Comparisons of Spirometric Reference Values and the Proportions of Abnormal Subjects among Male Smokers and Those Symptomatic in a Community Population

A comparison was performed of various population reference equations for several spirometric measurements. The objective was to determine if the different equations yielded similar distributions of the percent predicted values, similar reference values for asymptomatic nonsmokers (i.e., normal subjects), criteria (or cutoff points) for who might be defined statistically as normal, and similar proportions of smokers and symptomatic subjects who would fall below the criteria values and thus be considered statistically abnormal. The cross-sectional study population of adults in Tucson, subjects of an ongoing study, was used. Most of the equations yielded reasonably similar distributions of percent predicted values. Older and/or more unusual equations did not yield similar results. Using the lower 95th percentile criteria yielded similar cutoff points as reference values for statistical normality, and similar proportions of statistically abnormal among various smoking and symptomatic groups; some of the equations used did not. When the criteria used were the lower 1.645 SEE, or the published criteria, the proportions considered statistically abnormal were very low or very high; again, some equations gave very disparate results. The first conclusion was that the use of the lower 95th percentile criteria was more sensitive and specific. It was concluded also that most equations yield sufficiently similar results such that the choice of reference equation from those available would depend on other criteria.