Inpatient Palliative Care Use in Patients With Pulmonary Arterial Hypertension

Abstract

Background

Pulmonary arterial hypertension (PAH) is a progressive disease associated with significant morbidity and mortality. Despite the negative impact of PAH on quality of life and survival, data on use of specialty palliative care services (PCS) is scarce.

Research Question

We sought to evaluate the inpatient use of PCS in patients with PAH.

Study Design and Methods

Using the National (Nationwide) Inpatient Sample, 30,495 admissions with a primary diagnosis of PAH were identified from 2001 through 2017. The primary outcome of interest was temporal trends and predictors of inpatient PCS use in patients with PAH.

Results

The inpatient use of PCS was low (2.2%), but increased during the study period from 0.5% in 2001 to 7.6% in 2017, with a significant increase starting in 2009. White race, private insurance, higher socioeconomic status, hospital-specific factors, higher comorbidity burden (Charlson Comorbidity Index), cardiac and noncardiac organ failure, and use of extracorporeal membrane oxygenation and noninvasive mechanical ventilation were independent predictors of increased PCS use. PCS use was associated with a higher prevalence of do-not-resuscitate status, a longer length of stay, higher hospitalization costs, and increased in-hospital mortality with less frequent discharges to home, likely because these patients were also sicker (higher comorbidity index and illness acuity).

Interpretation

The inpatient use of PCS in patients with PAH is low, but has been increasing over recent years. Despite increased PCS use over time, patient- and hospital-specific disparities in PCS use continue. Further studies evaluating these disparities and the role of PCS in the comprehensive care of PAH patients are warranted.

FOR EDITORIAL COMMENT, SEE PAGE 2260

Pulmonary arterial hypertension (PAH) is a progressive disease characterized by elevated pulmonary artery pressure and pulmonary vascular resistance with normal left-sided filling pressures.¹ PAH has an estimated prevalence of 10 to 50 per 1 million and can be idiopathic, familial, or associated with conditions such as congenital heart disease, connective tissue disease, drugs or toxin use, and portal hypertension.^{2, 3, 4, 5} Symptoms of PAH include dyspnea, fatigue, chest pain, and dizziness. After right heart failure develops, symptoms also can include edema, weight gain, and abdominal bloating. PAH negatively impacts physical, social, and emotional health-related quality of life (QOL) and survival.^{6, 7, 8} PAH-specific medications improve symptoms, slow the progression of disease, and reduce hospitalizations, but are not curative and commonly are associated with significant side effects.^3,4

Despite improved survival over the last few decades related to advances in pharmacologic treatment options, PAH remains a progressive and fatal disease with 1-year mortality rates as high as 10% to 20%.^{9, 10, 11, 12} Hospitalization is associated with a particularly poor prognosis; patients with PAH and right heart failure have in-hospital mortality rates from 14% to as high as 30% to 48% if they are admitted to an ICU.¹³ Mortality also remains high after hospitalization. In patients who survive hospitalization, 35% die within 1 year of hospital discharge.¹⁴

Palliative care is a medical specialty of interdisciplinary clinicians aimed at improving QOL for patients with a serious or life-limiting illness, or both, as well as support for patients and families. Specialty palliative care services (PCS) are present in the vast majority of hospitals with more than 300 beds, and PCS use for patients who are facing serious illness with potentially life-limiting prognoses increasingly is becoming the standard of care.^15,16 Experts have advocated for earlier integration of PCS in the management of PAH, but small survey-based studies have found that PCS are underrecognized and underused.^{17, 18, 19}

A paucity of studies have evaluated the role of palliative care in patients with PAH, particularly those who are admitted to the hospital.^17,19,20 Therefore, we sought to investigate the national use of PCS by hospital admissions with PAH in the United States. We evaluated the temporal trends of PCS use, predictors of PCS use, and outcomes associated with the use of PCS in hospital admissions for PAH.

Methods

Study Population, Variables, and Outcomes

The National (Nationwide) Inpatient Sample (NIS) is the largest all-payer database of hospital inpatient stays in the United States. The NIS contains discharge data from a 20% stratified sample of hospitals and is a part of the Healthcare Quality and Utilization Project (HCUP), sponsored by the Agency for Healthcare Research and Quality.²¹ Information regarding each discharge includes patient demographics, primary payer, hospital characteristics (region; rural, urban nonteaching, or urban teaching; and small, medium, or large), principal diagnosis, up to 24 secondary diagnoses, and procedural diagnoses. The HCUP and NIS do not evaluate individual patients, but rather evaluate a single inpatient admission. Because the NIS is publicly available deidentified database, institutional review board approval was waived.

