Racial disparities in treatment patterns, healthcare resource use, and outcomes in patients with pulmonary arterial hypertension in the United States

Figures & data

Figure 1. (A) Study design; (B) study eligibility and patient disposition. ^aIndex PDE5i dosage, dosage form, and route of administration for PDE5i claims must indicate PAH use and ratio of pill count to days of supply must be ≥1. Abbreviations. CTEPH, chronic thromboembolic pulmonary hypertension; PAH, pulmonary arterial hypertension; PDE5i, phosphodiesterase (Type 5) enzyme inhibitor.

Table 1. Demographics at baseline in the Black patients and White patients with PAH.

Characteristic	Black patients (N = 836)	White patients (N = 2896)	SMD^a
Male sex, n (%)	205 (24.5)	1017 (35.1)	0.233
US region, n (%)			0.751
Midwest	132 (15.8)	665 (23.0)
Northeast	62 (7.4)	314 (10.8)
South	599 (71.7)	1165 (40.2)
West	43 (5.1)	752 (26.0)
Medicare Advantage insurance, n (%)	712 (85.2)	2357 (81.4)	0.101
Age on index date, mean years (SD)	67.4 (12.6)	70.6 (11.8)	0.259
Age category, years, n (%)			0.273
18 to <45	53 (6.3)	90 (3.1)
45 to <55	69 (8.3)	194 (6.7)
55 to <65	159 (19.0)	465 (16.1)
65 to <75	298 (35.6)	932 (32.2)
75 to <85	209 (25.0)	929 (32.1)
≥85	48 (5.7)	286 (9.9)
Highest education level, n (%)			0.65
High school diploma or lower	444 (53.1)	721 (24.9)
Lower than Bachelor’s degree	347 (41.5)	1727 (59.6)
Bachelor’s degree or higher	36 (4.3)	423 (14.6)
Unknown	9 (1.1)	25 (0.9)
Annual household income, US$, n (%)			0.76
<40,000	483 (57.8)	868 (30.0)
40,000 to <50,000	110 (13.2)	274 (9.5)
50,000 to <60,000	65 (7.8)	260 (9.0)
60,000 to <75,000	64 (7.7)	343 (11.8)
75,000 to <100,000	59 (7.1)	444 (15.3)
≥100,000	34 (4.1)	617 (21.3)
Unknown	21 (2.5)	90 (3.1)

^a SMD <0.1 predefined as indicating covariate balance between the cohorts (based on propensity score weighting/matching).

Abbreviations. PAH, pulmonary arterial hypertension; SD, standard deviation; SMD, standardized mean difference.

Table 2. Duration of follow-up.

	Median duration, months (IQR)
	Black patients	White patients	SMD^a
Overall cohort	(N = 836)	(N = 2896)
	14.72 (6.51–30.24)	15.71 (6.80–31.29)	0.031
Age <65 years at index	(n = 281)	(n = 749)
	14.08 (6.25–28.38)	16.04 (6.80–32.18)	0.081
Age ≥65 years at index	(n = 555)	(n = 2147)
	15.25 (6.92–30.70)	15.58 (6.79–31.13)	0.009

^a SMD <0.1 predefined as indicating covariate balance between the cohorts (based on propensity score weighting/matching).

Abbreviations. IQR, interquartile range; SMD, standardized mean difference.

Figure 2. Clinical characteristics of patients with PAH recorded during the 6-month baseline period. (A) Quan-Charlson Comorbidity Index score; (B) most common comorbidities (>20% of patients in either group); (C) PAH-related symptoms. SMD >0.1 indicates difference in index scores between the groups. ^aIndicates comorbidities with SMD >0.1, indicating difference in frequency between the groups. ^bIndicates symptoms with SMD >0.1, indicating difference in frequency between the groups. Abbreviations. COPD, chronic obstructive pulmonary disorder; QCCI, Quan-Charlson Comorbidity Index; PAH, pulmonary arterial hypertension; PVD, peripheral vascular disease; SD, standard deviation; SMD, standardized mean difference.

Table 3. Index PAH treatment pattern in the Black patients and White patients.

