (1). (For examples, will only the most seriously ill participate in a drug trial, or will the most vulnerable workers decline participation in a study. Will some drugs …