Tutorials in clinical research: part III. Selecting a research approach to best answer a clinical question.

OBJECTIVE This is the third in a series of sequential "Tutorials in Clinical Research." The objectives of this specific report are to enable the reader to rapidly dissect a clinical question or article to efficiently determine what critical mass of information is required to answer the question and what study design is likely to produce the answer. STUDY DESIGN Tutorial. METHODS The authors met weekly for 3 months exploring clinical problems and systematically recording the logic and procedural pathways from multiple clinical questions to the selection of proper research approaches. The basic elements required to understand the processes of selection were catalogued and field tested, and a report was produced to define and explain these elements. RESULTS Fundamental to a research approach is the assembly of subjects and the allocation of exposures. An algorithm leading to the selection of an approach is presented. The report is organized into three parts. The tables serve as a rapid reference section. The initial two-part narrative explains the process of approach selection. The examples section illustrates the application of the selection algorithm. CONCLUSIONS Selecting the proper research approach has six steps: the question, logic and ethics,identification of variables, data display considerations, original data source considerations, and selection of prototypical approaches for assembly of subjects. Field tests of this approach consistently demonstrated its utility.