Study Supports Simple Ratio for COPD Diagnosis

At a Glance

Uncertainty regarding how to diagnose chronic obstructive pulmonary disease (COPD) has posed significant problems for early detection and treatment of this common disease. 

Now, a new study of 24,207 adults has identified the best fixed threshold for predicting COPD-related hospitalizations and deaths. The results support current guidelines that, until this time, were based on the opinion of some experts and have remained controversial.

“Simplifying and standardizing the diagnosis of COPD has the potential to improve diagnosis, clinical care, and clinical research for this common and underdiagnosed chronic lung disease,” says senior author Elizabeth Oelsner, MD, MPH, Herbert Irving Assistant Professor of Medicine at Columbia University Vagelos College of Physicians and Surgeons.

The study was published June 25 in the Journal of the American Medical Association.



COPD is the third-leading cause of death worldwide and the fourth-leading cause of death in the United States. An estimated 24 million Americans have COPD, but only half have received a diagnosis.

The diagnosis of COPD requires a simple breathing test called spirometry. Spirometry measures the amount of air you can force out of your lungs in one second (forced expiratory volume in one second, FEV1) compared to the total amount of air you can force out of your lungs (forced vital capacity, FVC). A lower FEV1/FVC ratio indicates airflow obstruction, but where to “draw the line” between normal and abnormal levels to define and diagnose COPD is debated. 

Current recommendations define COPD by a FEV1/FVC ratio less than 0.70. However, this cutoff or “threshold” is based on the opinion of some experts rather than scientific evidence, and other experts have argued for a different threshold. This has led to controversy, inconsistent scientific reports, and confusion in clinical practice. The prevalence of COPD could vary by more than a third based upon the threshold used.


What the Researchers Did

Researchers, led by Oelsner, aimed to provide scientific evidence for the best threshold to diagnose COPD. To do this, they compared how well various FEV1/FVC ratio thresholds predict hospitalizations and deaths from COPD. They used data from the NHLBI Pooled Cohorts Study. This resource, developed at Columbia University, combined numerous existing multi-ethnic cohort studies to support large-scale studies of COPD. 


What the Study Found

In this large study of 24,207 adults, the best FEV1/FVC ratio threshold for predicting COPD hospitalizations and deaths was 0.71, which was similar to 0.70, the current clinical guideline. The fixed threshold of 0.70 proved as accurate, or more accurate, than other thresholds. 


Why It Matters

Validation of a single fixed ratio for diagnosis of COPD simplifies diagnosis, clinical care, and future clinical trials for this underrecognized and important chronic lung disease.


The study, titled “Discriminative Accuracy of FEV1/FVC Thresholds for COPD-Related Hospitalization and Mortality,” was published June 25 in the Journal of the American Medical Association.

Additional authors: Surya P. Bhatt (University of Alabama at Birmingham); Pallavi P. Balte (Columbia University Irving Medical Center); Joseph E. Schwartz (CUIMC); Patricia A. Cassano, (Weill Cornell Medical College, New York); David Couper (University of North Carolina, Chapel Hill); David R. Jacobs Jr (University of Minnesota); Ravi Kalhan (Northwestern University, Illinois); George T. O’Connor (Boston University); Sachin Yende (University of Pittsburgh and Veterans Affairs Pittsburgh Healthcare System); Jason L. Sanders (Brigham and Women’s Hospital, Boston); Jason G. Umans (MedStar Health Research Institute, Maryland); Mark T. Dransfield (University of Alabama at Birmingham); Paulo H. Chaves (Florida International University, Miami); and Wendy B. White (Tougaloo College, Mississippi).

Elizabeth Oelsner is supported by NIH grants R21HL129924 and K23HL130627. (See paper for full list). 

Columbia authors report no conflicts of interest related to this research. (See paper for other authors).