Abstract
Rationale
Interpretation of spirometry using race-specific reference equations may contribute to health disparities via underestimation of the degree of lung function impairment in Black patients. The use of race-specific equations may differentially affect patients with severe respiratory disease via the use of percentage predicted forced vital capacity (FVCpp) when included in the lung allocation score (LAS), the primary determinant of priority for lung transplantation.
Objectives
To determine the impact of a race-specific versus a race-neutral approach to spirometry interpretation on the LAS among adults listed for lung transplantation in the United States.
Methods
We developed a cohort from the United Network for Organ Sharing database including all White and Black adults listed for lung transplantation between January 7, 2009, and February 18, 2015. The LAS at listing was calculated for each patient under race-specific and race-neutral approaches, using the FVCpp generated from the Global Lung Function Initiative equation corresponding to each patient’s race (race-specific) or from the Global Lung Function Initiative “other” (race-neutral) equation. Differences in LAS between approaches were compared by race, with positive values indicating a higher LAS under the race-neutral approach.
Results
In this cohort of 8,982 patients, 90.3% were White and 9.7% were Black. The mean FVCpp was 4.4% higher versus 3.8% lower among White versus Black patients (P < 0.001) under a race-neutral compared with a race-specific approach. Compared with White patients, Black patients had a higher mean LAS under both a race-specific (41.9 vs. 43.9; P < 0.001) and a race-neutral (41.3 vs. 44.3; P < 0.001) approach. However, the mean difference in LAS under a race-neutral approach was −0.6 versus +0.6 for White versus Black patients (P < 0.001). Differences in LAS under a race-neutral approach were most pronounced for those in group B (pulmonary vascular disease) (−0.71 vs. +0.70; P < 0.001) and group D (restrictive lung disease) (−0.78 vs. +0.68; P < 0.001).
Conclusions
A race-specific approach to spirometry interpretation has potential to adversely affect the care of Black patients with advanced respiratory disease. Compared with a race-neutral approach, a race-specific approach resulted in lower LASs for Black patients and higher LASs for White patients, which may have contributed to racially biased allocation of lung transplantation. The future use of race-specific equations must be carefully considered.
Keywords: race, spirometry, transplant, LAS
Race-specific reference equations are currently recommended for the interpretation of spirometry (1), but their use has been increasingly called into question. Although average lung function from epidemiologic data of healthy populations has been observed to vary by race (2), recent studies have suggested that race-specific equations may inappropriately normalize the lower lung function seen among Black and Asian populations (3–5). This concern has raised questions regarding whether a race-neutral approach to spirometry interpretation, in which the same reference equation would be applied to all patients regardless of their race, is more appropriate (6–9).
By inflating percentage predicted spirometry values for Black and Asian patients, race-specific equations may lead providers to systematically underestimate the degree of lung function impairment for these racial groups. As a result, these equations may contribute to respiratory health disparities. Although evidence that race-specific equations inappropriately normalize lower lung function has resulted primarily from healthy populations, racial disparities associated with race-specific reference equations may be exaggerated among more severely diseased individuals, such as those being evaluated for lung transplantation.
Priority for lung transplantation in the United States has relied primarily on the lung allocation score (LAS) since 2005 (10). The LAS is calculated from multiple clinical variables indicative of disease severity, and it has historically included the percentage predicted forced vital capacity (FVCpp) (11), which is derived from race-specific reference equations. Although implementation of the LAS diminished racial disparities in lung transplantation compared with the pre-LAS era in terms of reduced risk of death on the waiting list and increased chance of undergoing lung transplantation for Black patients (12, 13), more recent data have shown that Black patients on the waiting list are still less likely to be allocated lungs in the post-LAS era compared with White patients (14, 15).
Although these disparities exist even when controlling for the LAS, the use of race-specific reference equations may have further contributed to the disparate outcomes by estimating a higher FVCpp and thus a lower LAS and priority for lung transplantation among Black patients (compared with a lower FVCpp and higher LAS for White patients). Given the emphasized need for the equitable distribution of lung transplants as a scarce resource (10, 16–18), and the need for further evidence of the clinical impact of including race in the interpretation of spirometry, understanding if and how race-specific spirometry interpretation can affect lung transplant allocation is necessary.
Thus, we sought to investigate the effect of race-specific versus race-neutral equations for the interpretation of spirometry on the LAS among Black and White adults listed for lung transplantation in the United States.
