The overall scheme for evaluating a new diagnostic test is presented in Fig. 2. The office-based practitioner has encountered a new diagnostic test (A). The purpose of the new test seems to be one that would be useful in an office practice, for the suspected disease is seen frequently enough to justify stocking an office test (B), and the benefits of the test, if true, would augment the delivery of patient care (C). From the manufacturer, appropriate literature references are obtained which compare the new test with either the current office test or an accepted standard; the article is from a peer review journal (D). This article is then examined more carefully (E). The conclusions contained in the Abstract, if true, would lead the practitioner to change his current office procedure. The Methods section of the article is now examined for evidence that an independent clinical standard of comparison was used; a patient population was described similar to that in the office practice; a wide spectrum of disease presentations was evaluated; a sensible definition of "normal" was utilized; the study was "blinded" to avoid bias; and a detailed description of the test techniques was provided. In the Results section, one notes the precision and inter-observer variability of the test, and an appropriate statistical analysis is verified. Finally, the Discussion is skimmed to find out whether the authors have commented on the utility of the test in an office setting.(ABSTRACT TRUNCATED AT 250 WORDS)