An increasing number of medications are being developed as either injectables or intravenous drugs. Physicians often administer these medications; thus, they are largely paid for via the medical benefit. A subset of these physician-administered outpatient drugs (PAODs), known as specialty medications, provide a highly sophisticated treatment, generally when there are few or no other treatment options available. Some of the benefits of specialty medications include the reduction in the number of relapses; prevention of disability progression; symptom management; maintenance and/or improvement of quality of life; and, sometimes, disease remission or cures. These specialty medications have piqued the attention of employers, more so than PAODs overall, because of their relatively high costs.
In this Issue Brief, we use data from the 2019 IBM® Marketscan® Commercial Claims and Encounters Database to analyze waste caused by pricing failure and measure site-of-treatment price differentials for PAODs. This analysis is important not only because of the observed price differentials but also because waste from site-of-treatment price differentials is being compounded by two other trends — the shifting of care from physician offices (POs) to more costly hospital outpatient departments (HOPDs) and the fact that prices for care in HOPDs are growing faster than PO prices. Ultimately, employers and workers bear the brunt of cost differences when HOPDs perform services that can be provided in less costly POs.
Key Findings:
- Just over one-half of PAODs were administered in HOPDs. One-third were administered in a PO, and 9 percent were received in other settings, such as a patient's home.
- Allowed charges were higher in HOPDs than in POs for all but two of the 72 PAODs examined in this study.
- In the aggregate, employers and workers would collectively save $10.3 billion annually if price differentials between HOPDs and POs were eliminated for the 72 PAODs examined in this paper. If we extended the savings to all PAODs, aggregate savings would be $14.1 billion each year.
- On a per-member, per-year basis, savings would be $80.21 for the 72 drugs examined in this paper and $110.03 for all PAODs if price differentials between HOPDs and POs were eliminated.
Employers could cut spending by $14.1 billion by shifting patients away from more costly HOPD settings or by negotiating site-neutral pricing for specialty medications. The $14.1 billion represents 1.5 percent of total health care spending on workers and their families.
This study was conducted through the EBRI Center for Research on Health Benefits Innovation (EBRI CRHBI), with the funding support of the following organizations: Aon, Blue Cross Blue Shield Association, ICUBA, JP Morgan Chase, Pfizer, and PhRMA.
Figure 1 ?????????????????? ?????????????? The p H Me Sim m W row or inn ev m ila e e ne , nt b st rDo A ly xt e r ion o a, a p r r st t w on rAllo e S ic e he g a ing p f d ic N vin n it in t ., we w is -for N a a gs he c b y a d - om s t ili nc he P tay o y a e na a s t p ym lt L sh ly r . hog o a -K tc ic r e e o a e rnt a a e ut r l p t t e ss he ing, - - sof ser ion r Dif oc sa -p v a ;eoc v ic nd f s sy ings. e e s w km r se St -t b p a a a - t spe y S s t For om c ne ie o w nding it B e m d -e . txa ew a D tof na e m is us , rt m p g - us Trea ele on e ine tm zin e , -rre ts of e he a tnt t he sea . me ; -20 d P m p reA 16 c nt e fin ah ha OD rint -. ? uni e" d e s Fa na - s sh t spe c c ont p nc tror ow n ic e rts A e ibut a le n th d nd/or ss s iff so ion e a out c rt- e ia t cnt im he on tof eia p d cinf r lp e ov W b oc plue e it tek - th T m w in e nc te e - e e a nt y em n HOP r s a osin of of ploy rpq ee eua fe m rD K cre s lit e ina r nt int a ng ynd -se a of bg p ae P Ised hy lif nhi O of esi s;-, b c the a o ia it for nd tor ns a a l lt , h is- Paul Fronstin is Director of the Health Research and Education Program at the Employee Benefit Research Institute Location, Location, Location: Spending Differences for ???????? Sa m ?????????????? ple Cha =racteristics, 2019 20 128.5 ?????????????? e spe som laarc gh nding e ertm itm ha ee d b I nit s, n th t ic eha ne a ia d tn th ion is fit t eion a e s inf a -e a se 5378. lue nd ov nd re nc e in a m rA a ed is ll g he of si p gon o r rce e p ost g nc ula or ae t sh e tc ion. A . ur a m