Methods
Study Sample
We conducted a retrospective study of all Black and White patients 18 years or older from the United Network for Organ Sharing (UNOS) database who were listed for lung transplantation between January 7, 2009, and February 18, 2015. We included only patients of Black and White race as reported in the UNOS database because Black and White patients have the greatest difference in FVCpp with race-specific compared with race-neutral equations (2). We excluded patients listed after February 18, 2015, as FVCpp was not incorporated for all patients when calculating the LAS after this date (19). Patients listed before January 7, 2009, were excluded, as a revision to the LAS was implemented at that time. In addition, we excluded any patients listed for multiorgan transplantation to ensure that LAS was the primary determinant of priority for transplantation.
Patients without reported values of forced vital capacity (FVC) were excluded, as FVCpp could not be determined in these patients. Last, we excluded all patients with missing data for any variable used to calculate the LAS to improve the validity of our LAS calculation.
LAS Calculation
The UNOS LAS calculator that was in effect from January 7, 2009, to February 18, 2015, was used to determine the LAS at listing for each patient (20). The LAS is calculated by first determining both a waiting list urgency measure and a post-transplantation survival measure, each of which is derived using an independent formula with a related set of clinical variables, and they denote the expected number of days survived out of the next 365 days either without transplantation or after transplantation. The LAS is then calculated by combining these two measures and normalizing to a scale ranging from 0 to 100. The formulas used to calculate waiting list urgency, post-transplantation survival, and the LAS, as well as the process used to validate these calculations, are found in the data supplement.
Race-Specific and Race-Neutral Approaches
We determined the LAS using both race-specific and race-neutral approaches. For the race-specific approach, we calculated FVCpp using the Global Lung Function Initiative (GLI) equation corresponding to each patient’s reported race as recorded in the UNOS database. For the race-neutral approach, we calculated FVCpp using the GLI “other” equation for all patients regardless of race. The resultant FVCpp was then used to calculate either a race-specific LAS (RS-LAS) or a race-neutral LAS (RN-LAS).
For both approaches, all percentage predicted values were calculated using the online GLI calculator using each patient’s absolute FVC, age, sex, and height from time of initial listing (21). Although GLI equations were not developed until 2012, we used them in this study to evaluate the potential impact of currently recommended equations. The GLI other equation was selected as our race-neutral approach because it does not include a term for race/ethnicity, is averaged from four racial/ethnic groups, and has been previously studied as such an approach (2–5, 22).
Analysis of Difference in LAS between Approaches
Demographics and descriptive variables obtained from the UNOS databases were compared across Black and White race using Student’s t test and the Mann-Whitney U test for continuous variables and chi-square or Fisher exact tests for categorical variables. Mean differences in spirometry values, LAS, and predicted survival (without transplantation and after transplantation) were compared between White and Black individuals under both race-specific and race-neutral approaches, as well as the mean differences between the race-specific and race-neutral approaches for White and Black individuals using the previously described approaches. Differences in percentage predicted values by race were plotted across the range of observed raw values to assess systematic differences.
To evaluate for varying differences by LAS, we plotted the RS-LAS and RN-LAS differences against RS-LAS score by race. We also defined bins by every 10 points of RS-LAS and assessed the average difference between RS-LAS and RN-LAS for each bin across races to better define the range of LAS values across which the greatest difference was observed. In addition, subgroup analyses by diagnosis groupings (group A, obstructive lung disease; group B, pulmonary vascular disease; group C, cystic fibrosis or immunodeficiency disorder; and group D, restrictive lung disease) were performed.
To evaluate the potential impact of excluding patients with missing data, we conducted a sensitivity analysis evaluating the difference in LAS between approaches when including patients who were previously excluded because of missing values for the LAS calculations. For these individuals, specified default values were used according to UNOS policy and practice (see Tables E1 and E2 in the data supplement) (20).
A two-sided P value <0.05 was considered to indicate statistical significance. All statistical analyses were performed using Stata version 17.0 (StataCorp).
Results
Demographics
A total of 8,982 patients met the inclusion criteria, of whom 8,114 (90.3%) were White and 868 (9.7%) were Black (see Figure E1). The disease severity of this cohort was demonstrated by a mean reported percentage predicted forced expiratory volume in 1 second (FEV1pp) of 38% and FVCpp of 51% among all patients.
Demographics and clinical characteristics of the study population at the time of listing are presented in Table 1. The most common diagnosis grouping was group D (restrictive lung disease), with 4,474 (49.8%) patients. In general, Black patients were younger, were more commonly female, were less commonly ever-smokers, and had a higher average body mass index at listing than White patients. Black patients had multiple indicators of greater disease severity at listing, including a higher median O2 flow rate (3.0 vs. 3.5 L/min), a lower median 6-minute-walk distance (875 vs. 770 ft), lower measured forced expiratory volume in 1 second (FEV1) (1.2 vs. 1.1 L), and lower measured FVC (2.1 vs. 1.7 L) compared with White patients (P < 0.001 for all).
