Helth Insurnce nd Rcil Disprities in Pulmonry Hypertension Outcomes Kishn S. Prikh, MD; Kthryn A. Stckhouse, MD; Stephen A. Hrt, MD; Thoms M. Bshore, MD; nd Richrd A. Krsuski, MD Pulmonry hypertension (defined s men pulmonry rtery pressure 25 mm Hg) encompsses rod spectrum of diseses nd ptient popultions. Prior efforts to chrcterize ptients with pulmonry hypertension focused on the suset of those with pure pulmonry vsculr disese, defined s pulmonry rteril hypertension (pulmonry cpillry wedge pressure 15 mm Hg nd pulmonry vsculr resistnce 3 Woods units). Given the disml 5-yer survivl rte of 34% in the Ntionl Institutes of Helth pulmonry rteril hypertension cohort 3 decdes go, 1 therpeutic dvnces trgeting functionl cpcity, disese progression, nd mortlity hve led to the vilility of endothelin receptor ntgonists (ERAs), 2-7 prostcyclins or their nlogs, 8-10 phosphodiesterse 5-inhiitors, 11-13 nd solule gunylte cyclse stimultors. 14,15 These dvnces in disese mngement hve improved the cre of ptients with pulmonry rteril hypertension, ut re expensive: the pproximte nnul costs re sildenfil, $12,761; osentn, $55,890; mrisentn, $56,736; iloprost, $92,146; epoprostenol (70 kg ptient), $33,153; nd treprostinil (70 kg ptient), $97,615. 16 Furthermore, 5-yer survivl, incorporting dvnces in modern therpies, remins low t 58% for ptients with pulmonry rteril hypertension, 17 nd the development of pulmonry hypertension is mong the most significnt predictors of poor outcomes in ptients with chronic hert filure nd lung disese. Alhough there is similr prevlence of pulmonry hypertension in lck (6.6%) nd white popultions (6.8%), 18,19 Blck individuls hve een shown to hve worse prognosis. 20-23 Furthermore, ge-stndrdized deth rtes for pulmonry rteril hypertension ptients from 2 seprte epidemiologicl studies, from 1980 to 1984 nd 2000 to 2002, hve diverged etween whites (incresing from 5.0%-5.3%) nd lcks (incresing from 4.8%-7.3%), 21 with little insight into why these rcil disprities currently exist. However, the Registry to Evlute Erly nd Long-term Pulmonry Arteril Hypertension Disese Mngement (REVEAL) registry of ptients with pulmonry rteril hypertension in the United Sttes reported no difference in time-to-recognition of disese from symptom ABSTRACT OBJECTIVES: Pulmonry hypertension portends poorer prognosis for lcks versus white popultions, ut the underlying resons re poorly understood. We investigted ssocitions of disese chrcteristics, insurnce sttus, nd rce with clinicl outcomes. STUDY DESIGN: Retrospective cohort study of ptients presenting for initil pulmonry hypertension evlution t 2 cdemic referrl centers. METHODS: We recorded insurnce sttus (Medicre, Medicid, privte, self-py), echocrdiogrphic, nd hemodynmics dt from 261 ptients (79% whites, 17% lcks) with new dignosis of pulmonry hypertension. Sujects were followed for 2.3 yers for survivl. Adjustment for covrites ws performed with Cox proportionl hzrds modeling. RESULTS: Compred with white ptients, lcks were younger (50 ± 15 vs 53 ± 12 yers; P =.04), with femles representing mjority of ptients in oth groups (80% vs 66%; P =.08) nd similr functionl clss distriution (clss 2/3/4: 30%/52%/16% lcks vs 33%/48%/14% whites; P =.69). Blcks dignosed with incident pulmonry hypertension were more frequently covered y Medicid (12.5% vs 0.7%) nd hd less privte insurnce (50% vs 61%; P =.007) thn whites. At presenttion, lcks hd more right ventriculr dysfunction (P =.04), ut similr men pulmonry rteril pressure (46 vs 45 mm Hg, respectively; P =.66). After djusting for ge nd functionl clss, lcks hd greter mortlity risk (hzrd rtio [HR], 2.06; 95% confidence intervl [CI], 1.18-3.44), which did not differ y rce fter dditionl djustment for insurnce sttus (HR, 1.74; 95% CI, 0.84-3.32; P =.13). CONCLUSIONS: In lrge cohort of ptients with incident pulmonry hypertension, lck ptients hd poorer rightside hert function nd survivl rtes thn white ptients. However, djustment for insurnce sttus in our cohort removed differences in survivl y rce. Am J Mng Cre. 2017;23(8):474-480 474 AUGUST 2017 www.jmc.com