Characteristic	Black patients (N = 836)	White patients (N = 2896)	SMD^a
Treatment index date on and after year 2020, n (%)	376 (45.0)	1424 (49.2)	0.084
Combination therapy, n (%)	120 (14.4)	419 (14.5)	0.003
PPA containing therapy, n (%)	85 (10.2)	425 (14.7)	0.137
ERA containing therapy, n (%)	183 (21.9)	671 (23.2)	0.031
Proportion of days covered, median (IQR)	0.96 (0.83–1.00)	0.99 (0.88–1.00)	0.155
Index treatment adherence (proportion of days covered ≥0.8), n (%)	547 (85.3)	1965 (89.9)	0.14
Days to discontinuation, KM median (IQR)	288.64 (351.46)	278.54 (340.63)	0.029
Index treatment discontinuation pattern, n (%)			0.102
1. Add a new PAH medication	130 (15.6)	419 (14.5)
2. Switch to a new PAH medication/regimen	25 (3.0)	80 (2.8)
3. Discontinue index PAH medication with restart after >90 days	104 (12.4)	306 (10.6)
4. Discontinue PAH medication without restarting	449 (53.7)	1697 (58.6)
5. Data cutoff/Insurance disenrollment	128 (15.3)	394 (13.6)

^a SMD <0.1 predefined as indicating covariate balance between the cohorts (based on propensity score weighting/matching).

Abbreviations. ERA, endothelin receptor antagonist; IQR, interquartile range; KM, Kaplan–Meier; PAH, pulmonary arterial hypertension; PPA, prostacyclin pathway agent; SMD, standardized mean difference.

Table 4. Health care resource utilization and costs during the 6-month baseline period.

Variable	Black patients (N = 836)	White patients (N = 2896)	SMD^a
Baseline health care resource utilization rate, per patient per month
Any all-cause hospitalization, n (%)	526 (62.9)	1584 (54.7)	0.168
Any all-cause ED visit, n (%)	212 (25.4)	644 (22.2)	0.073
Any all-cause ICU stay, n (%)	299 (35.8)	849 (29.3)	0.138
Number of all-cause health care resource utilization, median (IQR)	7.69 (5.52–11.03)	7.19 (5.02–10.20)	0.145
Number of PAH-related health care resource utilization, median (IQR)^b	0.67 (0.33–1.34)	0.84 (0.33–1.34)	0.005
Number of all-cause outpatient visits, median (IQR)	2.34 (1.50–3.51)	2.34 (1.34–3.51)	0.141
Number of non-PAH prescriptions, median (IQR)	2.67 (1.84–3.84)	2.51 (1.50–3.68)	0.099
Number of pulmonologist visits, median (IQR)	0.33 (0.00–0.84)	0.33 (0.00–0.67)	0.087
Number of cardiologist visits, median (IQR)	0.84 (0.33–1.50)	0.67 (0.33–1.34)	0.175
Baseline costs, US$, median (IQR), per patient per month^c
All-cause costs	64,805 (14,649–154,932)	33,348 (9217–119,368)	0.128
All-cause costs excluding pharmacy	63,216 (14,100–154,640)	31,858 (8316–118,630)	0.128
PAH-related costs^b	1403 (161–4531)	1346 (159–3237)	0.102
All-cause hospitalization costs	34,852 (0–124,061)	1227 (0–89,619)	0.121
All-cause ED visits costs	0 (0–1)	0 (0–0)	0.105
All-cause outpatient visits costs	6186 (1509–18,874)	6351 (1436–16,698)	0.028
All-cause outpatient prescription costs	380 (120–785)	313 (89–715)	0.022

^a SMD <0.1 predefined as indicating covariate balance between the cohorts (based on propensity score weighting/matching).

^b Utilization and costs are considered to be PAH-related if the claims list pulmonary hypertension as the primary diagnosis code.

^c To account for differences in pricing across health plans and provider contracts, Optum applies standard pricing algorithms to the claims data in Clinformatics Data Mart²⁷. These algorithms are designed to create standard prices that reflect allowed payments across all provider services. For example, professional service rates are standardized using a resource-based relative value scale approach²⁸.

Abbreviations. ED, emergency department; ICU, intensive care unit; IQR, interquartile range, PAH, pulmonary arterial hypertension; SMD, standardized mean difference.

Table 5. Health care resource utilization and costs during the follow-up period.