Table 1.
Demographics and clinical characteristics at time of lung transplantation listing
| White (n = 8,114) | Black (n = 868) | P Value* | |
|---|---|---|---|
| Age at listing, y, mean (SD) | 55.8 (13.1) | 52.6 (10.7) | <0.001 |
| Gender, n (%) | <0.001 | ||
| Female | 3,395 (41.8) | 520 (59.9) | |
| Male | 4,719 (58.2) | 348 (40.1) | |
| Height, cm, mean (SD) | 170.0 (9.9) | 168.4 (9.9) | <0.001 |
| BMI, kg/m2, mean (SD) | 25.4 (4.7) | 26.2 (4.7) | <0.001 |
| History of cigarette use, n (%) | <0.001 | ||
| No | 2,982 (36.8) | 394 (45.4) | |
| Yes | 5,130 (63.2) | 473 (54.6) | |
| Diagnosis group, n (%) | <0.001 | ||
| A (obstructive lung disease) | 2,797 (34.5) | 273 (31.5) | |
| B (pulmonary vascular disease) | 355 (4.4) | 69 (7.9) | |
| C (cystic fibrosis) | 1,000 (12.3) | 14 (1.6) | |
| D (restrictive lung disease) | 3,962 (48.8) | 512 (59.0) | |
| Supplemental O2 use, L/min, median (IQR) | 3.0 (2.0–5.0) | 3.5 (2.0–6.0) | <0.001 |
| 6-min-walk distance, ft, median (IQR) | 875 (578–1,143) | 770 (450–1,011) | <0.001 |
| Continuous mechanical ventilation, n (%) | 202 (2.5) | 18 (2.1) | 0.45 |
| ECMO, n (%) | 57 (0.7) | 10 (1.2) | 0.14 |
| FEV1, L, median (IQR) | 0.99 (0.63–1.63) | 0.98 (0.65–1.41) | 0.02 |
| FVC, L, median (IQR) | 1.92 (1.48–2.5) | 1.58 (1.18–2.03) | <0.001 |
| LAS at listing, reported, median (IQR) | 36.7 (33.5–43.6) | 38.1 (34.3–47.3) | <0.001 |
Definition of abbreviations: BMI = body mass index; ECMO = extracorporeal membrane oxygenation; FEV1 = forced expiratory volume in 1 second; FVC = forced vital capacity; IQR = interquartile range; LAS = lung allocation score; SD = standard deviation.
P values are based on Student’s t or Mann-Whitney U tests for continuous variables and chi-square or Fisher exact tests for categorical variables.
Individuals’ medical information used to calculate LAS is summarized in Table E3.
Percentage Predicted Spirometry
Spirometry values on the basis of race-specific and race-neutral approaches for White and Black individuals are shown in Table 2. Despite a lower raw FVC among Black patients compared with White patients, there was no statistically significant difference in FVCpp between White and Black patients using a race-specific approach (51.3% vs. 50.4%; P = 0.19). Moving to a race-neutral approach, however, FVCpp was higher by 4.4 percentage points on average for White patients and decreased by 3.8 percentage points for Black patients, resulting in an overall difference of 9 percentage points (55.7% in White patients vs. 46.7% in Black patients; P < 0.001).
Table 2.
Primary and secondary outcomes, compared by approach and stratified by race
| Race Specific | Race Neutral | Difference by Approach* | P Value† | |
|---|---|---|---|---|
| FEV1% predicted, mean (SD) | ||||
| White | 35.4 (21.8) | 38.0 (23.4) | +2.6 | <0.001 |
| Black | 38.6 (21.3) | 35.6 (19.6) | −3.0 | <0.001 |
| Difference by race‡ | −3.2 | +2.4 | — | — |
| FVC% predicted, mean (SD) | ||||
| White | 51.3 (18.0) | 55.7 (19.5) | +4.4 | <0.001 |
| Black | 50.4 (18.3) | 46.7 (16.9) | −3.8 | <0.001 |
| Difference by race‡ | +0.9 | +9.0 | — | — |
| LAS | ||||
| White | 41.9 (14.2) | 41.3 (14.1) | −0.6 | <0.001 |
| Black | 43.8 (15.4) | 44.3 (15.5) | +0.6 | <0.001 |
| Difference by race‡ | −1.9 | −3.0 | — | — |
| Predicted survival without transplantation, d, mean (SD) | ||||
| White | 294.8 (81.2) | 298.1 (80.4) | +3.3 | <0.001 |
| Black | 283.9 (87.3) | 280.6 (87.9) | −3.3 | <0.001 |
| Difference by race‡ | +10.9 | +17.5 | — | — |
| Predicted survival after transplantation, d, mean (SD) | ||||
| White | 318.1 (12.3) | 318.4 (12.1) | +0.3 | <0.001 |
| Black | 316.9 (11.9) | 316.6 (12.1) | −0.3 | <0.001 |
| Difference by race‡ | +1.2 | +1.8 | — | — |
Definition of abbreviations: FEV1 = forced expiratory volume in 1 second; FVC = forced vital capacity; LAS = lung allocation score; SD = standard deviation.