Insurnce nd Rce in Pulmonry Hypertension onset y rce. 24 We sought to investigte the ssocition of rce with survivl in popultion of ptients referred to 2 lrge tertiry cre centers for invsive evlution of pulmonry hypertension. Additionlly, given tht cost my e rrier to pulmonry rteril hypertension medictions, ssocited equipment for intrvenous/inhled dministrtion, nd specilized cre, we investigted the role of insurnce sttus on mortlity. TAKEAWAY POINTS Blck rce ws ssocited with 2-fold mortlity risk in ptients with incident pulmonry hypertension fter djustment for ge nd functionl clss. However, this reltionship ws no longer oserved fter dditionl djustment for insurnce. A totl of 261 consecutive ptients with pulmonry hypertension were included in our cohort t time of first evlution t 2 lrge tertiry referrl centers (Duke University Medicl Center nd Clevelnd Clinic). Africn Americn ptients hd more severe pulmonry vsculr disese t presenttion nd greter thn 2-fold risk of deth over study follow-up. Pulmonry hypertension is ftl disese nd ccess to medictions is pulic helth issue due to high costs nd rcil disprities. METHODS Consecutive ptients who were evluted t 2 pulmonry hypertension referrl centers for initil hemodynmic ssessment etween 1998 nd 2009, nd found to hve men pulmonry rtery pressure of t lest 25 mm Hg, were prospectively entered in dtse. Ech ptient underwent dignostic work-up, including echocrdiogrphy nd crdic ctheteriztion with vsodiltor testing, if indicted. Only ptients with known insurnce type were included in our study. The study ws pproved y the institutionl review ords t Duke University Medicl Center nd the Clevelnd Clinic Foundtion. Right hert ctheteriztion ws performed y the sme experienced crdiologist (RAK) in the crdic ctheteriztion lortories t the 2 medicl centers, using single end-hole, lloon flottion ctheter (either Brd Pulmonry Wedge Ctheter [Medtronic; Minnepolis, Minnesot] or Blloon Wedge Pressure Ctheter [Arrow Interntionl, Inc; Clevelnd, Ohio]). Contrst injections during left hert ctheteriztion, if performed, followed ll hemodynmic mesurements. Stndrd hemodynmic mesurements from right hert ctheteriztion were otined, including men right tril pressure, right ventriculr, systolic nd distolic pressures, pulmonry rtery systolic, distolic nd men pressures, nd men pulmonry cpillry wedge pressure. Men rteril lood pressure, crdic index, pulmonry vsculr resistnce, systemic vsculr resistnce, nd response to vsodiltor chllenge were lso recorded in the dtse when indicted. All echocrdiogrms were performed within 30 dys of crdic ctheteriztion using phse-rryed scnner with 4- or 8-MHz trnsducer, depending on ody hitus nd imge qulity. Stndrd 2-dimensionl imges were otined in the prsternl long- nd short-xis views, picl 2- nd 4-chmer views, nd sucostl view, nd Doppler grdients were otined to estimte vlvulr grdients. Assessment of right ventriculr size ws reported s grde from 0 to 3, where 0 = norml, 1 = mildly enlrged, 2 = modertely enlrged, nd 3 = severely enlrged. Right ventriculr function ws grded on 4-point scle for systolic dysfunction, where 0 = no dysfunction, 1 = mild dysfunction, 2 = moderte dysfunction, nd 3 = severe dysfunction. The degree of tricuspid regurgittion ws determined using color-flow Doppler nd ssigned grde from 0 to 4+, depending on the extent of color flow reltive to the right tril re. Estimted right ventriculr pressures were otined y pplying continuous wve Doppler to the tricuspid regurgittion jet to the Bernoulli eqution nd dded to the estimted right tril pressure. If tricuspid regurgittion velocity could not e dequtely visulized y color Doppler, sline microules were injected to improve imge qulity. Sline microule injection ws lso used to ssess for intrcrdic shunting if flow cross the intertril septum ws noted y color Doppler on sucostl imging. Assessments were mde y experienced imging crdiologists t oth cdemic medicl centers. Pulmonry hypertension work-up for ll ptients lso included chest rdiogrphy, (chest computed tomogrphy if chest rdiogrphy ws norml), ventiltion-perfusion scnning, full pulmonry function testing, electrocrdiogrphy, nd echocrdiogrphy, in ddition to ctheteriztion. Test results, demogrphics, nd medicl histories were otined from medicl records nd entered in the dtse. Insurnce sttus t the time of the ssessment ws recorded s either Medicre, Medicid, privte, or self-py, nd confirmed y