Thi reing ong sP .s r ha These e eMe rv q m e uir n w a dc ic em yd a he e spe rc e dle n o ic Bnding aa ens ut tnions e ing -fic of a - ia ha p tche rc oc ie ount v k e s de a W p tt ique e a it cd os h C of for td s a c hr la t18 he onic rim e p a s w sie tg My tre nif cit nt eh sus ent ic ion o loid a of ntp lLout y f e e e c uk high. t m - of e “p uni m loy -p ia oc te s ."” rk s Jour e field , tm spe or na ve nding al o lue so t f C s. ha lin a Bm n a icsed ong al on (EBRI). M. Christopher Roebuck is President and CEO of RxEconomics, LLC. This Issue Brief was written with assistance 2019 W Allow e exa edm pine aym d e pnt atsie w nte rpea y high ment ers f in HOP or all c Dla s t im ha s for n in tPhe Os 72 fo rP a All OD but s e txa wo m of ine the d in t 72he PA pO aD pe s re a xa nd mine found d in t tha his t fe stw udy. plan The Physician-Administered Outpatient Medications by Site of P t our he AOD tpot rs m io al a rOnc or w do e ult rolog k g , e pw ne op ye rula 34 a re lly m t(ion in ,3 ov b 6) e e : cda 43 our us cla 2e 3 im - sa of 43 s w m t2 he p 8. Ac he le ir r b e re ut :c le a s only tiv sed ely a Oc high cctount obc eo re st 19 ds. for , 20 14 17. d peroi:10.1 cent am 20 ong 0/J C us O.201 ers of 6.67 the. 41 7284 P.A ODs examined in from the Institute’s research and editorial staffs. Any views expressed in this report are those of the authors, and a mnnu emb al em rs p edaia id G n p e any n ric d te e hin rdiff g e out ren of tia p l w oca ks et$ 5,09 for the 9 b se ut drreugs. ache d M $ or 78 e ,674 spec ifi for ca one lly, 91 spe pceifi rcce on nt c of olog the y c inj lae im cts ion. hadT $0 he in t avehe rag e unit We also present alternative estimates that scale up to represent 100 percent of PAODs. This step assumes that the Trea Fi Frna onst llyin, P , ins ttme aeul, M ad of n . C int t hr rod istop uche ingr fin Roe anc buc iak l inc , and entB iv re uc s, w e Chi . St ch ma uart,y "or Loc m aa tion, y not L oc wor atk ion, , em Lp olo cayte ion: rs aC nd osins t Dur iffe errs c enc ould es for m ove the paper (Figure 6). should not be ascribed to the officers, trustees, or other sponsors of EBRI, EBRI-ERF, or their staffs. Neither EBRI nor Male 49% d p dr e rugs ic de uc d tnot ible iffe r st e field nt udie ia,l w d 81 (a i.e p s 2 e., r01 cte he nt p e rha e rcm e da nt $0 ini ( ng Fi coins gur 27e ur p3e a )rnc . cIee n o nt , a tof nd heP r98 A w OD or pe d e rs, c xpe eon ntndit ha avd e ur r $0 a eg s) ec , w op pould la ay n p m ha e ant yvm s. e etnt he Ov s t e sa r o a m H ll, p eOP ala v De n s r a m w ge e em r e H b OP te rr iple D p am y w m aha re knt tup s Augu st 19, 2021 • No. 536 p E a mtp ieloy nt • s fr erOnc U s oft om nit olog s< e HnOP = y t1 o r D y Me s t tro d w o ic ue se ine ot re he c s B m ost r is a si si sed tsh e ng s of a .ro ing n Sit tre to a e t m m of a ena Tr nt e g ba eyt m spe re em nt nd ov ," ing ing EBon RI the I spe sH su OP ce ia D B ltr s fr yie m f, om eno. dic ta he 49 tions ir8 (E ne b tm w ep c or a loy us k.e e Suc e of Bh a etne hen a fit ir high r Re rasea ng ce ost rm ch e . nt W he is n EBRI-ERF lobbies or takes positions on specific policy proposals. EBRI invites comment on this research. By Paul Fronstin, Ph.D., Employee Benefit Research Institute, and M. Christopher Roebuck, p tow lan p ard a y dm ed euc nts tible Fw ee m s re a a lm t eo ount POs f ed or t o the only sa m 1 p e euni rcetnt of of me tot dic al sp ation. ending The ,m ae nd dia cn oins uni ur t a pnc rice e a dm iffe ount rent ed ia l w to a only s 9 51% 81.6 p perce ernt ce. nt W . e did as the 72 drugs analyzed (i.e., 73 percent of PAOD expenditures). Also presented is the aggregate HOPD markup — m it ost com •c eom s ItP ns m oa on t y tit he m ue in tr e nt e , st t Ja a aail p m ff nua ount -m ha rod y rm s 20 ea w l c 20 he e yr e ). b a lt e < ne h ma $0. fit01 , int n.e ea na rly nc ae ll or cov ga eniz re Fd a igure tw ions ork 6 e (H rs h MOs av)e bcut ov e ca ra n b gee for ap p sp lie ec dia m ltor y m e e gd eic ne arta ion llys, a to a nd nyn early Ph.D., RxEco Age nomics , Years, LLC 41.8 w not Thu hic s, m eh is e xam os ine ss t of e tnt he tia he ll p y rp ic ot the ee d nt e iff xpe ia el s rendit a nt via ings fr ur l b ee -tw w om eeight e n HOP elim edina av Dte ing r aa nd gp er ot of ich ee d d rr ug iff