Positive values indicate greater values with a race-neutral approach.
P values are based on Student’s t tests.
Positive values indicate greater values among White patients.
Similarly, White patients had a 2.6 percentage point higher mean FEV1pp under a race-neutral approach compared with a race-specific approach (38.0% vs. 35.4%; P < 0.001), whereas Black patients had a 3.0 percentage point lower mean FEV1pp under a race-neutral compared with a race-specific approach (35.6% vs. 38.6%; P < 0.001). When using a race-specific approach, White individuals had a 3.2 percentage point lower mean FEV1pp than Black individuals, whereas when using a race-neutral approach, White individuals had, on average, a 2.4 percentage point higher FEV1pp than Black individuals (P < 0.001).
The differences in FVCpp and FEV1pp between the race-specific and race-neutral approaches varied over the range of raw FEV1 and FVC, and this difference appeared to increase as lung volume increased (Figure 1).
Figure 1.
Difference in percentage predicted spirometry between race-specific (RS) and race-neutral (RN) approaches as a function of measured spirometry. The difference in percentage predicted spirometry measures when using an RN compared with an RS approach is shown over the range of actual (measured) spirometry values for White and Black individuals. Positive values indicate higher percentage predicted spirometry under an RN approach. Each point represents an individual at the time of listing. Best fitting linear lines with 95% confidence intervals are presented. Differences in percentage predicted spirometry measures increase as measured lung volumes increase. FEV1 = forced expiratory volume in 1 second; FVC = forced vital capacity.
LAS
The LASs calculated using race-specific and race-neutral approaches to interpreting spirometry for White and Black patients are shown in Table 2. Compared with White patients, Black patients had a significantly higher mean LAS at listing under both race-specific (41.9 vs. 43.8; P < 0.001) and race-neutral (41.3 vs. 44.3; P < 0.001) approaches (Table 2).
On average, White patients had RN-LAS that were 0.6 points lower than their RS-LAS, compared with Black patients, who had RN-LAS that were 0.6 points higher than their RS-LAS (P < 0.001). Notably, the differences in LAS between approaches varied according to RS-LAS and were characterized by a U-shaped (or an inverted U–shaped) relationship, with differences in LAS more pronounced among those with RS-LASs between approximately 40 and 80 (Figure 2A). When the cohort was divided into bins by 10-point intervals, the average differences between RN-LAS and RS-LAS across the range of 40–80 were approximately 1 LAS point lower and higher, respectively, for White and Black patients (see Table E4).
Figure 2.
Difference in lung allocation score (LAS) and waiting list urgency between race-specific (RS) and race-neutral (RN) approaches as a function of RS-LAS at listing. (A and B) Differences in (A) LAS and (B) predicted survival without transplantation between the RS and RN approaches are shown over the range of the RS-LAS at listing. The RS-LAS on the x-axis was calculated using percentage predicted forced vital capacity from RS reference equations. Each point represents an individual at the time of listing. Best fitting quadratic lines with 95% confidence intervals are displayed for White and Black individuals. Positive values indicate (A) higher LAS and (B) greater number of predicted days survived under an RN approach.
In subgroup analyses by diagnosis groupings, the change in LAS was greatest in group B (−0.71 vs. +0.71; P < 0.0001) and group D (−0.78 vs. +0.68; P < 0.001). Patients in group B had the highest change in FVCpp (Table 3). The difference in LAS between approaches had a similar relationship with RS-LAS across all subgroups, and group D appeared to have the greatest proportion of patients with LASs between 40 and 80 (see Figure E2). Full results of subgroup analyses by diagnosis grouping are shown in Table 3.
Table 3.