ppointment schedulers t the time of the initil ppointment. Rce ws self-reported t the time of ctheteriztion. Long-term survivl ws ssessed using electronic helth records nd confirmed using the Socil Security Deth Index. Vriles of interest included the following: idiopthic pulmonry rteril hypertension (no ttriutle etiology of pulmonry hypertension), New York Hert Assocition (NYHA) clss (functionl clss 1: no symptoms nd clss 4: severe symptoms), systemic hypertension nd tril firilltion (y medicl history), serum sodium nd cretinine (venous lortory vlues drwn t time of hert ctheteriztion), nd rteril oxygen sturtion (otined y femorl rteril lood smple t time of hert ctheteriztion). Only white nd lck ptients were included in our study; ptients who self-identified s other rces were excluded. Descriptive sttistics were presented s men ± stndrd devition (SD) for continuous vriles nd s percentge for discrete vriles. Comprison of ctegoricl vriles ws performed using THE AMERICAN JOURNAL OF MANAGED CARE VOL. 23, NO. 8 475
TABLE 1. Demogrphic nd Clinicl Chrcteristics Demogrphics Whites (n = 205) Blcks (n = 45) P Medin ge, yers 57.7 ± 15.4 52.6 ± 12.3.02 Femle, % 65.9 80.0.08 Idiopthic PAH, % 41.9 28.6.12 Femle.003 WHO group clssifiction 1, % 59.5 40.0 2, % 11.7 17.8 3, % 19.5 15.6 4, % 3.9 4.4 5, % 5.3 22.2 NYHA clss.69 2, % 29.7 33.3 3, % 52.0 47.6 4, % 16.3 14.3 Systemic hypertension, % 39.0 42.2.74 Dietes, % 18.5 35.6.02 Atril firilltion, % 21.0 11.1.15 Serum sodium (mg/dl) 139.1 ± 3.2 138.4 ± 3.5.21 Serum cretinine (mg/dl) 1.0 ± 0.4 1.0 ± 0.5.69 Arteril O 2 sturtion 89.6 ± 14.7 89.0 ± 8.4.85 Insurnce sttus.007 Medicid, % 0.74 12.5 Medicre, % 31.2 33.3 Privte, % 61.6 50.0 Self-py, % 7.4 4.2 NYHA indictes New York Hert Assocition; O 2, oxygen; PH, pulmonry rteril hypertension; WHO, World Helth Orgniztion. Dt presented s men ± stndrd devition or frequency (%). Functionl clsses rnge from 1 to 4, with 4 hving the most severe symptoms. the χ 2 test or Fisher s exct test, where pproprite. Comprisons of continuous vriles etween groups were performed using 2-sided t tests nd 1-wy nlysis of vrince. Sttisticl significnce ws ssumed with P <.05. Survivl nlyses were performed using the Kpln-Meier nd Cox proportionl hzrds regression methods. Two proportionl hzrds models were utilized to investigte the effect of demogrphics, hemodynmics, echocrdiogrphic mesurements, nd insurnce sttus. Ech model ws constructed in forwrd step-wise mnner, investigting the effect of ech covrite nd potentil interctions individully. Potentil covrites included in the djustment models were vriles with seline differences hving P <.20, including known predictors of deth in the pulmonry hypertension popultion (including right tril pressure, pulmonry rtery pressure, pulmonry vsculr resistnce, right ventriculr function, crdic index, functionl sttus), nd rce. Bsed on the size of the cohort, the mximum numer of covrites included in ech model ws predetermined to e 5 to limit overfitting. Differences in the survivl functions were ssessed for significnce using the Wilcoxon test. All nlyses were performed using JMP version 12.0 softwre (SAS Institute; Cry, North Crolin). RESULTS A totl of 250 ptients (82% white, 18% lck) were included in the study. An dditionl 11 ptients self-identified s Asin/Pcific Islnder, Hispnic, or Ntive Americn were excluded from the nlysis. The medin follow-up period ws 2.3 yers (whites: 2.0; interqurtile rnge [IQR], 1.1-2.0 yers; lcks: 1.7 yers; IQR, 0.9-1.7; P =.02). Overll, lck ptients were younger, more likely to hve dietes, nd coverge with Medicid versus privte insurnce compred with white ptients (Tle 1). Etiology of pulmonry vsculr disese ws idiopthic pulmonry rteril hypertension in 42% of whites, ut only 29% in lcks. Functionl clss distriution ws similr for oth rces. Comprison of pulmonry hypertension chrcteristics ws lso performed using invsive hemodynmics nd echocrdiogrphy t presenttion. Blck individuls hd higher right tril pressure, decresed right ventriculr function, nd more severe tricuspid regurgittion thn whites. Other mrkers of pulmonry hypertension severity, including pulmonry vsculr resistnce, crdic index, nd right ventriculr size, did not differ y rce (Tle 2). No drug prescriing differences