Variable	Black patients (N = 718)	White patients (N = 2454)	SMD^a
Follow-up health care resource utilization rate, per patient per month
Any all-cause hospitalization, n (%)	311 (43.3)	985 (40.1)	0.064
Any all-cause ED visit, n (%)	201 (28.0)	548 (22.3)	0.131
Any all-cause ICU stay, n (%)	184 (25.6)	542 (22.1)	0.083
Number of all-cause health care resource utilization, median (IQR)	3.91 (1.23–7.18)	3.50 (1.16–6.30)	0.108
Number of PAH-related health care resource utilization, median (IQR)^b	0.33 (0.07–0.93)	0.35 (0.07–0.87)	<0.001
Number of all-cause outpatient visits, median (IQR)	1.04 (0.31–2.22)	0.86 (0.21–2.04)	0.141
Number of non-PAH prescriptions, median (IQR)^b	1.34 (0.37–2.63)	1.14 (0.35–2.45)	0.042
Number of all-cause prescriptions, median (IQR)	1.54 (0.47–2.90)	1.33 (0.42–2.77)	0.051
Number of pulmonologist visits, median (IQR)	0.00 (0.00–0.17)	0.00 (0.00–0.17)	0.01
Number of cardiologist visits, median (IQR)	0.09 (0.00–0.36)	0.05 (0.00–0.29)	0.061
Follow-up costs, US$, median (IQR), per patient per month^c
All-cause costs	10,284 (2089–40,832)	7533 (1420–32,641)	0.064
All-cause costs excluding prescriptions	5960 (780–36,276)	3849 (563–29,332)	0.064
PAH–related costs^b	526 (49–2589)	398 (39–2423)	0.027
PAH-related costs excluding prescriptions^b	172 (17–1269)	143 (13–1172)	0.045
All-cause hospitalization costs	<1 (<1–27,389)	<1 (<1–17,035)	0.058
All-cause ED visits costs	<1 (<1–89)	<1 (<1 – <1)	0.04
All-cause outpatient visits costs	1107 (168–4425)	791 (114–3714)	0.045
All-cause outpatient prescription costs	562 (111–1993)	500 (99–2356)	0.006
Hospitalization (total) length of stay, days
During baseline, median (IQR)	5.00 (0.00–14.00)	2.00 (0.00–10.00)	0.218
During follow-up, median (IQR)	0.00 (0.00–10.00)	0.00 (0.00–7.00)	0.142
Clinical outcomes during follow-up, n (%)
Lung transplant	2 (0.2)	8 (0.3)	0.007
Atrial septostomy	NA	NA	<0.001
Oxygen therapy	479 (57.3)	1642 (56.7)	0.012
Lung transplant, atrial septostomy, oxygen therapy or all-cause hospitalization	651 (77.9)	2184 (75.4)	0.058

^a SMD <0.1 predefined as indicating covariate balance between the cohorts (based on propensity score weighting/matching).

^b Utilization and costs are considered to be PAH-related if the claims list pulmonary hypertension as the primary diagnosis code.

^c To account for differences in pricing across health plans and provider contracts, Optum applies standard pricing algorithms to the claims data in Clinformatics Data Mart²⁷. These algorithms are designed to create standard prices that reflect allowed payments across all provider services. For example, professional service rates are standardized using a resource-based relative value scale approach²⁸.

Abbreviations. ED, emergency department; ICU, intensive care unit; IQR, interquartile range, NA, not applicable; PAH, pulmonary arterial hypertension; SMD, standardized mean difference.

Table 6. Inverse probability of treatment-weighted regression analysis of outcomes comparing Black patients versus White patients with PAH.

Adjusted covariates^a	IPTW-regression Ratio (95% CI): Black patients versus White patients
	All patients	Age <65 years	Age ≥65 years
Index treatment strategy: combination versus monotherapy
Odds ratio^a	0.92 (0.66–1.28)	0.64 (0.41–0.99)	1.13 (0.74–1.72)
Index treatment adherence
Odds ratio^a	0.8 (0.59–1.10)	0.54 (0.33–0.90)	0.98 (0.67–1.45)
All-cause health care resource utilization (rate: count per patient per month)
Rate ratio^a	0.97 (0.85–1.11)	1.24 (1.05–1.46)	0.89 (0.75–1.05)
All-cause cost (multiplicative change in mean: US$, per patient per month)
Cost ratio^a	1.18 (0.91–1.52)	1.75 (1.09–2.80)	0.99 (0.75–1.30)
Composite clinical outcome (lung transplant, atrial septostomy, oxygen therapy, or all-cause hospitalization)
Hazard ratio^a	1.05 (0.90–1.21)	1.24 (1.01–1.52)	0.98 (0.82–1.18)

^a Adjusted for demographics (age, sex, census region) plus comorbidity index, PAH-related symptoms, and procedures.

Abbreviations. CI, confidence interval; IPTW, inverse probability of treatment-weighted; PAH, pulmonary arterial hypertension.

OptumInsight. Clinformatics data mart user manual. 2007. Version 7.1. Initial Release November 2017.

 Google Scholar

Buntin MB, Escarce J, Goldman D, et al. Determinants of Increases in Medicare Expenditures for Physicians’ Services. Rockville (MD): Agency for Healthcare Research and Quality (US); 2003 Oct. (Technical Reviews, No. 7.) Appendix A, Payment for Physicians’ Services Under the Resource Based Relative Value Scale [cited 2023 Sep 10]. Available from: https://www.ncbi.nlm.nih.gov/books/NBK43879/pdf/Bookshelf_NBK43879.pdf

 Google Scholar

Data availability statement

The authors confirm that the data supporting the findings of this study are available within the article and its supplementary materials.