Z-scores for Spirometry Interpretation: Implications for Classifying Impairments in Lung Function

RATIONALE: Recent guidelines for spirometry interpretation recommend the use of z-scores rather than percent-predicted values to classify the severity of lung function impairment. The potential impact of this change remains unknown. OBJECTIVE: To evaluate the proportion of individuals whose severity classification would change with a transition to z-scores, and to assess how changes in classification are associated with respiratory symptoms and outcomes. METHODS: We evaluated two cohorts of individuals with spirometry: (1) individuals from NHANES III and (2) individuals with COPD from the Johns Hopkins Health System. Using the former and current ATS thresholds, all individuals were classified as normal, mild, moderate, or severe according to both their FEV1 percent-predicted and z-score. Individuals were classified as less severe, more severe, or unchanged with a shift from percent-predicted to z-score thresholds. We evaluated the association between a change in classification with dyspnea, cough, wheeze, phlegm, and mortality in the NHANES cohort, and with COPD exacerbations, all-cause hospitalizations, and cachexia in the COPD cohort using multivariable logistic or Cox regression analyses. RESULTS: A total of 14,863 individuals were included in the NHANES cohort and 14,238 in the COPD cohort. In the NHANES and COPD cohorts, respectively, the mean age was 44 and 64 years, and the mean FEV1 was 3.01 L and 1.75 L. Among the NHANES cohort, 1,497 (10%) individuals were classified as a lesser severity with z-scores, and 6,970 (49%) were classified as a lesser severity among the COPD cohort. No individuals were reclassified to a higher severity under z-scores. Compared to individuals with no change in their severity classification, those reclassified to lesser severity with z-scores had a corresponding lower risk of dyspnea (OR 0.76; 95% CI 0.62-0.92), mortality (hazard ratio [HR] 0.82, 95% CI 0.71-0.93), COPD exacerbation (OR 0.45, 95% CI 0.35-0.58), and all other outcomes except cough. CONCLUSION: A shift from percent-predicted values to z-scores for spirometry interpretation results in reclassification of severity for a substantial proportion of individuals and particularly among those with COPD. Individuals reclassified to lesser severity have a corresponding lower risk of clinically relevant outcomes, supporting the use of z-score threshold for spirometry interpretation.