What is a prospective study?

A prospective study is set up before the outcomes have happened, and data are gathered going forward. For diagnostic accuracy studies, STARD 2015 calls a study prospective if the plan for data collection preceded the tests; the STROBE explanatory paper reports a usage in which the split rests on when data were collected relative to when the study idea arose. According to STARD’s explanation, when the question is set before the tests, procedures and data collection can be tailored to the protocol.

In a prospective cohort, Setia writes, the investigator defines the cohort, measures exposure and potential confounders at baseline, then follows participants and assesses the outcome. Measurements can be more accurate and uniform, which helps with exposures that depend on recall or subjective assessment. Setia lists time and cost as limitations and loss to follow-up as a source of bias. Euser and colleagues summarize the trade-off as accuracy for prospective cohorts and efficiency for retrospective ones.

Setia notes that case-control participants can also be recruited prospectively, so “prospective” should not be treated as a synonym for “cohort”, although some authors use it that way (STROBE). Nor does it mean experimental: in ICMJE’s definition a clinical trial assigns an intervention to participants going forward, and ICMJE encourages, but does not require, registration of non-trial studies in which the researchers do not determine the exposure.