Difference in lung allocation score and percentage predicted forced vital capacity between approaches among diagnosis groups and stratified by race
| Difference in LAS*† |
Difference in Percentage Predicted FVC*† |
|||
|---|---|---|---|---|
| Diagnosis Group | White | Black | White | Black |
| A (obstructive lung disease) | −0.30 (0.21) | +0.33 (0.22) | +4.8 (1.5) | −3.9 (1.3) |
| B (pulmonary vascular disease) | −0.71 (0.32) | +0.70 (0.40) | +5.9 (1.8) | −4.6 (1.8) |
| C (cystic fibrosis) | −0.50 (0.25) | +0.44 (0.27) | +3.5 (1.0) | −3.2 (1.1) |
| D (restrictive lung disease) | −0.78 (0.45) | +0.68 (0.44) | +4.3 (1.5) | −3.6 (1.3) |
Definition of abbreviations: FVC = forced vital capacity; LAS = lung allocation score.
Values are presented as mean (standard deviation).
Positive values indicate greater values with a race-neutral approach.
P < 0.0001 for all comparisons between White and Black patients. P values are based on Student’s t tests.
Predicted Survival
Under a race-specific approach, predicted survival on the waiting list was on average higher among White than Black patients by approximately 11 days (294.8 vs. 283.9 days; P < 0.001). The difference in predicted survival under a race-neutral approach was greater among White patients, who had mean predicted survival that was 3.3 days longer, whereas Black patients had predicted survival that was 3.3 days shorter (P < 0.001) (Table 2). Similar to the LAS, the difference in predicted survival without transplantation between approaches was most pronounced in patients with RS-LASs ranging from 40 to 80 (Figure 2B), with a difference of approximately 1 week in predicted survival in opposite directions for each race (see Table E2).
For predicted survival after transplantation, White patients had mean predicted survival of 318.1 days compared with 316.9 days for Black patients (P < 0.001; Table 2). There was a small but statistically significant difference in predicted survival after transplantation between approaches (+0.3 vs. −0.3 days; P < 0.001).
Sensitivity Analyses
In the sensitivity analysis, an additional 3,510 patients were included, and the characteristics are shown in Table E5. Similar to the primary analysis, White patients had an RN-LAS that was 0.6 points lower than the RS-LAS, and Black patients had an RN-LAS that was 0.6 points higher (see Table S6).
Discussion
Our findings raise concern that a race-specific approach to spirometry interpretation may contribute to racial bias in respiratory disease through an impact on priority for lung transplantation. These results are particularly significant in the context of interest in an evidence-based understanding of how a race-specific approach influences patient care. Overall, in this population-based study of individuals listed for lung transplantation between 2009 and 2015, we demonstrated that a race-neutral approach, in which a single reference equation was applied to all patients regardless of race, would have resulted in a higher average LAS for Black patients and a lower average LAS for White patients compared with a race-specific approach.
Importantly, our results demonstrate how the use of race-specific equations may affect patients with severe respiratory disease. Recent data have suggested that race-specific equations may underestimate disease prevalence, severity, or outcomes among Black patients; however, these studies have generally assessed healthier cohorts that are not defined by the presence of respiratory disease and that consist of research participants (3–5, 22). In contrast, our cohort consisted of data from real-world patients with severe respiratory disease. This study fills a previously identified research gap by demonstrating that race-specific equations may have biased the care of those awaiting lung transplantation. More broadly, our finding reflects wider concerns in pulmonology that race-specific equations contribute to the undertreatment of respiratory disease among Black patients (3–5, 22).
In addition, our results demonstrated that the inclusion of race in estimating lung function had a clear impact on the LAS. Our findings align with previous suggestions that race-specific equations may reduce access to lung transplantation for Black compared with White patients (8, 9, 23). Although we did not investigate measures of access to transplantation, such as time to transplantation or probability of transplantation, the LAS is the primary determinant of priority for lung transplantation in the United States, and the fairness of the LAS has been identified as a key theme in considering equity of lung transplant allocation (10, 24, 25). Thus, by generating lower LASs for Black patients, race-specific equations may have contributed to previously reported racial disparities in access to lung transplantation (14, 15). In addition, because Black and White patients have the greatest magnitude of difference in FVCpp (and thus likely in LAS) between race-specific and race-neutral approaches, the average difference in LAS for Asian patients would likely be somewhere between −0.6 and +0.6, the differences found for White and Black patients, respectively (2). Regardless of the exact effect size, Asian patients would have been disadvantaged relative to White patients in terms of the LAS under a race-specific system compared with a race-neutral one.
