Retrospective Study

Retrospective studies are epidemiologic studies characterized by an exposure or outcome of interest that has occurred in the past. A retrospective study design enables an investigator or epidemiologist to identify possible associations between exposures and outcomes, but generally not causal relationships. The temporal relationship between exposure and outcome is an essential criterion for judging a causal relationship. Because the events of interest have occurred prior to beginning a retrospective study, biases may affect assessment of both the occurrence and timing of such events. However, retrospective studies are generally less costly and less time consuming than prospective studies. The two most common types of retrospective studies are retrospective case-control studies and retrospective cohort studies.

Retrospective case-control studies are characterized by the identification of cases prior to the beginning of the study. Within the retrospective case-control study design, people with the disease of interest (cases) are compared by exposure status to people without the disease of interest (controls). Advantages of retrospective case-control studies include a relatively short time period required for study, a relatively inexpensive study design, and use of a small population. Disadvantages of retrospective case-control studies include occasional difficulty in establishment of a temporal relationship, potential bias in assessment of exposure, occasional difficulty in selection of appropriate controls, incomplete information on exposure, limitation in human recall, and potential recall bias

A retrospective cohort study is also called a historical cohort or a nonconcurrent prospective study. The study designs of both the concurrent cohort study and the retrospective cohort study are identical, focusing on a comparison of incidence of disease in an exposed and a nonexposed population. However, in a retrospective cohort study design, historical data are used from the past to reduce the time frame of the study to obtain results sooner. Within the retrospective cohort study design, exposure is usually ascertained from historical records or interviews. Advantages of retrospective cohort studies include less expense than a concurrent cohort study and less time required for study. Disadvantages of retrospective cohort studies include occasional difficulty in establishment of a temporal relationship, susceptibility to bias both in assessment of exposure and outcome, information bias in quality and extent of information obtained from exposed and nonexposed people, biases from nonresponse and loss to follow-up, difficulty in selection of nonexposed comparison group, and changes over time in criteria and methods.