The utility of spirometry in diagnosing pulmonary restriction.

The aim of this retrospective study was to determine the utility of the spirometric measurements FVC, FEV1, and FEV1/FVC in diagnosing pulmonary restriction. Spirometry and lung volume measurements performed on the same patient visit were analyzed. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of (1) FVC<lower limit of normal (LLN) (NHANES III reference values) and (2) FVC<LLN and FEV1/FVC>or=LLN were compared to diagnose restriction based on lung volume measurements. In all, 18,282 pulmonary function tests from 8,315 patients were analyzed. Twenty-six percent of the patients (n=2,213) had restriction based on lung volume measurements. The sensitivity, specificity, PPV, and NPV of FVC<LLN to diagnose restriction based on lung volume measurement criteria were 88.6%, 56.8%, 39.9%, and 93.9%, respectively. The sensitivity, specificity, PPV, and NPV of FVC<LLN and FEV1/FVC>or=normal to diagnose restriction based on lung volume criteria were 72.4%, 87.1%, 64.4%, and 90.7%, respectively. Analysis of ROC curves showed that spirometric criteria based on FVC alone performed better (area under the curve=0.817) than those based on the combined criteria of FVC and FEV1/FVC (area under the curve=0.584). Consistent with earlier findings, the negative predictive value for a normal FVC (>or=LLN) to exclude pulmonary restriction was high in this series (up to 95.7%). Also, a spirometric diagnosis of "restriction" (FVC<LLN and FEV1/FVC>or=LLN) had a positive predictive value of 26.3-73.9%. On this basis, normal FVC can be regarded as excluding restriction with high reliability.