Using the NIS data from 2001 through 2017, a retrospective cohort of hospital admissions (≥ 18 years of age) with a primary discharge diagnosis of PAH (International Classification of Diseases [ICD], 9th Edition, Clinical Modification, code 416.0; ICD, 10th Edition, Clinical Modification, code I27.0) were identified.²² PCS use was identified using ICD, 9th Edition, Clinical Modification, code V66.7 or ICD, 10th Edition, Clinical Modification, code Z51.5, consistent with prior literature.^{23, 24, 25} The ICD, 9th Edition, Clinical Modification, code for palliative care encounter (V66.7) showed moderate sensitivity and high specificity (> 90%) in prior validation studies.^26,27 Prior validation studies demonstrated that when ICD, 9th Edition, Clinical Modification, code V66.7 is documented, > 90% patients receive specialist PCS consultation during an inpatient hospitalization.^26,27 This methodology was used previously in studies from the NIS database. Demographic characteristics, hospital characteristics, primary payer, acute organ failure, and cardiac and noncardiac procedures associated with each discharge were identified using previously validated methodology by our group (e-Table 1).^{25,28, 29, 30, 31, 32, 33, 34} The Deyo’s modification of the Charlson Comorbidity Index was used to identify the burden of comorbid diseases.³⁵ In accordance with HCUP definitions, we used median household income of the patient’s zip code of residence as a proxy measure of a patient’s socioeconomic status (SES).³⁶

The primary outcome was the use of PCS in admissions with PAH during this 17-year period. Secondary outcomes included temporal trends of PCS use; predictors of PCS use; and association of PCS use with do-not-resuscitate (DNR) status, length of stay, costs, in-hospital mortality, and discharge disposition. Costs were adjusted for inflation.

Statistical Analysis

As recommended by the HCUP and NIS, survey procedures using discharge weights provided with the NIS database were used to generate national estimates. Using the trend weights provided by the HCUP and NIS, samples from 2001 through 2011 were reweighted to adjust for the 2012 HCUP and NIS redesign.³⁷ To compare categorical and continuous variables, χ ² and t tests, respectively, were used. Logistic regression was used to analyze trends in PCS use over time (referent year, 2001). Univariate analysis for trends and outcomes was performed and outcomes were represented as OR with 95% CI. To account for differences in available PAH therapies over time, we also analyzed outcomes stratified by decade (2001-2010 and 2010-2017) (e-Tables 1-3). Multivariate logistic regression analysis incorporating age, sex, race, primary payer status, SES, hospital characteristics, comorbidities, acuity of illness, cardiogenic shock, acute organ failure, and cardiac and noncardiac procedures was performed for predictors of PCS use. For the multivariate modeling, regression analysis with purposeful selection of statistically (liberal threshold, P < .20 in univariate analysis) and clinically relevant variables was conducted.

The inherent restrictions of the NIS database related to research design, data interpretation, and data analysis were reviewed and addressed.³⁷ Pertinent considerations include not assessing individual hospital-level volumes (because of the changes to sampling design detailed previously), treating each entry as an admission as opposed to individual patients, restricting the study details to inpatient factors because the NIS database does not include outpatient data, and limiting administrative codes to those previously validated and used for similar studies. Two-tailed P values of < .05 were considered statistically significant. All statistical analyses were performed using SPSS version 25.0 software (IBM Corp.).

Results

In the period between January 1, 2001, and December 31, 2017, an estimated 30,495 hospital admissions with a primary diagnosis of PAH. PCS use was noted in 682 (2.2%) of all admissions. In the period from 2001 through 2017, a steady increase in PCS use occurred from 0.5% in 2001 to 7.6% in 2017 (Fig 1A). In a multivariate logistic regression analysis incorporating age, sex, race, primary payer status, SES, hospital characteristics, comorbidities, acuity of illness, cardiogenic shock, acute organ failure, and cardiac and noncardiac procedures, a nearly five-fold increase in PCS use occurred in 2017 (OR, 5.27; 95% CI, 2.94-9.44; P < .001) compared with 2001 (Fig 1B). The use of PCS stratified by patient and hospital characteristics is presented in Figures 2 and 3. A steady increase in PCS referrals was found, with a notable inflection point around 2009 across most patient and hospital characteristics.

