Race-Neutral Pulmonary Function Testing – Pulmonology Advisor

In July, physician leaders at Boston Medical Center offered a roadmap for transitioning from race-specific to race-neutral reference equations for pulmonary function testing (PFT) in an article published in Chest.¹  The transition to race-neutral PFT reference equations by the urban safety-net hospital, which performs 3500 to 4000 PFTs annually, represents another milestone in medicine’s long journey toward health equity — and away from racist notions of the past.

“PFT results impact decision making across the healthcare system,” wrote Amos Wu, MD, and his colleagues from Boston Medical Center. “PFTs are widely used to assess lung function and to identify pathologic deviations by comparing patient’s results to reference equations derived from a healthy cohort.”¹  

Until recently, the reference equations recommended for interpreting PFT results assumed that a healthy cohort of Black individuals would have lower lung function scores than a healthy cohort of White individuals. After mounting debate and emerging evidence challenging this assumption, the American Thoracic Society (ATS) issued a recommendation in 2023 for the use of race-neutral average reference equations for PFT interpretation. As the 2023 Official ATS Statement: Race and Ethnicity in Pulmonary Function Test Interpretation pointed out, “the use of race and ethnicity in reference equations for PFT interpretation codifies race and contributes to health disparities by norming differences in lung function.”² 

Physician leaders at Boston Medical Center saw transitioning to race-neutral PFT equations as a move toward health equity for hospital patients, 75% of whom identified as being persons from underserved populations. The 514-bed, 70-specialty hospital began planning the transition to race-neutral spirometry reference equations shortly after the ATS recommendation was released, and that plan officially became a reality on May 20, 2024.¹

“No concerns have been reported to date,” noted Wu and colleagues.  However, they added, “the full impact of race-neutral equations on clinical decision-making remains uncertain.”¹

Race and Spirometry

According to the 2023 ATS statement on PFT, the origins of race-based approaches to spirometry are embedded in deep prejudice and overt racism that have been present since before the United States began. Early investigators found Black Americans had worse lung function scores than White Americans, which were seen as confirmation of race-based biological differences that were used to justify slavery.² 

The belief Black Americans had inherent biological differences persisted even after emancipation. In 1896, Race Traits and Tendencies of the American Negro, a book by Frederick Hoffman, eugenics promoter and biostatistician of the Prudential Life Insurance Company, argued Black Americans’ lesser lung capacity provided proof of their physical inferiority. Black American intellectuals such as W.E.B. DuBois decried this scientific racism, asserting differences in lung function reflected the social conditions experienced by Black Americans.²

The idea that differences in lung function might be attributable to something other than race evolved gradually, as the pulmonary community grew to appreciate disparities between socioenvironmental factors affecting Black and White Americans, such as tobacco smoke and pollution exposure, nutrition, pulmonary infections, and working conditions. Yet this new understanding did not prompt changes in how PFTs were interpreted for Black Americans. In fact, some saw race-based adjustments as a way to avoid discrimination; for example, race-based PFT interpretation allowed Black American workers to qualify for jobs requiring a minimum level of pulmonary function, such as in the cotton industry.²

As PFTs became routine in clinical practice, the pulmonary community worked to standardize normal PFT reference values. These numbers, based on data from healthy populations, allowed PFT results to be compared across different PFT labs, replacing the previous practice of using distinct reference ranges from lab to lab.

In a 1999 paper, John Hankinson and colleagues used data from the third National Health and Nutrition Examination Survey (NHANES III), which represented a large sample of the American population, as a basis for PFT reference values. When adjusted for age and sex, Black Americans had lower mean PFT scores than White Americans, and to a lesser extent, than Mexican Americans. Although Dr Hankinson and his colleagues had considered a race-neutral approach, their use of distinct race-specific equations enabled more precise interpretation.³

“They used the best methodologies available at the time, but the approach valued specificity over sensitivity. To get the highest specificity, to know that a person with disease truly has disease, you use a race-specific equation, because you set the bar higher for what is abnormal,” explained Aaron D Baugh, MD, a pulmonary and critical care physician and an assistant professor of medicine at the University of California, San Francisco. Dr Baugh was also lead author of the 2022 article, Reconsidering the utility of race-specific lung function prediction equations,⁴ and the 2023 article, Towards a race-neutral system of pulmonary function test results interpretation.⁵

Later PFT equations, such as those used by the Global Lung Function Initiative (GLI), also employed a race-based approach, using race-specific equations for predicted PFT values in individuals classified as White, Black, North Asian, and South Asian, based on analysis of the larger GLI dataset. A composite GLI-OTHER equation was used for individuals of mixed race, but it was heavily weighted toward values for White individuals due to the sample sizes.

The Move Toward Race-Neutral PFTs

For the past few years, said Dr Baugh, a growing number of voices have decried the drawbacks of the race-specific approach to PFT interpretation, sparking vigorous debate. Some expressed concern that race-specific equations might inadvertently normalize reduced lung function within a racial group, rather than highlight underlying inequities and discrimination. There were also concerns that a race-specific approach might perpetuate a purely biological conception of race, when current thinking considers race to be a social construct and an unreliable proxy for genetic differences.²

In 2020, an article published in the New England Journal of Medicine highlighted and criticized the use of race-based approaches in many medical fields, including nephrology, cardiology, urology, obstetrics, and pulmonology, noting their propensity to perpetuate or amplify race-based health inequities,⁶ said Meredith C McCormack, MD, MHS, director of the division of pulmonary and critical care medicine at Johns Hopkins Medicine in Baltimore, Maryland.