Notably, we found that the difference in LAS between race-specific and race-neutral approaches was most pronounced across the range of values wherein a change in LAS might have a pronounced impact on waiting list time. A slight change in LAS among patients with very high or very low LASs will be of minimal impact on their time on the waiting list. However, we found the greatest magnitude of difference in LAS between approaches in patients with LASs between 40 and 80, which is in the middle of the LAS range. The importance of the LAS in this range is emphasized by Organ Procurement and Transplantation Network data demonstrating that an increasing proportion of wait-listed patients have LASs >50 and that the median LAS at transplantation is consistently greater than 40 (26).
To provide additional context for our findings, the coefficients for the LAS (see Table E5) indicate that a change in arterial partial pressure of carbon dioxide of 15 mm Hg would produce a roughly equivalent change in the LAS as found in our study between a race-specific and a race-neutral approach. This analogy reflects that the difference in LAS between approaches in our study could be indicative of meaningful underlying clinical changes.
Furthermore, we found that the choice of a race-specific or a race-neutral approach had the greatest impact on patients with group D diagnoses, the most common grouping. This is likely because group D included a higher proportion of patients listed with LASs in the range of 40–80, as shown in Figure E2. Interestingly, there was a similarly high difference between approaches in group B despite comparatively few patients listed with LASs of 40–80. As shown in Table 3, this is likely because patients in group B had the greatest change in FVCpp, which itself is likely a result of their greater underlying lung function (Figure 1).
Our results are also novel in demonstrating the impact of race-specific equations on percentage predicted spirometry values in patients with advanced respiratory disease, and these data suggest that race-specific equations have a greater impact on respiratory health disparities earlier in the course of disease progression. The differences in percentage predicted values between race-specific and race-neutral approaches were observed to be larger at higher absolute lung functions. This explains why we found on average a 4% difference in FVCpp between approaches, compared with prior studies in healthier populations that revealed differences closer to 7% (3–5). The differing impact of race-specific versus race-neutral approaches across lung function severity suggests that a race-specific approach is most likely to bias care at earlier stages of disease. This upstream impact is particularly relevant to lung transplantation because differences in FVCpp can affect whether a patient is listed for transplantation, and the subsequent clinical and psychological effects of not being listed may be substantial.
In interpreting our findings, it must be noted that the FVCpp was included only for patients in group D after February 18, 2015, and it was recently removed altogether from the LAS in September 2021 (11, 27). Although patients listed after February 18, 2015, were not included in our study, it is likely that race-specific equations would have had a very similar impact on the LAS for those in group D (who would have had FVCpp incorporated into their LASs until September 2021) as those in our cohort on the basis of the similarity in FVCpp coefficient values in the LAS equation before and after that date (20). In addition, the lung transplant allocation system is currently pending further change from the LAS to the continuous allocation system (28), which will still include current components of the LAS but will also incorporate biological factors (e.g., donor compatibility), patient access, and placement efficiency. These changes demonstrate that the U.S. lung allocation system is continually evolving and iteratively updated. Spirometry metrics (FVCpp or otherwise) are highly likely to be considered for inclusion in future updates given their broad importance in assessing respiratory disease severity. Thus, our results provide a strong rationale to carefully consider which spirometry metrics should be evaluated for inclusion in the future.
Strengths and Limitations
Our study has a few limitations. First, our results do not demonstrate whether a race-specific or a race-neutral approach is a more accurate measure of lung disease, but they do provide novel evidence on how race-specific equations may contribute to racial disparities in lung transplantation.
Second, we relied on race as reported electronically, which may be misreported in the electronic health record, particularly among minorities (29). However, race as listed in the electronic health record is probably what is used by the embedded algorithms that produce FVCpp. This practice means that our results are likely indicative of clinical practice but also underscores inherent problems with using race in clinical algorithms such as spirometry equations.
Last, our study cannot conclude how a race-specific versus a race-neutral approach affects outcomes such as the probability of transplantation or time to transplantation. Multiple other factors beyond LAS influence this, such as size, blood type, preformed antibodies, and listing center (30, 31). However, because of the primary importance of LAS in determining allocation, the use of a race-specific versus a race-neutral interpretation of spirometry has the potential to have some effect on access to transplantation.
Conclusions
Compared with a race-neutral approach to spirometry interpretation, a race-specific approach results in lower LASs for Black patients and higher LASs for White patients. As such, a race-specific approach may have contributed to decreased access to lung transplantation among Black compared with White patients on the waiting list. Race-specific equations may promote inequitable care, and their future use must be carefully considered.
