Using spirometry to rule out restriction in patients with concomitant low forced vital capacity and obstructive pattern

Abstract

Background: Different formulas have been proposed to exclude restriction based on spirometry, however none of them have specifically tested the patients whose spirometry show both obstruction and a low forced vital capacity (FVC).

Study objective: The study was designed to create an algorithm that would better predict the absence of restriction in such patients.

Design: Retrospective analysis of prospectively collected data.

Methods: A cohort of consecutive adults that underwent complete pulmonary function testing from 2002-2004 was analyzed. The data was randomly split into two groups to allow for derivation and then validation of a predictive formula. Patients were randomly assigned into either a "derivation" or "validation" group. In the derivation group, stepwise logistic regression was used to determine a formula and optimal cut-off value for the variable with the best discriminative capacity. The formula was applied subsequently to the validation group to test the results and compared to previously published formula.

Results: The study group contained 766 patients. We determined that the variable with the highest association with TLC was [(FEV(1)/FVC) % predicted/FVC % predicted]. A value of ≥1.11 was found to be the maximal cutoff to predict the absence of restriction. The formula was applied to a validation group (n=397) and performed better than prior published algorithm with a sensitivity, specificity, positive predictive value and negative predictive value of 95%, 44%, 22%, and 98%, respectively.

Conclusion: Our formula performs superior to the previously published algorithms in patients with concomitant low FVC and obstruction to exclude restriction.

Keywords: Algorithm; restriction; spirometry; total lung capacity..

Fig. (1)

Area under the curve based on our formula (AUC = 0.81).

Fig. (2)

Area under the curve based on Glady et al. formula (AUC = 0.71).