Figure 1.

Graphs showing temporal trends in PCS referrals in pulmonary arterial hypertension (PAH) hospital admissions. A, Unadjusted temporal trends of PCS use in PAH hospital admissions between 2001 and 2017 (P < .001 for trend over time). B, Adjusted ORs (with 95% CIs) for PCS use in all PAH hospital admissions between 2001 and 2017 (with 2001 as the referent). Adjustment for age, sex, race, primary payer status, socioeconomic status, hospital characteristics, comorbidities, acuity of illness, cardiogenic shock, acute organ failure, and cardiac and noncardiac procedures (P < .001 for trend over time). PCS = palliative care services.

Figure 2.

A-D, Graphs showing temporal trends in PCS referrals in hospital admissions with pulmonary arterial hypertension (PAH) stratified by patient characteristics. Unadjusted temporal trends of PCS use in hospital admissions with PAH between 2001 and 2017 stratified by age (A), sex (B), race (C), and Charlson Comorbidity Index (D) (P < .001 for trend over time). See Figure 1 legend for expansion of abbreviation.

Figure 3.

A-C, Graphs showing temporal trends in PCS referrals hospital admissions with pulmonary arterial hypertension (PAH) stratified by hospital characteristics. Unadjusted temporal trends of PCS use in PAH admissions between 2001 and 2017 stratified by hospital location and teaching status (A), bed size (B), and region (C) (P < .001 for trend over time). See Figure 1 legend for expansion of abbreviation.

Compared with those not receiving PCS, the cohort receiving PCS was older and more likely to be White, and significant differences were found in primary payer and SES. Patients receiving PCS demonstrated a higher comorbidity burden and were more likely to be admitted during the weekend to medium and large urban teaching hospitals with geographic variation (all P < .001) (Table 1).

Table 1.

Baseline and In-Hospital Characteristics of PAH Admissions Stratified by PCS Use

Characteristic	PCS Use (n = 682)	No PCS Use (n = 29,813)	P Value
Age, y	62.2 ± 16.2	55.9 ± 17.4	< .001
Female sex	77.1	76.6	.78
Race			< .001
White	74.1	65.1
Black	17.9	17.8
Othersa	7.9	17.1
Primary payer			< .001
Medicare	55.5	45.9
Medicaid	10.7	16.5
Private	31.6	31.3
Othersb	2.2	6.3
Weekend admission	21.6	14.3	< .001
Quartile of median household income for zip code			< .001
0-25th	16.5	23.7
26th-50th	30.3	24.8
51st-75th	28.9	24.7
75th-100th	24.3	26.8
Charlson Comorbidity Index			< .001
0-3	33.1	58.3
4-6	42.8	31.9
≥ 7	24.0	9.8
Hospital teaching status and location			< .001
Rural	2.1	7.0
Urban nonteaching	23.5	22.4
Urban teaching	74.4	70.6
Hospital size by no. of beds			< .001
Small	7.9	12.5
Medium	18.6	14.4
Large	73.4	73.1
Hospital region			< .001
Northeast	15.1	23.3
Midwest	27.5	19.5
South	29.1	32.7
West	28.3	24.5
Heart failure	9.5	3.7	< .001
Cardiogenic shock	10.1	1.4	< .001
Acute pulmonary embolism	2.2	1.8	.38
Acute noncardiac organ failure
Respiratory	39.3	11.0	< .001
Renal	40.1	13.1	< .001
Hepatic	1.5	1.1	.36
Hematologic	21.6	9.1	< .001
Neurologic	12.5	1.1	< .001
Coronary angiography	28.9	20.8	< .001
Percutaneous coronary intervention	0.0	1.2	< .001
Right heart catheterization	24.5	29.1	.01
Pulmonary artery catheterization (continuous monitoring)	9.9	10.9	.40
Extracorporeal membrane oxygenation	1.5	0.3	< .001
Fiber-optic bronchoscopy	2.9	2.4	.38
Invasive mechanical ventilation	9.5	4.0	< .001
Noninvasive mechanical ventilation	16.7	3.7	< .001
Acute hemodialysis	2.3	0.8	< .001

Data are presented as percentage or mean ± SD unless otherwise indicated. PAH = pulmonary arterial hypertension; PCS = palliative care services.

^aHispanic, Asian, Native American, others.

^bUninsured, no charge, others.

