- 51 Downloads
A study in which the outcome of interest has already occurred when the study initiated is commonly referred to as retrospective study.
Some investigators have used the terms retrospective and case-control interchangeably. This usage is misleading since study designs other than case-control studies can be retrospective. For example, in a retrospective cohort study (Okasha et al. 2002) examining the relationship between body mass index (BMI) during early adulthood and cancer death, both the exposure and the outcome had occurred at the initiation of the study. University students’ weight and height information was obtained from annual physical records at the University of Glasgow health service from 1948 to 1968. Cancer mortality was determined based on a central registry in Scotland that identifies the cause of death (National Health Service Central Register). Subjects with BMI in the highest quartile were found to be at higher risk of cancer death than individuals in the lowest BMI quartile.
In a retrospective study, data may be collected based on clinical records, employment records, or memory. These retrospective data can be more prone to bias, such as recall bias, than prospective designs studying the same association.
Some investigators may also use the term “retrospective” in a different way, referring to the timing of the measurements of the exposure and outcome; studies in which the exposure is measured after the outcome is measured can be referred to as retrospective (Rothman and Greenland 1998, pp. 74–75).