Footnotes
Supported by National Heart, Lung, and Blood Institute grant T32HL007534 (J.H.B.) and National Institutes of Health grants K23 HL153778 (A.B.), R01HL152419 (M.C.M.), R61HL157845 (M.C.M.), R01HL154860 (M.C.M.), P50ES018176 (M.C.M.), and P2CES033415 (M.C.M.). The funders had no role in the study design, data collection, analysis, or interpretation of this study.
Author Contributions: Conception and design of the study: J.H.B., A.B., K.J.P., C.A.M., and M.C.M.; data acquisition: J.H.B., P.S., E.L.B., C.A.M., and M.C.M.; analysis and data interpretation: J.H.B., A.B., K.J.P., P.S., E.L.B., C.A.M., and M.C.M.; drafting and critical revision of the manuscript: J.H.B., A.B., K.J.P., P.S., E.L.B., C.A.M., and M.C.M.
This article has a data supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.
Author disclosures are available with the text of this article at www.atsjournals.org.
References
- 1. Graham BL, Steenbruggen I, Miller MR, Barjaktarevic IZ, Cooper BG, Hall GL, et al. Standardization of spirometry 2019 update: an official American Thoracic Society and European Respiratory Society technical statement. Am J Respir Crit Care Med . 2019;200:e70–e88. doi: 10.1164/rccm.201908-1590ST. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Quanjer PH, Stanojevic S, Cole TJ, Baur X, Hall GL, Culver BH, et al. ERS Global Lung Function Initiative Multi-ethnic reference values for spirometry for the 3-95-yr age range: the global lung function 2012 equations. Eur Respir J . 2012;40:1324–1343. doi: 10.1183/09031936.00080312. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Baugh AD, Shiboski S, Hansel NN, Ortega V, Barjaktarevic I, Barr RG, et al. Reconsidering the utility of race-specific lung function prediction equations. Am J Respir Crit Care Med . 2022;205:819–829. doi: 10.1164/rccm.202105-1246OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Elmaleh-Sachs A, Balte P, Oelsner EC, Allen NB, Baugh A, Bertoni AG, et al. Race/ethnicity, spirometry reference equations, and prediction of incident clinical events: the Multi-Ethnic Study of Atherosclerosis (MESA) lung study. Am J Respir Crit Care Med. 2022;205:700–710. doi: 10.1164/rccm.202107-1612OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. McCormack MC, Balasubramanian A, Matsui EC, Peng R, Wise RA, Keet CA. Race, lung function and long-term mortality in the National Health and Examination Survey III. Am J Respir Crit Care Med . 2022;205:723–724. doi: 10.1164/rccm.202104-0822LE. [DOI] [PubMed] [Google Scholar]
- 6. Anderson MA, Malhotra A, Non AL. Could routine race-adjustment of spirometers exacerbate racial disparities in COVID-19 recovery? Lancet Respir Med . 2021;9:124–125. doi: 10.1016/S2213-2600(20)30571-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Schluger NW. The vanishing rationale for the race adjustment in pulmonary function test interpretation. Am J Respir Crit Care Med . 2022;205:612–614. doi: 10.1164/rccm.202112-2772ED. [DOI] [PubMed] [Google Scholar]
- 8. Kaminsky DA. Is there a role for using race-specific reference equations? Yes and no. Am J Respir Crit Care Med . 2022;205:746–748. doi: 10.1164/rccm.202201-0006ED. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Bhakta NR, Balmes JR. A good fit versus one size for all: alternatives to race in the interpretation of pulmonary function tests. Am J Respir Crit Care Med . 2022;205:616–618. doi: 10.1164/rccm.202201-0076ED. [DOI] [PubMed] [Google Scholar]
- 10. Egan TM, Murray S, Bustami RT, Shearon TH, McCullough KP, Edwards LB, et al. Development of the new lung allocation system in the United States. Am J Transplant . 2006;6:1212–1227. doi: 10.1111/j.1600-6143.2006.01276.x. [DOI] [PubMed] [Google Scholar]
- 11.Miller E.2020. https://optn.transplant.hrsa.gov/media/4206/bp_202012_updated-cohort-for-calculation-of-the-lung-allocation-score.pdf
