BRS - Subjects

BRS - How Do I Get Subjects?

To recruit subjects, think about the type of INFERENCES you want to make from your study (and to whom you want the study results to apply). Your study subjects should look as much as possible like the population of patients you hope to apply the results to - otherwise, the study results might be useless in the real world.

This selection is accomplished by INCLUSION and EXCLUSION CRITERIA. These are rules that cover who gets into the study and who should not. The more restrictive these rules are, the less widely applicable the results will be. The less restrictive the rules, the more chance there is of confounding or of interaction with the intervention, which may result in a study that shows negative results despite an true benefit.

The SETTING of the study is important to maximize generalizability. The natural history of a condition may differ significantly between a tertiary care setting vs. a primary care setting, between an urban and rural setting, and possibly between a US study and a study from another country.

The RECRUITMENT METHOD is important to avoid bias. For most studies, once you define the setting, you must recruit subjects in a way that doesn't result in limited validity or generalizability. If you're studying the Emergency Department treatment of a condition, then recruiting from 9am to 5pm on weekdays doesn't really give you the same range of patients that an ED sees over the course of a week - so the study's results may not generalize to real-world, 24-7 Emergency Medicine. The best recruitment methods are either consecutive recruitment - where every eligible patient in a period of time is screened for inclusion - or sampling - whereby a random sample of eligible subjects are screened

Finally, for a randomized controlled trial, it's important to ensure ALLOCATION CONCEALMENT. The investigator that is recruiting subjects should not know the group to which they will be assigned. Otherwise, they may bias the study by deciding for or against recruitment of a subject based on the group they'll be in rather than based on the inclusion/exclusion criteria.