The PCS cohort demonstrated higher acuity of illness as noted by higher rates of heart failure; cardiogenic shock; acute noncardiac organ failure; and use of cardiac, respiratory, and renal support systems (Table 1). Multivariate predictors of PCS use are detailed in Table 2. In a multivariate logistic regression analysis, White race, private insurance, higher SES, admission to an urban hospital, small hospital by number of beds, hospital region (Midwest, South, and West), higher comorbidity burden (Charlson Comorbidity Index), heart failure and cardiogenic shock, acute noncardiac organ failure, and use of extracorporeal membrane oxygenation and noninvasive mechanical ventilation independently were predictive of increased PCS use, whereas right heart catheterization, invasive mechanical ventilation, and bronchoscopy were associated with decreased PCS use (Table 2).

Table 2.

Multivariate Predictors of PCS Use in Admissions With PAH

Predictor (Total N = 30,495)	OR	95% CI		P Value
		Lower Limit	Upper Limit
Age group, y
18-49	Reference category
50-59	0.68	0.51	0.90	.01
60-69	1.23	0.92	1.64	.16
70-79	0.86	0.61	1.22	.41
≥80	0.91	0.62	1.34	.65
Female sex	1.20	0.99	1.46	.06
Race
White	Reference category
Black	1.09	0.87	1.38	.45
Othersa	0.40	0.32	0.50	< .001
Primary payer
Medicare	Reference category
Medicaid	1.16	0.85	1.58	.36
Private	1.43	1.15	1.79	.002
Othersb	0.54	0.31	0.95	.03
Quartile of median household income for zip code
0-25th	Reference category
26th-50th	1.92	1.49	2.48	< .001
51st-75th	1.74	1.35	2.26	< .001
75th-100th	1.21	0.92	1.59	.17
Hospital teaching status and location
Rural	Reference category
Urban nonteaching	6.33	3.14	12.77	< .001
Urban teaching	5.74	2.87	11.46	< .001
Hospital size by no. of beds
Small	Reference category
Medium	0.50	0.37	0.68	< .001
Large	0.65	0.51	0.84	.001
Hospital region
Northeast	Reference category
Midwest	2.55	1.95	3.33	< .001
South	1.58	1.21	2.05	.001
West	2.35	1.80	3.07	< .001
Charlson Comorbidity Index
0-3	Reference category
4-6	2.08	1.63	2.66	< .001
≥ 7	3.91	2.90	5.27	< .001
Heart failure	1.84	1.36	2.49	< .001
Cardiogenic shock	2.06	1.46	2.91	< .001
Acute organ failure
Respiratory	2.98	2.47	3.58	< .001
Renal	2.07	1.71	2.49	< .001
Hepatic	0.47	0.24	0.95	.04
Hematologic	2.27	1.83	2.80	< .001
Neurologic	5.56	4.04	7.65	< .001
Right heart catheterization	0.78	0.64	0.95	.01
Pulmonary artery catheterization (continuous monitoring)	0.86	0.65	1.14	.29
Extracorporeal membrane oxygenation	3.61	1.51	8.64	.004
Invasive mechanical ventilation	0.66	0.46	0.93	.02
Noninvasive mechanical ventilation	2.78	2.17	3.56	< .001
Fiber-optic bronchoscopy	0.46	0.25	0.85	.01
Acute hemodialysis	0.73	0.41	1.30	.29

See Table 1 legend for expansion of abbreviations.

^aHispanic, Asian, Native American, others.

^bUninsured, no charge, others.

The cohort with PCS consultation showed greater use of DNR status (46.2% vs 1.8%), longer length of stay (12.9 ± 20.4 days vs 7.2 ± 9.8 days), higher hospitalization costs ($130,434 ± $172,417 vs $56,499 ± $144,316), and higher in-hospital mortality (52.8% vs 6.4%; all P < .001). PCS consultation remained an independent predictor of mortality; other significant predictors were age, race, payer, hospital characteristics (size and region), patient comorbidities, and use of mechanical circulatory and noncardiac organ support (e-Table 2). The PCS cohort had fewer discharges to home (25% vs 68.9%) and more frequent discharges to skilled nursing facilities (27.9% vs 7.7%; all P < .001). PCS use was associated independently with greater use of DNR status and fewer discharges to home even after adjusting for age, race, primary payer status, SES, hospital characteristics, comorbidities, acuity of illness, cardiogenic shock, out-of-hospital cardiac arrest, acute organ failure, and cardiac and noncardiac procedures (adjusted OR, 25.49 [95% CI, 20.63-31.48; P < .001] and 0.12 [95% CI, 0.09-0.15; P < .001], respectively). When stratified by decades (2001-2010 and 2010-2017), results overall were similar. The cohorts with PCS consultation had higher prevalence of DNR status, longer length of stay, increased hospital costs, and increased in-hospital mortality (e-Table 3). Notably, a higher prevalence of DNR status, decreased length of stay, higher costs, and increased discharges to nursing homes was found in more recent years (e-Table 3).

