ERS/ATS Lung Guidelines Endorse Global Lung Initiative for Interpretation of Spirometry, DLCO, and Lung Volumes

Aaron B. Holley, MD


November 28, 2022

The European Respiratory Society (ERS) and American Thoracic Society (ATS) just updated their standards for interpreting lung function testing. The last joint society update from these organizations was in 2005, and much has changed since then. The authors cover multiple topics of varying complexity, each worthy of their own review. Pardon my vanity as I attempt to provide a cursory overview of the reference equation recommendations in under 1500 words.

The 2005 interpretation standards recommend the National Health and Nutrition Examination Survey (NHANES III) equations for referencing spirometry results to normal for White persons, Hispanic persons, and Black persons. They identify commonly applied reference equations for interpreting diffusion capacity for carbon monoxide (DLCO) and lung volumes but stop short of endorsing any one equation over another. In the 2022 version it's all about the Global Lung Initiative (GLI) for spirometry, DLCO, and lung volumes.

For spirometry, the chief differences from NHANES III is that GLI has a much larger sample size, separates the population into heterogeneous ethnic groups, extends all the way to age 95 (NHANES III stops at 80), and collapses White persons and Hispanic persons into one category. NHANES III was drawn directly from the US population and uses only three ethnic/race categories: Caucasian, African American, and Hispanic. Our group found that differences between the two equations in a US population are pretty minimal. The difference in age range leads to proportionately larger discrepancies between equations when applied to individuals older than 65 years, with Black persons particularly affected.

The GLI equations for DLCO and transfer factor for CO (TLCO) have much larger sample sizes and generally ensured a more rigorous patient selection compared with those derived in the past. They include only studies that used modern equipment, test performance techniques, and statistics. Both have been compared with other reference sets and both include purely White populations. Finally, neither use weight or body mass index as a co-variate, and each include a large proportion of overweight and obese individuals.

The 2022 ERS/ATS guidelines go on to tackle the thorny issue of whether and how to factor in race/ethnicity when interpreting lung testing. Here they come out unequivocally against applying a fixed correction factor when applying results from White persons to non-White persons. To justify this they cite another study my group published in 2010 that found fixed correction factors were inferior to the statistically derived lower limit of normal derived directly from a non-White population. They also cite the NHANES III paper on reference equations for the same reason. They do allow for using reference equations where race is a covariate when the individual being studied is of a race or ethnicity that was included in the population used to derive the equation.

When assessing the impact and value of the ERS/ATS Lung Function Interpretation Standards, I believe a few caveats are in order. First, there's been a move toward simplification of spirometry reports. This makes a lot of sense. To the extent that simplification is a goal, recommending the same reference sets for everyone is important for patient care at the individual and population levels and for research.

Beyond recommending NHANES III for spirometry in North American patients in the 2005 guidance, no other reference sets were specified for any other population or test. Laboratories were free to use whatever set they believed best represented the population they cared for. For the physicians interpreting lung testing, there was often no way to know which equation was used. Even if they were lucky enough to receive a report that listed the equation that had been applied, there was a better than average chance they had never used or seen it before, meaning they were ignorant of its limitations. GLI levels the playing field for everyone and removes some uncertainty, assuming all labs adopt the 2022 ERS/ATS recommendations.

This will come at a cost. For example, at labs in the United States, there's reason to believe that NHANES III would be preferable to GLI when interpreting spirometry in patients under age 65 years. Differences between sets should be minimal, but the derivation population for NHANES III is far less heterogeneous than that used for GLI and is drawn directly from the United States. Furthermore, GLI complicates the assessment of racial/ethnic origin by using complex groups and separating Chinese individuals by whether they were born above or below the Huaihe River and Qinling Mountains. Determining which ethnic "label" the respiratory therapist should assign won't always be easy.

Using race/ethnicity at all has also been questioned. I'm unable to do justice here to the pros and cons of including race/ethnicity in the interpretation of spirometry, but other experts have written thoughtful papers on this topic. In my opinion, it is somewhat contradictory for the ERS/ATS to come out against a fixed correction factor while allowing for race/ethnicity as a covariate. The papers they cite to justify its abandonment found that a statistically derived equation for the given race/ethnicity being tested is always preferable; however, they both concluded that fixed correction factors actually perform quite well. Unequivocally coming out against correction factors but also recommending equations for DLCO and lung volumes derived from purely White populations leaves the interpreting clinician in a bind. For reasons we don't entirely understand, self-reported race explains a portion of the measured variability in DLCO and lung volumes just as it does for spirometry. Irrespective of any individual's opinion on whether race/ethnicity should be used, practitioners should know that the percent predicted values for their patient's spirometry will factor it in whereas those for DLCO and lung volumes will not. This could lead to a muddled interpretation.

In summary, the ERS/ATS interpretation standards are incredibly well done and very thoughtful. They move us toward standardization using the very precise GLI equations. It's important for readers and clinicians to realize that although the unresolved discussion of race/ethnicity is important, with or without its incorporation GLI explains a large portion of the variability in lung function testing. To optimize precision, clinicians are best off understanding when race/ethnicity is and is not being factored in and using their clinical judgement to optimize the interpretation. This is exactly what the ATS and ERS have always recommended we do.

Aaron B. Holley, MD, is a professor of medicine at Uniformed Services University in Bethesda, Maryland, and a pulmonary medicine/critical care physician at MedStar Washington Hospital Center in Washington, DC. He covers a wide range of topics in pulmonary, critical care, and sleep medicine.

Follow Medscape on Facebook, Twitter, Instagram, and YouTube


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.