were seen etween lck nd white ptients. During follow-up, 55% of ptients received t lest 1 pulmonry rteril hypertension-specific therpy (ERA, prostnoid or phosphodiesterse 5-inhiitor) nd 16% received comintion pulmonry hypertension-specific therpy. The most commonly utilized phrmcologic gents in descending order of frequency included sildenfil (used in 29.6% of ptients), osentn (16.8%), epoprostenol (15.8%), treprostinil (6.8%), nd iloprost (2.8%). None of these gents were ssocited with improved survivl y univrite nlysis. Differences in seline chrcteristics etween ptients who survived nd those who died were exmined (Tle 3). Older ge, worse functionl clss, dietes, serum sodium nd cretinine vlues, nd insurnce sttus were ssocited with deth y completion of the study follow-up period. Anlysis y rce reveled tht 20 of 45 (44%) lck ptients nd 67 of 205 (33%) white ptients hd died t the end of the 2.3-yer follow-up period (Figure). With djustment for seline differences in ge nd NYHA functionl clss, lcks hd incresed risk for deth (hzrd rtio, 2.06; 95% confidence intervl [CI], 1.18-3.44; P =.012) (Tle 4). However, when insurnce sttus ws dded to the model, lcks nd whites hd no sttisticlly significnt difference in survivl rtes (P =.13). 476 AUGUST 2017 www.jmc.com
Insurnce nd Rce in Pulmonry Hypertension TABLE 2. Hemodynmics nd Echocrdiogrphic Mesurements of Whites nd Africn Americns t Initil PH Evlution Vrile Whites (n = 205) Blcks (n = 45) P Invsive hemodynmics Right tril pressure (mm Hg) 10.3 ± 5.7 13.4 ± 8.6.02 Men PA pressure (mm Hg) 44.7 ± 16.0 45.8 ± 15.2.66 PCW pressure (mm Hg) 13.5 ± 6.9 13.1 ± 7.5.73 Crdic index (L/min/m 2 ) 2.5 ± 0.9 2.5 ± 0.9.79 PVR (Wood units) 8.6 ± 7.0 10.2 ± 7.5.26 Systolic lood pressure (mm Hg) 96.5 ± 15.2 103.6 ± 15.3.03 SVR (Wood units) 22.1 ± 9.7 22.4 ± 6.0.86 Echocrdiogrphy RV size (grde) 1.5 ± 1.2 1.8 ± 1.2.09 RV function (grde) c 1.1 ±.1 1.5 ± 1.2.04 RV systolic pressure (mm Hg) 70.5 ± 25.5 68.9 ± 24.7.73 LV ejection frction, % 56.3 ± 6.7 57.9 ± 5.4.15 LV systolic dysfunction, % 9.3 6.8.69 TR (grde) d 2.0 ± 1.1 2.5 ± 1.2.009 LV indictes left ventriculr; PA, pulmonry rtery; PCW, pulmonry cpillry wedge; PH, pulmonry hypertension; PVR, pulmonry vsculr resistnce; RV, right ventriculr; SVR, systemic vsculr resistnce; TR, tricuspid regurgittion. Dt presented s men ± stndrd devition for continuous vriles nd medin ± interqurtile rnge for ctegoricl vriles. RV size grde: 0 = norml, 1 = mildly enlrged, 2 = modertely enlrged, 3 = severely enlrged. c RV function grde: 0 = norml, 1 = mild dysfunction, 2 = moderte dysfunction, 3 = severe dysfunction. d Tricuspid regurgittion grde: 0 = none, 1 = trivil, 2 = mild, 3 = moderte, 4 = severe. DISCUSSION Among 250 ptients referred to 2 lrge pulmonry hypertension referrl centers, lcks hd worse survivl from time of evlution. However, dditionl djustment for insurnce sttus ttenuted the ssocition of rce with outcomes, suggesting tht insurnce sttus plys n importnt role in pulmonry hypertension outcomes. lcks with pulmonry hypertension hd more mrkers of dvnced pulmonry vsculr disese, decresed right ventriculr function, nd were more likely to hve dietes compred with whites. Ptients who died were lso more likely to hve dietes, which my e mrker for endothelil dysfunction nd the resulting pulmonry venous hypertension. 25 Despite incresed disese severity nd dietes, pulmonry hypertension etiology (including pulmonry hypertension secondry to left-sided hert filure) did not vry etween ptients who survived nd those who died. Finlly, the percentge of our sujects lcking insurnce ws lower thn tht of the generl US popultion. The ssocition of rce nd helth outcomes in pulmonry hypertension, which crries high mortlity s n isolted disese nd when ssocited with ostructive lung disese, pulmonry TABLE 3. Demogrphic nd Clinicl Chrcteristics, Cohorts Seprted y Survivors nd Nonsurvivors t End of Study Follow-up Period Chrcteristic Surviving (n = 163) Nonsurviving (n = 87) P Medin ge, yers 53.3 ± 14.7 63.3 ± 13.3 <.001 Femle 68.7 66.7.74 Rce.13 White 84.7 77.0 Blck 15.3 23.0 Follow-up intervl, yers 2.3 ± 0.2 1.2 ±0.8 <.001 Pulmonry hypertension WHO group clssifiction.33 1 55.2 57.5 2 14.7 9.2 3 16.0 24.1 4 4.3 3.4 5 9.8 5.7 NYHA clss <.001 2 40.0 