1520Prospective Vs Retrospective Cohort Studies: Is a Consensus Needed?

Abstract

Focus of Presentation There has been a lot of confusion among epidemiologists, especially students and early career epidemiologists, on definitions of prospective and retrospective cohort studies versus the actual practice. Many retrospective studies are reported and reviewed as prospective studies, and this is expected to lead to drawing wrong conclusions from studies, as well as leading to the disruption of the evidence-based hierarchy especially when it comes to systematic reviews and meta-analysis. There is a need, therefore, for consensus by the expert community to clarify every aspect of this issue. Findings Most epidemiological textbooks define prospective cohort study as a type of cohort study where researchers develop and conduct the study on all subjects before they start developing outcomes while in retrospective cohort study, the researchers go back in time to study exposure history for subjects whom at least some of them have developed the outcomes. Some experts, though, define both study types based on the time of development of the hypothesis or research question in relation to data collection. That is, those who develop the research question before data collection are considered to be conducting a prospective study and it is a retrospective study when developed after data collection (Miettinen, O.S. 1985). In some cases, retrospective studies are defined as the cohort studies whose data collections are from the records or historical data, while prospective studies are when data are collected prospectively (Setia M.S. 2016). Meanwhile, some experts consider any follow up study to be prospective even if historical data is used while they define retrospective studies as case control studies (Last, J.M. 2000, Doll, R. and Hill, A.B. 1954) These disparities in definitions lead to confusion among young epidemiologists as to which study type, they are conducting. This leads to the question of how important it is to differentiate the two study types. Is it really necessary? Does it affect my study results? Some experts are of the opinion that the need to differentiate between the two is minimal, as a retrospective cohort study conducted very well could be superior to an average prospective one in terms of evidence-based hierarchy. And due to the ambiguity and inconsistent nature of these labels, some even recommend abandoning the labels all together (Wylie C. E. 2015, Vandenbroucke, J.P. 1991). However, we are aware that data collected in retrospective studies might be prone to issues like incomplete records, recall bias and confounding, thereby leading to decreased internal validity of studies. This could be averted if the same study were conducted retrospectively. Conclusions/Implications It has become apparent that it is very essential for the expert community to address this methodological issue in Epidemiology and try to reach some consensus on this matter with regards to whether these subdivisions should be maintained and if so, how they should be uniformly defined and also to have a standard uniform methodological guideline that could be used internationally. This could serve as a guide for young epidemiologists and could ensure the uniformity of research worldwide. Key messages There is a confusion, especially among young epidemiologist on the definitions and use of prospective and retrospective cohort studies Expert community, especially the International Association of Epidemiology, should address these concerns and if possible, have a consensus to form a standard guideline.

Evaluating agreement between bodies of evidence from randomized controlled trials and cohort studies in medical research: a meta-epidemiological study

“…We did not explore whether disagreement was larger between RCTs compared to prospective and retrospective cohort studies, respectively. The corresponding information was reported in a suboptimal manner, and researchers may use inconsistent nomenclature [ 113 , 114 ]. Second, we did not evaluate the methodological quality of the included systematic reviews, but given that we focused on high-impact journals, we assumed that published systematic reviews are of reasonably high methodological quality.…”