As the debate continued, concerns arose about the practical challenges switching from race-based to race-neutral PFTs, given their use in other contexts such as federal disability and life insurance evaluations. It was also argued that a race-neutral approach would reduce the specificity of PFT values and inhibit progress toward equitable health care for Black patients.⁷ 

It had long been the accepted belief that race-based calculations did not result in clinically meaningful differences, said Dr Baugh — a belief that was founded on prior evidence from mostly healthy people. This perspective began to change after the publication of 5 key papers suggesting that race-neutral equations might lead to meaningful clinical differences in self-reported symptoms, emphysema, hospitalizations rates, mortality, and more.^4,8,9,10,11 “Looking at the new evidence, we can definitely say that there actually is a clinical impact,” said Dr Baugh.

In part due to this new data, the ATS issued its 2023 statement on race-neutral PFTs. The ATS now recommends use of the GLI-GLOBAL race-neutral reference equation. Unlike GLI-OTHER, GLI-GLOBAL includes a weighted average of the data in the original GLI-ethnicity specific equations.¹²

Clinical Implications

A trade-off exists between sensitivity and specificity with respect to these equations. Although increasing sensitivity may allow certain diseases to be flagged earlier, it comes with a greater likelihood of false-positive results. In contrast, a more specific equation ensures more patients are accurately diagnosed but it may miss some borderline patients who actually have lung disease.

The use of GLI-race neutral equations enhances the sensitivity of findings for Black individuals, producing a higher incidence of abnormal findings. The same is true for Asian Americans, albeit to a lesser extent, whereas the sensitivity is lower for White and Hispanic Americans. However, said Dr Baugh, when using GLI-race neutral equations, the sensitivity of findings for White and Hispanic Americans is reduced much less than it is increased for Black Americans.

According to Dr Baugh, the increased PFT sensitivity for Black individuals may be helpful given the barriers to high-quality care faced by this population. It might also help practitioners catch subtle signals of lung disease early in the disease course. “I would rather have to explain to a patient about a falsely abnormal result than have a patient who really had disease that I didn’t catch,” he said.

Importantly, he stressed, most people’s PFT findings will not substantively alter with the use of race-neutral equations. People who are clearly healthy would be found healthy using either version of the test, and severe disease will also be reflected on either test. “The largest swing is for African Americans, and it’s only about an 8% to 9% difference in the people in whom the test might flag,” said Dr Baugh.

Both Dr Baugh and Dr McCormack emphasized the importance of patient information to put PFT values in context — for example, getting more detailed medical history, exercise tests, imaging tests, etc, especially for borderline cases. “When someone’s falling on either side of a threshold that might demarcate abnormal vs normal, we should really dig in and pay more attention to additional ways of assessing,” said Dr McCormack.

It is also critical to look at how an individual’s PFT results change over time, said Dr Baugh, as trends are more important than single point values. He pointed out that the whole question of the value of race-neutral vs race-specific PFTs might be thought of as debating what a “normal” PFT value should be. “One thing you know is that the person is normal to themselves, so changes from the patient’s baseline are always most important,” he added.

Using trends and supplementary information can help prevent both underdiagnosis and overdiagnosis, which can have clinical and nonclinical implications. For example, under the race-based equations, a Black individual trying to qualify for a job as a firefighter might be more likely to be flagged on screening PFT tests, but additional tests and follow up screening might make it possible to move forward with employment. Such an approach is also needed in other contexts — such as evaluating whether a person is safe to undergo thoracic surgery. If PFT results are borderline, cardiopulmonary testing could provide reassurance that a patient is a suitable candidate for the operation.

Dr McCormack said it may be helpful for clinicians to look at trends in terms of PFT values that are expressed in absolute values, such as FEV₁, which would not change with the move to race-based PFT tests. When interpreting FEV₁ as a percent of predicted or when using a z score, the patient is compared against the norm (ie, healthy individuals), so these values change with the use of race-neutral vs race-specific approaches, she explained.

Not all PFT labs will be able to move to race-neutral approaches right away; some computing software used to calculate PFT values will need to be updated first. Dr McCormack noted that some software that does not yet have the GLI-GLOBAL calculations installed may have the GLI-OTHER, which can serve as a rough proxy.

The Boston Medical Center Roadmap

Boston Medical Center’s transition to race-neutral equations for interpreting PFTs used an 8-step quality improvement framework known as the Kotter Change Model. “The process begins by creating a sense of urgency to motivate action, followed by building a coalition of key shareholders to lead the change. The change is then initiated through communication, empowerment, and identification of short-term wins. Finally, the model focuses on implementation and sustained change,” Dr Wu and colleagues explained.¹

Key steps in the process were:

• establishing multidisciplinary leadership for the change process;
• forming a strategic vision and communication plan;
• having direct communication and close partnership with their PFT software manufacturer regarding the transition process; and,
• educating the hospital’s respiratory therapists and technologists who perform PFTs.

On a practical level, developing a strategy to minimize disruption to test interpretation and reporting is key, the authors noted. “Logistical areas of consideration are the current state and need for updates (including reference equations, software, and hardware), adjustments to operations, and PFT report updates.” They further noted that their PFT software manufacturer’s expertise in system transitions and race-neutral reference equations “was crucial to our success.”¹

Importantly, the race-neutral approach to PFTs has been generally well-received by the nearly 1000 physicians working at Boston Medical Center, Dr Wu and colleagues reported. “While there was a minority of expressed concern for the potential of negative unintended consequences, no clinicians voiced a desire to maintain race-specific reference equations.”¹