Discussion

In this nationally representative descriptive study evaluating PCS use in PAH, we noted very limited use of PCS in only 2% of hospital admissions. A serial increase in PCS use was found during the study period across all patient and hospital categories, but the overall rate remained low over time. White race, higher comorbidity burden, higher SES, private insurance, admission to an urban hospital, small hospital by number of beds, geographic region (Midwest, South, and West), heart failure and cardiogenic shock, acute noncardiac organ failure, and use of extracorporeal membrane oxygenation and noncardiac organ support were independent predictors of PCS use. PCS use was associated with a higher prevalence of DNR status and identified a higher risk cohort with greater in-hospital resource use, higher in-hospital mortality, and less frequent discharges to home, suggesting delayed PCS consultation and higher rates of therapeutic failure in this population.

Despite the poor prognosis and significant symptom burden associated with PAH, our study found a low rate of PCS use in PAH hospital admissions. PAH is a progressive disease associated with impaired QOL and survival.^6,7,38,39 Experts have advocated for early implementation of PCS to improve QOL in PAH, but data on use of palliative care in PAH are very scarce.^17,19,20 Prior survey-based studies have provided some insight into PCS use in PAH, but to our knowledge, this is the first study to characterize use of PCS in PAH during hospital admissions. A survey of PAH specialists showed low PCS referral (40%) for impaired QOL and symptoms, whereas another survey of PAH patients described a high prevalence of impaired QOL, but low patient awareness and use of PCS.^17,19 PAH hospitalizations are increasing, and patients admitted have a higher comorbidity burden with more advanced right heart failure and often die in the hospital setting.^20,22,40 Although the ideal timing of PCS consultation is not known, PCS use during hospitalizations can be helpful to address issues related to symptoms, QOL, and advanced care planning, particularly if these issues have not been addressed in the outpatient setting. Given the high morbidity and mortality associated with PAH even after hospitalization, hospital admissions without PCS use represent a missed opportunity.¹⁴

The low inpatient use of PCS of < 10% of PAH admissions to hospital is similar to that found in studies of patients with other cardiopulmonary conditions, such as acute myocardial infarction (4.5%), heart failure (1.2%), and COPD (1.7%).^25,41,42 Concordant with prior studies, an increase in PCS use occurred with time, even after adjusting for baseline demographics and comorbidities.^25,42 This may be related to an increased awareness of the benefit of PCS in other diseases such as heart failure and cancer or increased availability of PCS.^43,44 Interestingly, a notable inflection point occurred in 2009. Although the exact reason for this observation is not known, it may be related to recognition of palliative care and hospice as a medical subspecialty with board certification in 2008 or identification of palliative care by the National Priorities Partnership as one of six priority areas in 2008.^45,46

Discrepancies in PCS use were present and varied according to type of hospital practice, geographic region, insurance status, comorbidity burden, SES, and race. Similar disparities have been observed in prior studies of heart failure and cardiogenic shock resulting from myocardial infarction.^25,42,47 We identified specific groups with lower rates of PCS use. Patients with worse SES and “other” race, for example, showed lower rates of PCS use. These groups may benefit from advocacy efforts to increase access to PCS. In contrast, patients with private insurance showed higher rates of PCS use. Potential reasons for these disparities cannot be elucidated from this study, but may include differential access to PCS, unconscious biases related to usefulness of PCS, and misconception of PCS as end-of-life care.^25,42 Further studies to explore reasons for these disparities in PCS use are needed to minimize inequalities and to ensure uniform access to PCS.

We suspect that some differences in PCS use, such as the lower use in rural vs urban hospitals, is related to availability of PCS services. The Pulmonary Hypertension Association recently encouraged accredited Pulmonary Hypertension Care Centers to incorporate PCS into clinical practice, emphasizing its importance in PAH management. Access to expert quality care and specialty services such as PCS are an additional reason for patients with PAH to be managed at an accredited Pulmonary Hypertension Care Center.