sit e-e rspe s of entcia ifi tls rce a w HtOP ould meD nt ini m , b a te ia rk cll ups ayus b e e — for r edflect em riv ost e ed d of a in s te he am g g p P rloy A eg OD e art- e s a st sa nd uvd ings ied, 6 Distribution of Out-of-Pocket Health Care Spending, 2019 ne one tw -ha or • k lf (4 p Pla a5y n. m pe P errnt c ov e nt a ide m ) ount rof s ctould he s w me ra re e rspond e m in isa si ng p bly a . n w lowit eh a ring sep theairr aptreic ceost s so th shar aing t the tie y rm for ay trhe etm ur.n int Amo ong the t hose networ wk or . k Thi ers, 45 s strategy Suggested citation: Fronstin, Paul, and M. Christopher Roebuck, “Location, Location, Location: Spending Differences 18–24 ins L vero ur y c e fe rw -a p a w t id eion ra em aount d, mini L s. st o e Hc row eda eout vteion rsi , d to e ta, he n HOP L do egc r D e ea or t t ot Pion O. al sp e: nding Sp fae llsn ad s ai n resu glt Dif of elim fer inate ing nsic tee -ofs -t15% refor atme nt price divided by the aggregated counterfactual HOPD cost. E Fr ndno onstin, P te aul, M s . Christopher Roebuck, Jason Buxbaum, and A 7 . Mark Fendrick, "Do People Choose Wisely After p ha er s it ce •s lim nt ha Cila tva eim t ions as cw op a ea s w re y m m eell. I ant rk, e t a d m And ge a as d y53 18 c not a – p p 64 it ew , a rO c t or v e e eral d nt k . w lha ell in a ve c Aoins ge re d a 18 ur s –w a 64 nc it , h lim Pe hy . s ic The iit an e-d A a dm p vre ini ov rsa tered ide ge rc O c op utp hoi a ati y cen m es t e Dr nt or ug is in a Us $1 ers re 0a 9s aw nd hetrhe e paow vee ra rg ful e for Physician-Administered Outpatient Medications by Site of Treatment,” EBRI Issue Brief, no. 536 (August 19, 2021). 25–34 18% differentials, workers and their families should share in this savings as premiums fall. Figure 3 Phys Sa ici tisfya ing n- HAd ealth P m 8laini n Des duc tte ible re s? d Ev ide Ou ncet Fr p om a t the ien Uset of M Low e-d Vaica lue Hetaion lth Cas re Se bry vic Si es," t Ee BRI of Iss ue 90% 72 72 c hospit oinsur aa l sy ncs et e ism 26 s lip m eit rc p ea nt y.er s’ In apbla ilins ty w to ite h th xclude ree o cr e rm taor in hig e tieh rs of -costc ost prov sh ide arring locfor ations pre sc from ripti ton heir d rne ugs tw(or the k. most common ???????? Introduction 35–44 Average and Median Percentage Price Differential per Unit Base ???????? d on 72 21% ?? 1 ???????? We deletB erdie cfla , no. ims w 51it 6 (E h unit mpsloy =1 ee b eB ce ane usfit e itRe wsea ould rc h I hav ns e tb ite ut ee n, im Oc ptoss obe ible r 20 t20 o d ). etht ^ ertm ps:/ ine /ww whe wt.e he brr i.or theg se /do cla cs/ im ds inv efault olv - ed See Fronstin, Roebuck, and Stuart 2020. p Trea lan desit gn) me , aven r ?????????????????? at ge copayme ???????? nts are ???????????? $11 for fir =st ? -tie ?????????? r drugs ?????????????? (i.e., gene / ric ? s), ?? $35 for • second-tier drugs (i.e., preferred Copyright Information: This report is copyrighted by the Employee Benefit Research Institute (EBRI). You may copy, ?? ?? ???????? Physician-Administered Outpatient Drugs This H Iss igue h U Brsie e45–54 frs is of theH te hir alt d h in a Ca ser rei e Sse e rxa vices mining site-of-treatment price differentials for health care services. First, we 1 + ???????????? 23% 79% ?? National D sour rug c eC/e od be ri -(is Nsu DC e) -- bb ra iesed f/eb b rill i_i ing b_5 (16 vs. _low HCPvC aS lhe uni alt th -b -22 ased oc) t2 . 0 C.p lad im f?s for sfvrsn fiv =e56 d6 rugs 93a 2f whe _6r . e units=1 was the only 9 ?? =1 ?? =1 80% brands), $62 for third-tier drugs (i.e., nonpreferred brands), and $116 for fourth-tier drugs (i.e, specialty drugs). p 2 rint, or download this report solely for personal and noncommercial use, provided that all hard copies retain any and See Fronstin and Roebuck 2021. determined that hospital outpatient departments (HOPDs) charged 81 percent more for oncology medications than By Paul Frons 55–64 tin, Ph.D., Employee Benefit Research Institute, and M. Christophe 23% r Roebuck, 250% 74% R pla e us fe ible re v nc alue es ( because all NDCs had the same dosage/strength as the HCPCS unit) were retained. Finally, as in our Average coinsurance rates are 18 percent for first-tier drugs, 25 percent for second-tier drugs, 37 percent for third-tier all copyright and other applicable notices contained therein, and you may cite or