A conceptual framework for pulmonary function testing

Abstract

Pulmonary function (PF) testing is traditionally depicted as a sequence of separate activities in which a “test” is performed producing “results” which are subsequently “interpreted.” An alternative view of PF testing is presented which identifies two problems that must be solved: an inductive problem and a deductive problem. Solving the inductive problem yields two sets of models both of which are needed for the solution of the deductive problem. One set of models, called physiology models, describes physical processes in the respiratory system. The second set, termed pathology models, relates parameters of the physiology models and clinical measures to functional status category, treatment and prognosis. A key component of the deductive problem is the PF test, itself, which is the test of the hypothesis that a subject belongs to a particular functional category specified by a given pathology model. Since this hypothesis is formulated before pulmonary function (physiology model) parameters are evaluated, knowledge of the implications (interpretation) of any observed parameter values (results) is an integral part of the “test” and therefore must be understood before “results” are obtained. Even though this framework is described in the context of PF testing, it can be applied to medical decision making in general.