Establishing a solid research question and appropriate methodology for selecting study participants forms the basis for good study design. For example, the diminished reputation of retrospective studies can be traced to problems with defining the research question and study population. In retrospective studies, one sets out to use data collected in the past, usually for a purpose other than the research study being executed. Therefore, the investigator does not have the opportunity to tailor the data collection to the research question or control the selection of study participants. The potential mismatch between the collected data, the intended research question, and the intended study population is a major cause, if not the main cause of bias in retrospective studies. Prospective studies offer an investigator the opportunity to collect all data relevant to the research question and to control the selection of study participants to minimize selection bias. It is up to the investigator to take full advantage of this opportunity through adequate planning of the study. Retrospective studies are not without merit, however, because they may be the only practical way to study rare diseases or patient outcomes. Whatever the study design, problems in defining the study population should either be addressed in the study or stated as a limitation of the study. In this article, I have discussed the importance of considering the research question and selection of study participants and have provided some basic related advice (Table). These two aspects of clinical research have a substantial effect on whether a study's results are scientifically worthwhile and ultimately useful to others. Attention to these issues should be helpful to the various roles involved in the life of a study: the investigator planning the study, the reviewer of the resulting manuscript, and the reader of the published research article.
ASJC Scopus subject areas
- Radiology Nuclear Medicine and imaging