[Skip to Content]
[Skip to Content Landing]
Original Investigation
June 25, 2019

Discriminative Accuracy of FEV1:FVC Thresholds for COPD-Related Hospitalization and Mortality

Author Affiliations
  • 1Division of Pulmonary, Allergy, and Critical Care Medicine and the UAB Lung Health Center, University of Alabama at Birmingham
  • 2Division of General Medicine, Columbia University Medical Center, New York, New York
  • 3Division of Nutritional Sciences, Weill Cornell Medical College, Ithaca, New York
  • 4Gillings School of Global Public Health, University of North Carolina, Chapel Hill
  • 5Division of Epidemiology and Community Health, School of Public Health, University of Minnesota, Minneapolis
  • 6Division of Pulmonary and Critical Care Medicine, Northwestern University, Chicago, Illinois
  • 7Division of Pulmonary, Allergy, Sleep, and Critical Care, Boston University, Boston, Massachusetts
  • 8Department of Critical Care Medicine, University of Pittsburgh and Veterans Affairs Pittsburgh Healthcare System, Pittsburgh, Pennsylvania
  • 9Division of Pulmonary and Critical Care Medicine, Brigham and Women’s Hospital, Boston, Massachusetts
  • 10MedStar Health Research Institute, Hyattsville, Maryland
  • 11Benjamin Leon Center for Geriatric Research and Education, Florida International University, Miami
  • 12Undergraduate Training and Education Center, Tougaloo College, Tougaloo, Mississippi
  • 13Department of Epidemiology, Mailman School of Public Health, Columbia University Medical Center, New York, New York.
JAMA. 2019;321(24):2438-2447. doi:10.1001/jama.2019.7233
Key Points

Question  What is the discriminative accuracy of various thresholds for the ratio of the forced expiratory volume in the first second to the forced vital capacity (FEV1:FVC) for predicting chronic obstructive pulmonary disease (COPD)-related hospitalization and mortality?

Findings  Among 24 207 participants from 4 US general population–based cohorts, the optimal fixed threshold for discriminating COPD-related events was 0.71 (C statistic for the optimal fixed threshold, 0.696). The discriminative accuracy of the 0.71 threshold was not significantly different than that of the 0.70 threshold (difference, 0.001) but it was more accurate than a lower-limit-of-normal threshold derived from population-based reference equations (difference between the optimal ratio threshold vs the model using the LLN threshold, 0.034). The 0.70 threshold provided optimal discrimination in a subgroup analysis of ever smokers and in adjusted models.

Meaning  These results support the use of FEV1:FVC less than 0.70 to identify individuals at risk of clinically significant COPD.

Abstract

Importance  According to numerous current guidelines, the diagnosis of chronic obstructive pulmonary disease (COPD) requires a ratio of the forced expiratory volume in the first second to the forced vital capacity (FEV1:FVC) of less than 0.70, yet this fixed threshold is based on expert opinion and remains controversial.

Objective  To determine the discriminative accuracy of various FEV1:FVC fixed thresholds for predicting COPD-related hospitalization and mortality.

Design, Setting, and Participants  The National Heart, Lung, and Blood Institute (NHLBI) Pooled Cohorts Study harmonized and pooled data from 4 US general population–based cohorts (Atherosclerosis Risk in Communities Study; Cardiovascular Health Study; Health, Aging, and Body Composition Study; and Multi-Ethnic Study of Atherosclerosis). Participants aged 45 to 102 years were enrolled from 1987 to 2000 and received follow-up longitudinally through 2016.

Exposures  Presence of airflow obstruction, which was defined by a baseline FEV1:FVC less than a range of fixed thresholds (0.75 to 0.65) or less than the lower limit of normal as defined by Global Lung Initiative reference equations (LLN).

Main Outcomes and Measures  The primary outcome was a composite of COPD hospitalization and COPD-related mortality, defined by adjudication or administrative criteria. The optimal fixed FEV1:FVC threshold was defined by the best discrimination for these COPD-related events as indexed using the Harrell C statistic from unadjusted Cox proportional hazards models. Differences in C statistics were compared with respect to less than 0.70 and less than LLN thresholds using a nonparametric approach.

Results  Among 24 207 adults in the pooled cohort (mean [SD] age at enrollment, 63 [10.5] years; 12 990 [54%] women; 16 794 [69%] non-Hispanic white; 15 181 [63%] ever smokers), complete follow-up was available for 11 077 (77%) at 15 years. During a median follow-up of 15 years, 3925 participants experienced COPD-related events over 340 757 person-years of follow-up (incidence density rate, 11.5 per 1000 person-years), including 3563 COPD-related hospitalizations and 447 COPD-related deaths. With respect to discrimination of COPD-related events, the optimal fixed threshold (0.71; C statistic for optimal fixed threshold, 0.696) was not significantly different from the 0.70 threshold (difference, 0.001 [95% CI, −0.002 to 0.004]) but was more accurate than the LLN threshold (difference, 0.034 [95% CI, 0.028 to 0.041]). The 0.70 threshold provided optimal discrimination in the subgroup analysis of ever smokers and in adjusted models.

Conclusions and Relevance  Defining airflow obstruction as FEV1:FVC less than 0.70 provided discrimination of COPD-related hospitalization and mortality that was not significantly different or was more accurate than other fixed thresholds and the LLN. These results support the use of FEV1:FVC less than 0.70 to identify individuals at risk of clinically significant COPD.

×