quote small portions of the report In prior research, we found that 10 percent of the population accoun 1t for 70 percent of health care spending p Fihy nasi lly cia , w n o e ff conc ices (P lude Os) by, c ca olc nturla ollin ting g t for he d pre ug rc em nt ix a age nd of trte ot aa tm l he ent a ltint h ca ens re it y sp . e W nding e conc tha lude t the d ta ha gg t,r e ha ga dt e P O pot pe ric nt eia s p l sa rev ving ailes d in 3Ph.D., Fulton, B R rex nt Eco Pe D. rnomics 20 so17 n . C"o Hv , ee a Llt rLC e h C d are Market C stoncentra thtion Trends In The United States: Evidence And Policy p rSeeveious Ins tit ana ute ly o ses, f Mew dic e ine trim om f th ed e tNa ails tional at t he Ac a 1 10 de mie and s 99 2010, pe S rc hra ent nk, iles Ro of gstta he d (ad nrd ug P-aspe rekh cifi 20 c)19 uni . ts’ distributions to reduce Berwick, Donald M., and Andrew D. Hackbarth. 2012. "Eliminating Waste in US Health Care." JAMA 307 (14): 1513- drugs, 70a %nd 28 percent for fourth-tier drugs. provided that you do so verbatim and with proper citation. Any use beyond the scope of the foregoing requires EBRI’s (Fronstin and Roebuck 2019). Within this group, 50–60 percent reached not only their deductible but also their the HOPD setting, payers could have saved $9,766 per oncology patient in 2016. Next, we analyzed 25 outpatient would reRe pre sponse sent Po . lis." cyhH oe ld ae ltrh Affairs 36 (9): 1530-1538. doi:10.1377/hlthaff.2017.0556. 62% the influence of outliers that likely represented data entry errors. 4 1516. 201% prior express permission. For permissions, please contact EBRI at permissions@ebri.org. Seem W axim inn um et a l. ou 2018 t-of-.p ocket limit. As a result, much of the health care 2these people use is subject to low or no cost services consist Sp ing ou of selab, imaging, and selected specialty medications. For each of the 25 health care services 200% 24% Ther e is evidence that patient cost sharing has an impact on use of specialty medications. A recent systematic review of 60% sharing, which lowers the savings potential for these plan members from any reduction in price differentials. Institute of Medicine of the National Academies. 2010. The Healthcare Imperative: Lowering Costs and Improving 5 A T A G L A N C E examined, HOPDs charged more per unit of service than POs; costs were higher by between 15 percent and 531 Using the rema C ini hing ld/o (tch le ea r ne de dp ) ednad te a, ntfor each claim, we divided the cost by the number of units. We subsequently See Cooper et al. 2019. 15% Carls, Ginger S., M. Christopher Roebuck, Troyen A. Brennan, Julie A. Slezak, Olga S. Matlin, and Teresa B. Gibson. the literature concluded that reductions in specialty drug use were associated with higher cost sharing, with stronger Report availability: This report is available on the internet at www.ebri.org Employers would continue to realize savings if these plan members switched from high-cost to low-cost Results Out comes. Edited by Pierre L. Yong, Robert S. Saunders and LeighAnne Olsen. Washington, DC: The National percent, with a median difference of 91 percent. In aggregate, we estimated that employer health plan sponsors and calculated the T y ap ve er a og f eH ce ost alth pe Pl r uni ant by drug and place of service, namely: 2012. "Impact of Medication Adherence on Absenteeism and Short-Term Disability for Five Chronic Diseases." e 6 ffects for non-initiation or abandonment of a prescription at the pharmacy and somewhat smaller or no effects for refill See Figure 9.9 in https://www.kff.org/report-section/ehbs-2020-section-9-prescription-drug-benefits/. 50% providers, because employers pay for nearly all of these claims. 150%Academies Press. plan enrollees could have saved more than $11 billion in 2018 on these services if HOPDs had charged PO prices. This HMO/EPO 11 14% An increasing number of medications are being developed as either injectables or intravenous drugs. Physicians often Journal of Occupational and Environmental Medicine 54 (7): 792-805. doi:10.1097/JOM.0b013e31825463e9. behavior once therapy was initiated. More recent studies have also found that member cost sharing is an impediment ???????? ???? W 7 here Do Plan Members Receive Physician-Administered Outpatient Medications? ?? and ?? See Figure 9.10 in https://www.kff.org/report-section/ehbs-2020-section-9-prescription-drug-benefits/. represented ab PPO out 1 /PO perS cent of total health care spending ?? on wor ?? kers and their families. These price differentials are a 12 59% administer these medications; thus, they are largely paid for via the medical benefit. A subset of these physician- Table of Contents to optimal specialty medication use among Medicare beneficiaries covered by Part D. Koller, Christopher F., and Dhruv Khullar. 