14.1 3 47.2 58.8 4 10.7 25.9 Systemic hypertension 36.8 44.8.22 Dietes 16.0 32.2.003 Atril firilltion 16.6 24.1.15 Serum sodium (mg/dl) 139.4 ± 2.8 138.1 ± 4.0.006 Serum cretinine (mg/dl) 0.9 ± 0.4 1.2 ± 0.5 <.001 Arteril O 2 sturtion, % 90.0 ± 13.7 88.4 ± 14.0.50 Insurnce sttus, %.004 Medicid 1.9 3.5 Medicre 21.4 47.4 Privte 69.9 42.1 Self-py 6.8 7.0 NYHA indictes New York Hert Assocition; O 2, oxygen; WHO, World Helth Orgniztion. Dt presented s men ± stndrd devition, men (rnge), or frequency (%). Rnging from clss 1 to 4, with 4 reflecting the most severe symptoms. firosis, nd left-sided hert filure, 26-35 hs not een well understood. This is the lrgest study to dte to exmine the ssocition etween insurnce nd rcil disprities in outcomes for pulmonry hypertension, nd the only study to exmine this ssocition mong ptients with pulmonry hypertension in WHO groups 2 to 5. Prior studies exmining rcil differences in pulmonry rteril hypertension hve consistently shown tht lcks hve poorer outcomes, ut the studies hd significnt limittions. Lilienfield et l found cler rcil differences in pulmonry rteril hypertension survivl using dt from the CDC, ut ll dignosis nd disese chrcteriztion ws sed on Interntionl Clssifiction THE AMERICAN JOURNAL OF MANAGED CARE VOL. 23, NO. 8 477
FIGURE. Survivl of Ptients With Newly Dignosed Pulmonry Hypertension y Rce Over 2.3-Yer Follow-up Proportion Surviving Survivors 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 0 0.5 1 1.5 2 2.5 Yers Since Pulmonry Hypertension Referrl 205 191 178 166 157 46 38 34 32 28 Whites Blcks of Diseses, Ninth Revision, Clinicl Modifiction (ICD-9-CM) codes. 23 Furthermore, their nlysis used dt from 1979 to 1996 in the sence of mny new pulmonry rteril hypertension therpies tht hve emerged in the lst 2 decdes, nd re more difficult to interpret in the context of current mngement prctices. A more recent study of pulmonry rteril hypertension prognostictors identified rce s one of the most significnt vriles ssocited with deth, with n djusted survivl hzrd rtio of 4.3 for nonwhites compred with whites. 22 However, further explortion of this strong ssocition ws limited y smll smple size of nonwhite ptients (15 nonwhites, including 6 lcks), nd socioeconomic nd insurnce sttus ws not reported. In the United Sttes, lcks experience more life-yers lost s result of hypertension, HIV, dietes, nd trum, 34 nd they re lso less likely to undergo procedures for hert disese including ngiogrphy nd coronry rtery ypss surgery resulting in poorer outcomes following crdiothorcic surgery. 33,36,37 These findings persist fter controlling for socioeconomic fctors. Thus, rcil disprities in helthcre hve een well estlished, nd our study extends these findings to pulmonry hypertension. 28,29,31,32,34,35 Hypotheses regrding discrepncies for other disese sttes hve cited differences in genetics, income, eduction, ccess to medicl insurnce nd providers, ffordility of medicl cre, geogrphy, nd psychologicl ises mong ptients nd providers. 38-45 In our popultion, lcks hd poorer right ventriculr function, ut overll, no significnt differences in clinicins ssessment of functionl clss compred with whites. Right ventriculr function is known key determinnt of outcome in pulmonry hypertension, nd lck ptients without privte insurnce my hve een referred lter in their disese course, nd presented with more difficult to tret disese. This discrepncy lso suggests tht clinicl scertinment of functionl clss, which is often used to se mngement decisions, my lso e suject to importnt inccurcies, s shown in hert filure popultions. 46 Our results highlight the need for further pulic helth study in understnding how to mnge pulmonry hypertension most effectively. Although time-from-symptom onset to pulmonry rteril hypertension dignosis likely does not differ y rce, 24 the poorer right ventriculr function oserved in lcks on presenttion my TABLE 4. Proportionl Hzrds Regression Models NYHA Functionl Model Age β Vlue Clss β-vlue Model A c 0.04 (P <.001) Model B d 0.03 (P =.006) Clss 1: 0.99 (P =.23) Clss 2: 0.48 (P =.19) Clss 3: 0.37 (P =.25) Overll P =.003 Clss 1: 0.91 (P =.27) Clss 2: 0.54 (P =.16) Clss 3: 0.35 (P =.30) Overll P =.005 Insurnce Sttus β Vlue Survivl Hzrd Rtio (lck vs white) 95% CI P N/A 2.06 1.18-3.44.01 Medicid: 0.97 (P =.24) Medicre: 0.07 (P =.86) Privte: 0.79 (P =.03) Overll P =.05 1.74 0.84-3.32.13 CI indictes confidence intervl; NYHA, New York Hert Assocition. Bet vlue is the stndrd regression coefficient used in the sttisticl model. Rnging from clss 1 to 4, with 4 reflecting the most severe symptoms. NYHA functionl clss β vlue is reltive to NYHA clss 4 nd insurnce sttus β vlue. c Adjusted for ge nd NYHA functionl clss. d Adjusted for ge, NYHA functionl clss, nd insurnce sttus. 478 AUGUST 2017 www.jmc.com
