RESUMO
PURPOSE: A first objective was to provide an overview of advantages and cautions around so-called retrospective clinical studies. A second objective was to provide guidelines for strong studies that can make a valid contribution to the clinical literature, whether these studies are prospective, retrospective, experimental, or observational. METHOD: Invited commentaries were solicited from statistical and study design experts. RESULTS: The strength of a clinical study does not lie so much with its point in time relative to data generation, as it lies with study design. In fact, quite surprisingly, data collected in the past can be modeled to create a prospective study, if appropriate. One distinctive strength of observational studies-which are sometimes but not always retrospective-is the ability to obtain a large corpus of data from medical databases rapidly, as sometimes warranted by pressing health care policy and practice issues. CONCLUSION: Retrospective studies, often considered inferior to prospective, randomized, and controlled clinical trials, can have strength and validity often not recognized in the hierarchy of clinical data.
