A lung function threshold for survival? - FEV1Q and mortality in patients with COPD and chronic respiratory failure.

INTRODUCTION

The FEV1 quotient (FEV1Q), calculated as the index between FEV1 and a theoretical lower survivable FEV1 threshold of 0.4L for females and 0.5L for males, has been investigated as a novel method of interpreting results from lung function testing. The applicability of the FEV1Q in populations with chronic respiratory failure has not been studied, and the continuous association between the FEV1Q and mortality is unknown.

METHODS

Longitudinal analysis of data from the DISCOVERY database. First percentile values of FEV1 were determined. The predictive ability of FEV1Q and FEV1 %-predicted values for overall and respiratory mortality were compared using Cox and Fine-Gray regression models with C-statistics. The continuous association between FEV1Q and mortality was evaluated using a restricted cubic spline.

RESULTS

A total of 5,711 patients (61 % females) with oxygen-dependent COPD were studied. First-percentile values of FEV1 were 0.3L for females, and 0.4L for males. Higher levels of FEV1Q were associated with a lower risk of overall and respiratory mortality when adjusting for age, sex, height, smoking status, A-a-gradient, and education. For overall mortality, FEV1Q and FEV1 %-predicted models had identical C-statistics of 0.60 (95 %CI 0.59-0.61). The association between FEV1Q and overall mortality was J-shaped, with a threshold of increased risk at FEV1Q values < 1.0.

CONCLUSION

While first-percentile values of FEV1 were lower in this cohort than in previous studies, a population threshold for increased mortality risk was identified at FEV1Q levels corresponding to those originally presented. For individual subjects, neither FEV1Q, nor FEV1 %-predicted, were identified as useful predictors of mortality.