2019. "The Commercial Differential for Hospital Prices: Responses From This m 40a %y come as a surprise, especially since plan members are seeing their out-of-pocket costs increase mainly due to 3 The Che rd ne isw tr, ibut Mic ion o hael, f tr Za ec akt m Ce oop nt e sett r, E ings for ugene LPA arO sen Ds- H isa lloc show k, na nd in Fi Fig ona ure Sc 2. ot PA t O Mor Dst on. are 20 prov 19. " ideA dr e in t Hhr eae lteh C loc aa re tions Ser: vic HOP es Ds; form of waste, H cR ont Aributing to the 20–30 percent of overall spending that may be considered wasteful. 8 14% administered outpatient drugs (PAODs), known as specialty medications, provide a highly sophisticated treatment, 98% I nt Se rod e Fi uc gure tion 9. ................................ 11 in https://www.kff................................. org/report-section/ehbs ................................ -2020-section-9-prescrip ................................ tion-drug-benefits/. ......................... 3 r Ther ising efor hea e St lt , a h p dtreug s a la-n d spe nd eE c difi m uc c p tloy H ible OP es. rD s." The mJour arfa kups na ct l o is( i.e f th tha ., te p pA rla ic m n me ee d ric iff am e n M rb ee nt e rs us ia dic lsa )ing l w Ae ss P re o A c OD cia alc ts, e ion ula spe t E1 ed-c 2. d a ia slly : oi:10. spec10 ialt 0y1/ m jaem dic a.201 ation 9.9 s, 27 we5. re 100% POs; and Shop othepra , b w le hic ? E h mos videnc t e com from monly the in Consu clude m s ptthe ion o paf Low tient’s er hom -Lim eb . Jus MRI t ." o vN eB r E one R Wor -half king (58P p ae precre N nto. ) of 24P 86 AOD 9 (N s a w tiona ere l Cost sharing mH aSA y ha -e vle ig iit bs lim le he ita s w lth he pla n it n comes to managing use of high-cost medications. Many plans 13% have maximum generally when there are few or no other treatment options available. Some of the benefits of specialty medications 9 mostly high users of health care services in 2019. Three-quarters used at least $3,000 of health care, and median Backg 3r 0ound % .......................................................................................................................................................... 3 I n th Seeis Fi I gss ure ue 9. B6 rie in f, http wes foc ://w us w w on .kff w.a ost rge / re ca pus orte-d se b cy tio pn/ ric eing hbs-fa 2020 ilure -s a ec nd tioc nond -9-puc ret sed crip ation not-he drug r a-na bely ne sifits s of /. site-of-treatment ???????? administered in HOPDs. One-third (34 percent) were administered in a PO, and 9 percent were administered in another dollar am Bount ureas for u of m Ece onomi dicatio c ns Re . sea And rca h) ll p . lans have an overall maximum out-of-pocket limit, which is not difficult to include the reduction in the number of relapses; prevention of disab ?? ility progression; symptom management; L i, Pengxiang, S o Tia urcny e:a E n Hu, X mployeiny e Ba en Y nefiu, t RSa eslim ear cC hha Inhin stit, utN ea eb sila tim D aa tehod s ba w sa ela d , oMa n ar dim ssia ni s Btlum rativ, e A em nry o lR. lmeP ne t ta tin t, d and Jalpa A. ?? ???????? overall spending was nearly $8,000. These plan members were likely to not only reach their deductible but also to often price differentials for outpatient drugs. Specifi????????????cally, we expand = 1 an -d( update ) our prior analyses to include 72 physician- ?? 10 ???? D sett ata ing and . Methods ................................................................................................................................................ 4 reach when som claim eone s da tis a .using a high-cost medication. Where specialty medications are covered by the medical plan See Figure 9.6 in https://www.kff.org/report-section/ehbs-2020-section-9-prescription-drug-benefits/. maintenance and/or improvement of quality of life; and, sometimes, d ?? isease remission or cures. These specialty Doshi. 2017. "Impact of Cost-Sharing Increases on Continuity of Specialty Drug Use: A Quasi-Experimental ?? 