- 12. Wille KM, Harrington KF, deAndrade JA, Vishin S, Oster RA, Kaslow RA. Disparities in lung transplantation before and after introduction of the lung allocation score. J Heart Lung Transplant . 2013;32:684–692. doi: 10.1016/j.healun.2013.03.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Lederer DJ, Benn EKT, Barr RG, Wilt JS, Reilly G, Sonett JR, et al. Racial differences in waiting list outcomes in chronic obstructive pulmonary disease. Am J Respir Crit Care Med . 2008;177:450–454. doi: 10.1164/rccm.200708-1260OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Mooney JJ, Hedlin H, Mohabir P, Bhattacharya J, Dhillon GS. Racial and ethnic disparities in lung transplant listing and waitlist outcomes. J Heart Lung Transplant . 2018;37:394–400. doi: 10.1016/j.healun.2017.09.017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Riley LE, Lascano J. Gender and racial disparities in lung transplantation in the United States. J Heart Lung Transplant . 2021;40:963–969. doi: 10.1016/j.healun.2021.06.004. [DOI] [PubMed] [Google Scholar]
- 16.Prentice MA.2018. https://optn.transplant.hrsa.gov/media/2565/geography_publiccomment_201808.pdf
- 17. Egan TM. Ethical issues in thoracic organ distribution for transplant. Am J Transplant . 2003;3:366–372. doi: 10.1034/j.1600-6143.2003.00039.x. [DOI] [PubMed] [Google Scholar]
- 18. U.S. Department of Health and Human Services, Health Resources and Services Administration. 42 CFR part 121: Organ Procurement and Transplantation Network; final rule. Fed Reg . 1999;64:56650–56661. [PubMed] [Google Scholar]
- 19.Organ Procurement and Transplantation Network. 2015. https://optn.transplant.hrsa.gov/media/1576/policynotice_20150201.pdf
- 20.United Network for Organ Sharing. 2020. https://unos.org/wp-content/uploads/unos/lung-allocation-score.pdf
- 21.Global Lung Function Initiative. 2021. http://gli-calculator.ersnet.org
- 22. Liu GY, Khan SS, Colangelo LA, Meza D, Washko GR, Sporn PHS, et al. Comparing racial differences in emphysema prevalence among adults with normal spirometry: a secondary data analysis of the CARDIA lung study. Ann Intern Med . 2022;175:1118–1125. doi: 10.7326/M22-0205. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Bhakta NR, Kaminsky DA, Bime C, Thakur N, Hall GL, McCormack MC, et al. Addressing race in pulmonary function testing by aligning intent and evidence with practice and perception. Chest . 2022;161:288–297. doi: 10.1016/j.chest.2021.08.053. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Merlo CA, Weiss ES, Orens JB, Borja MC, Diener-West M, Conte JV, et al. Impact of U.S. lung allocation score on survival after lung transplantation. J Heart Lung Transplant . 2009;28:769–775. doi: 10.1016/j.healun.2009.04.024. [DOI] [PubMed] [Google Scholar]
- 25. Schnellinger EM, Cantu E, Kimmel SE, Szymczak JE. A conceptual model for sources of differential selection in lung transplant allocation. Ann Am Thorac Soc . 2023;20:226–235. doi: 10.1513/AnnalsATS.202202-105OC. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Valapour M, Lehr CJ, Skeans MA, Smith JM, Uccellini K, Lehman R, et al. OPTN/SRTR 2017 annual data report: lung. Am J Transplant . 2019;19:404–484. doi: 10.1111/ajt.15279. [DOI] [PubMed] [Google Scholar]
- 27.United Network for Organ Sharing. 2021. https://unos.org/news/changes-to-the-updated-cohort-for-calculation-of-the-lung-allocation-score-las-are-now-in-effect/
- 28.Organ Procurement and Transplantation Network. 2022. https://optn.transplant.hrsa.gov/professionals/by-organ/heart-lung/lung-continuous-distribution-policy/?j=628641&sfmc_sub=84493098&l=14_HTML&u=27931812&mid=100001876&jb=4#TK_Policy
- 29. Samalik JM, Goldberg CS, Modi ZJ, Fredericks EM, Gadepalli SK, Eder SJ, et al. Discrepancies in race and ethnicity in the electronic health record compared to self-report. J Racial Ethn Health Disparities . 2022 doi: 10.1007/s40615-022-01445-w. [DOI] [PubMed] [Google Scholar]
- 30. Benvenuto LJ, Arcasoy SM. The new allocation era and policy. J Thorac Dis . 2021;13:6504–6513. doi: 10.21037/jtd-2021-17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Benvenuto LJ, Anderson DR, Kim HP, Hook JL, Shah L, Robbins HY, et al. From the Columbia University Lung Transplant Program Geographic disparities in donor lung supply and lung transplant waitlist outcomes: a cohort study. Am J Transplant . 2018;18:1471–1480. doi: 10.1111/ajt.14630. [DOI] [PubMed] [Google Scholar]