Evaluating diagnostic tests.

Abstract

The presentations at this conference will discuss new technologies and rapid scientific developments that have resulted in new diagnostic tests for periodontal disease, musculoskeletal imaging, temporomandibular joint dysfunctions, and incipient coronal and root dental caries. However, for many of these claims, there has been insufficient scientific support regarding the sensitivity and specificity of the tests, or their ability to predict the percent of cases in which the disease or condition progresses to the next state of development. Research is needed that will yield the basic diagnostic parameters of new diagnostic tests, i.e., their accuracy, precision, sensitivity, specificity, positive predictive value, and negative predictive value. The purpose and methods for calculating each of these measures are described in this paper. Five questions are then presented that will need to be addressed in future research regarding new diagnostic tests: (1) Does the scientific theory of the test fit with our current body of knowledge? (2) Have the efficacy parameters of the test been reliably determined? (3) How does the test affect clinical decision-making? (4) Does using the test improve the patient's health? and (5) Is the added expense of the test justified by increased effectiveness or by avoiding other health expenditures?