50% reach their maximum out-of-pocket limit. 18% 20% C ad ong mini rest ss eiona red l B out udg pate ie t nt Offi drcug e. s (P 2012 AOD . Offs s). eTog tting ethe Effre, ctthe s of se Pm reesc dic ripti ations on D arc ug count Useed on for Me 49 di cp ae re rc ’s Sp ent of ending all cla for im s a Mend dica 7l 3 because they are physician administereed, outpatient cost sharing (i.e., deductibles and coinsurance) would apply. medications Study." ha Nov t e e H : p e H a ique M lth Se O= dh te r he a vlic t h a e ts Re m te ant inse ion o ten aa rc nf e h. ce m A oc rp g cloy a en ss ie ze r ad s, m ti o Oc n;or tE ob e P e O so t r= 19 eha x, cln P 20 us17 iv A eOD . pd ro oi:10.1 s viov dee r roar11 ll g, a1 n b/1 ie zc a 47 a tius o5 n-e ;67 P of P 73 O the .= 12 pir r74 e rfe e 4. rla r etd iv ely high costs. 11 Data and Study Sample ........................................................................................................................................ 4 14% See Doshi et al. 2016. p Ther erce ent w a of Se s a r qv ll spe uit ice es. anding C bong it of ron evss aP riona iA aOD tion in l B s p udg atid he etv ia p Offi e trhe cce ent . m ht aegt de pic :/ of a/ww l b PA ene OD w.c fit sb . o.g bFoll eing ov ow /s aing id te m s/ ini our dst efa e drult e erd iv /f in HOP a ile tion o s/cbofile D f d s a rug snd /a -spe tP ta Os. cc hm ifiOv ce p n ertric s/ one e 4 37 m-a 41 ha rkl-ups f We report the average and median HOPD markup across the 72 drugs. However, spe pnding rovide ron orgta he niz se atiP oA nOD ; PO s S w =ould poin tb oef s su erb vj ic ee c;t H to RA the =h e sa alm the r e m im ab xim ursum em eout nt a -of rra-n pg oecm ke et n tli ;m Hit S, Aw =hic heah lt h c ould easily be The overall sample used an average of $1,690 in prescription drugs in 2019, and the subset of those using any of the 8% An10 alysis % ................................................................ ® ................................ ® ................................................................. 5 b 12 w ye rH eOP adD m Me s r inid est la ice a tr iv lOff ee d tsets in t o Phe Os, w -11 H -O 29 e P -D c 1a 2.p or lcula d Pf. O t eb e est tw im ee an 30 ted sa and vings 70 tp ha ertc e could nt of b the e r e tim aliz ee . dI n cont if the r highe ast, v re p ry ric fe es c w PA haO rgD es d w by e rH e OPDs 7% savings account. In th See is Do Iss sue hi e B t raie l. f, 2016 we ;us W einn, dat Ke a fr aom ting ,t he and 20 Du 19 s eItz Bina M 2016 Mark;e a tsc nda L ni e t Ca ol. m2017 merc.ia l Claims and Encounters Database to r Ma earcthe in, A d a nne s the Bse ., Mic meadh icH ataions rtma a n, reB oft enj ea n hig min Wa h-cost shingt as w on, ell. A a Thu ron s, w Catlin e t , ur and n oa ur nd a ttthe ent N ion ationa to al He nothe altrh Expe approndit ach ur to e 0% 72 PAODs used an average of $3,515. While the subset used more prescription medicines than the overall population, In the last stage of the analysis, we estimated the potential savings that could be realized if PO prices prevailed in the c aould dmini bst ee e rlim ed ina in tthe ed . 'ot her' setting 10 percent of the time or more. Average Median analyze waste caused by pricing failure and measure site-of-treatment price differentials for PAODs. This analysis is Results ................................ Accounts Team. 2017................................ . "National Health Spe................................ nding: Faster Growth I ................................ n 2015 As Coverage ................................ Expands And Utilization 7 managing the cost of specialty medications: focusing on site-of-treatment price differentials of such medications. 13 they also used more health care in general than the overall population. As a result, we find that pharmacy spending as Author estimates from the 2020 Annual Social and Economic Supplement to the Current Population Survey. C ooper, Zack, Stuart Craig, Martin Gaynor, Nir J. Harish, Harlan M. Krumholz, and John Van Reenen. 