Insurnce nd Rce in Pulmonry Hypertension e due to genetic fctors, comorid conditions, or fewer interctions with helthcre providers, leding to more ggressive disese t presenttion to the pulmonry hypertension referrl center. lcks re less likely to hve privte insurnce, nd we found tht the type of insurnce t the time of referrl my interct with rce for survivl. Insurnce my e surrogte for overll socioeconomic sttus, nd it lso my hve implictions for ese of ccess to referrl centers nd potentil pulmonry hypertension therpies. Limittions The grnulrity of our nlysis is limited y lck of dt on modes of deth. However, distriution of pulmonry hypertension etiology did not significntly differ etween survivor nd nonsurvivor groups, nd so it is less likely tht our findings re driven y single pulmonry hypertension sutype or ssocited comorid conditions. Ptient dherence rtes nd physicin likelihood to prescrie pulmonry rteril hypertension therpies when indicted were not ssessed, ut, given their reltionship with insurnce sttus, would e difficult to interpret in the context of this study. Once the decision to prescrie pulmonry rteril hypertension therpy ws mde, we were le to ssess tht specific pulmonry rteril hypertension therpies did not differ y rce lone. Finlly, we noted wek correltion etween ge nd Medicre sttus in our popultion in our model: due to the discrete informtion provided y ech vrile nd lck of significnt multi co-linerity, we kept oth vriles in the djustment model. Clinicl Implictions Clinicins cring for ptients with pulmonry hypertension should recognize tht insurnce sttus is n importnt prognostictor of incresed mortlity fter djustment for ge nd functionl clss, regrdless of pulmonry hypertension WHO clssifiction. Ptients with pulmonry hypertension who re lcking privte insurnce represent high-risk supopultion tht would enefit from further pulic helth reserch nd thorough evlution of insurnce options if necessry monitoring nd tretments re not ville. CONCLUSIONS Blcks referred for evlution of pulmonry hypertension hve worse right hert function nd susequent prognosis compred with whites fter initil dignosis. However, djusting for insurnce sttus wekens the rcil outcomes disprities. Pulmonry hypertension workup nd mngement often require multiple tests for dignosis nd monitoring, specilist cre, nd, if wrrnted, costly therpies. A etter understnding of vrying options ville to ptients with newly dignosed pulmonry hypertension cross the spectrum of insurnce types my identify pulic helth mesures tht would improve detection nd moridity nd mortlity ssocited with the disese. n Author Affilitions: Duke Clinicl Reserch Institute (KSP, RAK), Durhm, NC; Division of Crdiology, Duke University Medicl Center (KSP, TMB, RAK), Durhm, NC; Clevelnd Clinic Lerner College of Medicine, Cse Western Reserve University (KAS), Clevelnd, OH; University of Michign (SAH), Ann Aror, MI. Source of Funding: Mnuscript preprtion supported y NIH grnt 5T32GM086330-05 (KSP). Author Disclosures: Dr Krsuski hs grnts pending from Actelion nd is consultnt for Actelion nd Byer. The remining uthors report no reltionship or finncil interest with ny entity tht would pose conflict of interest with the suject mtter of this rticle. Authorship Informtion: Concept nd design (TMB, KSP, KAS, SAH, RAK); cquisition of dt (SAH, RAK); nlysis nd interprettion of dt (KSP, KAS, RAK); drfting of the mnuscript (KSP, KAS, RAK); criticl revision of the mnuscript for importnt intellectul content (KSP, KAS, SAH, RAK); sttisticl nlysis (KSP, KAS, RAK); otining funding (RAK); provision of ptients or study mterils (RAK); dministrtive, technicl, or logistic support (TMB, SAH, RAK); nd supervision (TMB). Address Correspondence to: Richrd A. Krsuski, MD, Deprtment of Crdiovsculr Medicine, Duke University Medicl Center, 2301 Erwin Rd, Durhm, NC 27710. E-mil: Richrd.krsuski@duke.edu. REFERENCES 1. Rich S, Dntzker DR, Ayres SM, et l. Primry pulmonry hypertension. ntionl prospective study. Ann Intern Med. 1987;107(2):216-223. 