Patients who received PCS had a higher acuity of illness; they were more likely to have heart failure and cardiogenic shock and showed a higher use of extracorporeal membrane oxygenation. These findings highlight selective and late use of PCS, which also has been described in other conditions.^25,42,47 Despite multiple studies highlighting the critical role of palliative care in patients with chronic medical conditions and higher benefit with early involvement,^44,48 palliative care consultation often is considered synonymous with end-of-life cares and is often underused.⁴⁹ In contrast, some invasive procedures associated with more aggressive care, such as right heart catheterization and invasive ventilation, were associated with lower PCS use, whereas less aggressive measures, such as noninvasive ventilation, were associated with higher PCS use. We suspect that these differences may be reflective of differences in goals of care and potentially may be related to differences in preferences associated with intubation.

PCS use was associated with a higher prevalence of DNR status, a longer length of stay, higher hospitalization costs, and increased likelihood of being discharged to a skilled nursing facility. Studies of heart failure have described a similar association between PCS use and DNR status.^25,47 Palliative care specialists help patients establish goals of care aligned with their wishes through advanced communication skills, and this may account for differences in DNR status.⁵⁰ Similar to some studies in heart failure and cardiogenic shock using the same database, but contrary to others, we noted a longer length of stay and higher hospitalization costs in patients with PCS use.^25,42,47,51 This is likely because the patients who received PCS were sicker (higher comorbidity index and increased need for advanced organ support) and older than those who did not.

Patients who received PCS showed higher mortality, which is most likely multifactorial and related to higher illness acuity and comorbidity burden, higher prevalence of DNR status, and other unmeasured confounding factors. In multivariate analysis, the association between PCS use and mortality persisted after adjusting for comorbidity burden, organ failure, and need for advanced organ support, which suggests that unmeasured factors such as frailty and treatment failure confound the relationship between PCS and mortality. Because of limitations of the database, it is not possible to assess the timing of PCS use, so it is not known whether PCS was used late by patients with prolonged hospitalizations. Given the high mortality rates associated with PCS use, it is clear that PCS was used more at the end of life and less in patients who survived hospitalization. This is consistent with a prior survey of PAH specialists in which providers stated that they used PCS primarily for patients who were actively dying at the end of life.¹⁷ Although it is possible that patients would benefit from early PCS in the outpatient setting or during a hospitalization to improve symptoms and QOL and to decrease overall health-care costs, these outcomes cannot be assessed within this database.

In summary, to our knowledge, our study is the first to evaluate the use of PCS in patients admitted with PAH and reports a low overall rate of use. A significant increase in PCS use occurred from 2001 through 2017, which is encouraging. Although the appropriate timing for PCS consultation in PAH is unclear because of a paucity of data, experts recommend early PCS use to improve symptoms and QOL and to address unmet needs of PAH patients.^52,53 To implement these recommendations, it is clear that improved and systematic PCS use in patients with PAH is needed.

Study Limitations

This study has several limitations, some of which are inherent to the analysis of a large administrative database. Because the NIS relies on discharge diagnosis for identifying patients, errors in coding could have led to some hospitalizations to be missed and possibly other causes of pulmonary hypertension to be included, particularly in early years of the study because of a lack of PAH awareness. Data on medication and PAH cause are not available in this database, so the effect of these on PCS use and patient outcomes could not be evaluated. The HCUP attempts to mitigate potential errors in the NIS by using internal and external quality control measures. Important factors such as terminal extubation, QOL, and mortality after hospitalization could not be identified reliably in this hospital admissions database. Also, the reasons for obtaining a PCS referral, the role of PCS on symptom mitigation, and improvement in QOL and relationships could not be assessed in this administrative database. The timing of PCS consultation and whether patients already had undergone a consultation with palliative care before admission is unclear and also cannot be ascertained through this database. It is possible that these patients could have undergone PCS previously in the outpatient setting or during a different inpatient encounter that was not captured on the NIS admission. Finally, ICD codes, which were used for inclusion, may not be specific for the diagnosis of PAH. Despite these limitations, this study addressed an important knowledge gap highlighting the national use of PCS in PAH.

Conclusions

In this large, nationally representative contemporary cohort study, use of PCS was documented in only 2% of all admissions for PAH despite the high morbidity and mortality associated with this condition. Despite increasing trends in the adoption of PCS, significant patient- and hospital-specific disparities in implementation remain. Further dedicated studies evaluating these disparities and the integration of PCS into the comprehensive care of PAH patients are warranted.