2019. "Hospital HOPD setting. To do so, using the drug prevalence rates measured in our sample, we first extrapolated the number of 0% important not only because of the observed price differentials but also because waste from site-of-treatment price Increases." Health Affairs 36 (1): 166-176. Accessed October 19, 2017. doi:10.1377/hlthaff.2016.1330. Source: Authors’ analysis of IBM MarketScan administrative enrollment and claims data. a pWhere ercentaD ge o of Plan tot M al he em O autp b lters h ca atient R re ece spe ivnding e Phy si wcian as low -Ad er m ain mist ong In epa red Outp P tiA enOD t usati ers t ent haM n a edi mong cati o the ns? e nt ................................ Pir ha er ma pop cyulation. Phar....... mac y 7 Thi s analy Prsi ics is es G im repw or Subst tant not ant ia jus lly t b Fa est caeus r Tha e of n P the hy ob sic ser ian P ved r ic pe ric s For e diff Heospit rentia alls -B a bsed etw eCean reH I OP n 20 Ds a 07nd –14 P." Os H b ea ut lt h Af also fa bir es c a38 use 14 HOPD patients for each drug that would be expected in the U.S. population of adults, using an estimate of 128.5 million See Koller and Khullar 2019. differentials is being compounded by two other trends — the shifting of care from physician offices (POs) to more costly Potential Aggregate Savings spending accounted for 24 percent of spending for all adults in our sample but only 17 percent among adults who used 13 waste from (2) : si18 te-4of -18 -tr9. d eatm oi:10.1 ent pr 37 ice 7 /h diff ltha ere ffnt .201 ials8. is 05 b42 eing 4. compounded by two other trends. First, care is shifting from Source: Authors’ analysis of IBM MarketScan administrative enrollment and claims data. adult Hosw wD ith o All emo pwe loym d e Pn aty -m baents sed he Diff alter h p by lan Si s.te of Treatm Fent igure ? ................................ 2 .................................................... 8 Martin, Anne B., Micah Hartman, David Lassman, Aaron Catlin, and and the National Heatlh Expenditure Accounts Data and Methods hospit 15 al outpatient departments (HOPDs) and the fact that prices for care in HOPDs are growing faster than PO prices. See Fulton 2017. Ana at lea lys st is one of the 72 PAODs examined in the paper (Figure 5). I Pn th Os teo am gor gre e gca ost te, lye m HOP ploy Ds. I ers n 2004 and w D o , irstri a kp ep rb s roxim u w tiould on a to ecf ly olle V 94 ari c tp o ive u ers ly ce Ph sa nt v y of e sici $ c10 he an .3 m -A ot bd ill he mi ion ra np ip ste ye inf r ryed us eaions r if p w ric ee re d a iff de m rini ent st ia els re b de in P tweO en s, Team. 2021. "National Health Care Spending In 2019: Steady Growth For The Fourth Consecutive Year." Ultimately, employers and workers bear the brunt of cost differences when HOPDs perform services that can be Potential Aggregate Savings ................................................................................................................................. 8 ^ ???????? ???????? Doshi, Jalpa A., Tianyan Hu, Pengxiang Li, Amy R. Pettit, Xinyan Yu, and Marissa Blum. 2016. "Spe4cialty Tier-Level Cost Outpatient Drugs, by Site of Service 16 H but OP b Dys a 20nd 14 PtOs hat w pe erre c eent lim ag ina e ha ted d for drop the pe ??d 72 t o PA 57 OD = p s e ???????????????????? re cxa ent m ine with a d in t cor his re p spond a •p 128 er ing ( .5 Fi ??????????????gsh urift e 4t)ow . If w arde H eOP xte Dnde s. d Se tc he ond sa, vp ings to a rices forll See White and Whaley 2019. ?? ?? Using the 2019 data, for each of the 72 medications (indexed by d), by place of service (denoted by HOPD, PO and Data and Study Sample Health Affairs 40 (1): 14-24. doi:10.1377/hlthaff Figure 5 .2020.02022. 70% p C ro ovnc idelus d in l ion and ess costly P IOs m.pli cations for Employers and Insurers Sharing and Biologic Agent Use in the Medicare Part D Initial Coverage Period Among Beneficiaries With Member Savings ................................................................................................................................................... 