2. Chnnick RN, Simonneu G, Siton O, et l. Effects of the dul endothelin-receptor ntgonist osentn in ptients with pulmonry hypertension: rndomised plceo-controlled study. Lncet. 2001;358(9288):1119-1123. 3. Liu C, Cheng J. Endothelin receptor ntgonists for pulmonry rteril hypertension. Cochrne Dtse Syst Rev. 2005;(1):CD004434. 4. McLughlin VV, Gine SP, Brst RJ, et l; Treprostinil Study Group. Efficcy nd sfety of treprostinil: n epoprostenol nlog for primry pulmonry hypertension. J Crdiovsc Phrmcol. 2003;41(2):293-299. 5. McLughlin VV, Siton O, Bdesch DB, et l. Survivl with first-line osentn in ptients with primry pulmonry hypertension. Eur Respir J. 2005;25(2):244-249. 6. Ruin LJ, Bdesch DB, Brst RJ, et l. Bosentn therpy for pulmonry rteril hypertension. N Engl J Med. 2002;346(12):896-903. 7. Skoro-Sjer N, Lng I. The role of treprostinil in the mngement of pulmonry hypertension. Am J Crdiovsc Drugs. 2008;8(4):213-217. 8. Brst RJ, Ruin LJ, Long WA, et l; Primry Pulmonry Hypertension Study Group. A comprison of continuous intrvenous epoprostenol (prostcyclin) with conventionl therpy for primry pulmonry hypertension. N Engl J Med. 1996;334(5):296-301. 9. McLughlin VV, Genthner DE, Pnell MM, Rich S. Reduction in pulmonry vsculr resistnce with long-term epoprostenol (prostcyclin) therpy in primry pulmonry hypertension. N Engl J Med. 1998;338(5):273-277. 10. Olschewski H, Simonneu G, Glie N, et l; ; Aerosolized Iloprost Rndomized Study Group. Inhled iloprost for severe pulmonry hypertension. N Engl J Med. 2002;347(5):322-329. 11. Benedict N, Seyert A, Mthier MA. Evidence-sed phrmcologic mngement of pulmonry rteril hypertension. Clin Ther. 2007;29(10):2134-2153. 12. Sstry BK, Nrsimhn C, Reddy NK, Rju BS. Clinicl efficcy of sildenfil in primry pulmonry hypertension: rndomized, plceo-controlled, doule-lind, crossover study. J Am Coll Crdiol. 2004;43(7):1149-1153. 13. Singh TP, Rohit M, Grover A, Mlhotr S, Vijyvergiy R. A rndomized, plceo-controlled, doule-lind, crossover study to evlute the efficcy of orl sildenfil therpy in severe pulmonry rtery hypertension. Am Hert J. 2006;151(4):851.e1-e5. 14. Ghofrni HA, D Armini AM, Grimminger F, et l; CHEST-1 Study Group. Riocigut for the tretment of chronic thromoemolic pulmonry hypertension. N Engl J Med. 2013;369(4):319-329. doi: 10.1056/NEJMo1209657. 15. Ghofrni HA, Glie N, Grimminger F, et l; PATENT-1 Study Group. Riocigut for the tretment of pulmonry rteril hypertension. N Engl J Med. 2013;369(4):330-340. doi: 10.1056/NEJMo1209655. 16. McLughlin VV, Archer SL, Bdesch DB, et l; ACCF/AHA. ACCF/AHA 2009 expert consensus document on pulmonry hypertension: report of the Americn College of Crdiology Foundtion Tsk Force on Expert Consensus Documents nd the Americn Hert Assocition: developed in collortion with the Americn College of Chest Physicins, Americn Thorcic Society, Inc., nd the Pulmonry Hypertension Assocition. Circultion. 2009;119(16):2250-2294. doi: 10.1161/CIRCULATIONAHA.109.192230. 17. Thenppn T, Shh SJ, Rich S, Gomerg-Mitlnd M. A USA-sed registry for pulmonry rteril hypertension: 1982-2006. Eur Respir J. 2007;30(6):1103-1110. 18. Choudhry G, Jnkowich M, Wu WC. Prevlence nd clinicl chrcteristics ssocited with pulmonry hypertension in Africn-Americns. PLoS One. 2013;8(12):e84264. doi: 10.1371/journl.pone.0084264. 19. Ene I, Ghio S, Bongrzoni A, et l. Echocrdiogrphic ltertions suggestive of pulmonry hypertension in the Itlin ultrsonogrphy lortories. Epidemiologicl dt from the INCIPIT study (INCidence of Pulmonry Hypertension in Itlin ultrsonogrphy lortories). G Itl Crdiol (Rome). 2010;11(5):402-407. 20. Dvis KK, Lilienfeld DE, Doyle RL. Incresed mortlity in Africn Americns with idiopthic pulmonry rteril hypertension. J Ntl Med Assoc. 2008;100(1):69-72. 21. Hyduk A, Croft JB, Ayl C, Zheng K, Zheng ZJ, Mensh GA. Pulmonry hypertension surveillnce United Sttes, 1980-2002. MMWR Surveill Summ. 2005;54(5):1-28. 22. Kwut SM, Horn EM, Berekshvili KK, et l. New predictors of outcome in idiopthic pulmonry rteril hypertension. Am J Crdiol. 2005;95(2):199-203. THE AMERICAN JOURNAL OF MANAGED CARE VOL. 23, NO. 8 479
23. Lilienfeld DE, Ruin LJ. Mortlity from primry pulmonry hypertension in the United Sttes, 1979-1996. Chest. 2000;117(3):796-800. 24. Brown LM, Chen H, Hlpern S, et l. Dely in recognition of pulmonry rteril hypertension: fctors identified from the REVEAL Registry. Chest. 2011;140(1):19-26.doi: 10.1378/chest.10-1166. 25. Aernethy AD, Stckhouse K, Hrt S, et l. Impct of dietes in ptients with pulmonry hypertension. Pulm Circ. 2015;5(1):117-123. doi: 10.1086/679705. 26. Bursi F, McNlln SM, Redfield MM, et l. Pulmonry pressures nd deth in hert filure: community study. J Am Coll Crdiol. 2012;59(3):222-231. doi: 10.1016/j.jcc.2011.06.076. 27. Seeger W, Adir Y, Brer JA, et l. Pulmonry hypertension in chronic lung diseses. J Am Coll Crdiol. 2013;62(suppl 25):D109-116. doi: 10.1016/j.jcc.2013.10.036. 28. Epstein AM, Aynin JZ. Rcil disprities in medicl cre. N Engl J Med. 2001;344(19):1471-1473. 29. Koroukin SM. Rcil disprities in helth cre. JAMA. 2002;287(22):2942; uthor reply 2943. 30. Lederer DJ, Benn EK, Brr RG, et l. Rcil differences in witing list outcomes in chronic ostructive pulmonry disese. Am J Respir Crit Cre Med. 2008;177(4):450-454. 31. Mensh GA, Mokdd AH, Ford ES, Greenlund KJ, Croft JB. Stte of disprities in crdiovsculr helth in the United Sttes. Circultion. 2005;111(10):1233-1241. 32. Nelson A. Unequl tretment: confronting rcil nd ethnic disprities in helth cre. J Ntl Med Assoc. 2002;94(8):666-668. 33. Rothenerg BM, Person T, Zwnziger J, Mukmel D. Explining disprities in ccess to high-qulity crdic surgeons. Ann Thorc Surg. 2004;78(1):18-24; discussion 24-15. 34. Wong MD, Shpiro MF, Boscrdin WJ, Ettner SL. Contriution of mjor diseses to disprities in mortlity. N Engl J Med. 2002;347(20):1585-1592. 35. Cooper R, Cutler J, Desvigne-Nickens P, et l. Trends nd disprities in coronry hert disese, stroke, nd other crdiovsculr diseses in the United Sttes: findings of the ntionl conference on crdiovsculr disese prevention. Circultion. 2000;102(25):3137-3147. 36. Rcil/ethnic differences in crdic cre: the weight of the evidence. Kiser Fmily Foundtion wesite. https://kiserfmilyfoundtion.files.wordpress.com/2002/09/6040r-rcil-nd-ethnic-differences-in-crdiccre-report.pdf. Pulished Octoer 2002. Accessed Septemer 16, 2016. 37. Lillie-Blnton M, Mddox TM, Rushing O, Mensh GA. Disprities in crdic cre: rising to the chllenge of Helthy People 2010. J Am Coll Crdiol. 2004;44(3):503-508. 38. Gry TL, Nryn KM, Gregg EW, Beckles GL, Sddine JB. Rcil/ethnic differences in the helthcre experience (coverge, utiliztion, nd stisfction) of US dults with dietes. Ethn Dis. 2003;13(1):47-54. 39. Hider AH, Chng DC, Efron DT, Hut ER, Crndll M, Cornwell EE 3rd. Rce nd insurnce sttus s risk fctors for trum mortlity. Arch Surg. 2008;143(10):945-949. doi: 10.1001/rchsurg.143.10.945. 40. Hrgrves JL, Hdley J. The contriution of insurnce coverge nd community resources to reducing rcil/ethnic disprities in ccess to cre. Helth Serv Res. 2003;38(3):809-829. 41. Husmnn LR, Jeong K, Bost JE, Irhim SA. Perceived discrimintion in helth cre nd helth sttus in rcilly diverse smple. Med Cre. 2008;46(9):905-914. doi: 10.1097/MLR.0013e3181792562. 42. Lpu-Bul R, Ofili E. From hypertension to hert filure: role of nitric oxide-medited endothelil dysfunction nd emerging insights from myocrdil contrst echocrdiogrphy. Am J Crdiol. 2007;99(6B):7D-14D. 43. Long JA, Chng VW, Irhim SA, Asch DA. Updte on the helth disprities literture. Ann Intern Med. 2004;141(10):805-812. 44. Mt-Greenwood E, Chen DB. Rcil differences in nitric oxide-dependent vsorelxtion. Reprod Sci. 2008;15(1):9-25. doi: 10.1177/1933719107312160. 45. Skinner J, Weinstein JN, Sporer SM, Wennerg JE. Rcil, ethnic, nd geogrphic disprities in rtes of knee rthroplsty mong Medicre ptients. N Engl J Med. 2003;349(14):1350-1359. 46. Yp J, Lim FY, Go F, Teo LL, Lm CS, Yeo KK. Correltion of the New York Hert Assocition Clssifiction nd the 6-minute wlk distnce: systemtic review. Clin Crdiol. 2015;38(10):621-628. doi: 10.1002/clc.22468. Full text nd PDF t www.jmc.com 480 AUGUST 2017 www.jmc.com