9 hospit PAODs a, l-a bg ag sed reg a out te p sa atvie ings w nt carould e ar eb g er $1 ow4.1 ing b fa illst ion er p tha er n p yea hry . siThi cian p s rerp icre essent mor s 1.5 p e gene er ra ce lly n. t For of t ot exa al he mpa lelt,h ca a rercee nt sp e st ndin udy g on OTHER), we measured: Distribution of Health Care Spending, 2019 17 ® ® Thi Ss st ee http udy ma s://w ke ws us w.we sj.c of ot m/ hea rticle 2019 s/IBM meth oMa dolog rkey tsc -ho aw n-the Co-m ws m j-e arna cia lyl C zela d-im hos a spita nd l-p Eric ncing oun -d te ar ta s D -11625583571 atabase (CC.A E). The CCAE N Oe ur xt fin , w deings ha derivev de t im hep p lic ot aetnt ion ias for l per -bp ot ath e ient m,p p loy er- eyre s a arnd (PP ins PYur ) e sa rs. vings There fora e re a ca h d num rug be ar ss of um aing ctions tha tthe the se ob thir ser d-v pead r ta yv p ea ra ye gr es Rheumatoid Arthritis." Arthritis Care & Research 68 (11): 1624-1630. Accessed October 19, 2017. w found orkerts a hatn d b etthe we ire d n 2007 epend e ant nd s.20 14, prices for hospital-based outpatient care increased 25 percent, while physician Roebuck, M. Christopher, J. Samantha Dougherty, Robert Kaestner, and Laura M. Miller. 2015. "Increased Use Of Key C onc Fin lusion dinag nd s: Implications for Employers and Insurers ........................................................................................ 10 Aged 18–64, Overall Aged 18–64, Physician-Administered Outpatient Drug Users datab 60a % se contains membe 58% r enrollment information as well as adjudicated inpatient and outpatient medical and a ca nnu n ta ak l e H. OP FirD st c , ost the y re c flect an e ex de trhe t p reest ss im uraet e on d H hospit OPD a m ls a rtk o up. sh ift In o from the rd is wc ooun rds, w tede c c ha alc rg ula e c te ont d trhe ac td s b iffa esed renc on e ba e tm wult een th iple e o f 18 5 http://onlinelibrary.wiley.com/doi/10.1002/acr.22880/full. See • http Tot ://a vb l num idcente ber. r o of rgm /fre em que bently rs (-N a= sk10 ed.8 -q ue mill stions ion) /in t . he sample. prices g 70% rP ew re sc 6 p ripti ercon entD .r ugs Ultim Re ad te uc lye , s Me employ dica el C rs a ost nd s I wn M orke ed rs b icae id arP tohe pula brtunt ions of ." H cost ealt d h Af iffer fa eir nc s e34 s w (he 9): n HOP 1586-D 1s p 593. erform Re pha fe rr m ea nc cy e sc ................................ laims. We constructed................................ an analytical dataset of ................................ adults (ages 18–64................................ ) who were continuousl ......................... y enrolled in 11 65% a Me ctd ua icl aa re v etro ag som e ae nnu ota he l H r OP pros D pce ost ctiv a end catse her a count te. Ee m rfa ploy ctua erl (un s could der a P ls O o p erxe ice rts) su ac vh p erarg eess aur nnu e on al HOP healt Dh p cos latns . Thi to s m do e the thod sa me On a per-member, per-year basis, savings would be $80.21 for the 72 drugs and $110.03 for all PAODs if price • Total number of patients for each drug (P ). ser vic • es t Jus hat t ov can b er one e pr-ov haide lf of d P in le AOD sss cw ost erle y a Pd Os or minist in st dered a nd in HOP -alone D s. la b One or -im thir ag ding werfa e ca ili dtm ieini s. stered in a PO, and 9 percent doi:0.1377/hlthaff.2015.0335. 19 See Fronstin and Roebuck 2014. Doshi, Japla A., Pengxiang Li, Vrushabh P. Ladage, Amy R. Pettit, and Erin A. Taylor. 2016. "Impact of Cost Sharing on employment-based health plans in 2019. Members in capitated plans were excluded. A total of 10.8 million individuals with the hospitals in their networks. A coalition of employers across Illinois, Iowa, and Wisconsin is already trying holds treatment intensity constant in the HOPD setting. Specifically, differ 50 e% ntials between HOPDs and POs were eliminated. Endnotes ........................................................................................................................................................... 13 • w Ae vreer a rg ee ce num ived bin oth er of e cr la sett ims p ings, er psu atc ie h a nt s a for p ea atcie h nt dr 'sug hom (X e ).. d 60% 14 Specialty Drug Utilization and Outcomes: A Review of the Evidence and Future Directions." American Journal of m 20et these criteria. Sample averages for the following characteristics are reported in Figure 1: gender, age, relationship 56% this. Roe Sebeuc Che k, M. rneC whr eis t a top l. 2019. her, Joshu a N. Liberman, Marin Gemmill-Toyama, and Troyen A. Brennan. 2011. "Medication • Average annual total cost (COST ) (i.e., allowed amount) per patient for each drug, as well as the member • Allowed charges were higher in HOP d Ds than in POs for all but two of the 72 PAODs examined in this study. ???????? ???????? ?? Managed Care 22 (3): 188-197. to policyholder, and plan type. ???????? Backgr Ad ohe und renc e Leads To Lower Health Care Use And Fig Cu ost res D 4 espite Increased Drug Spending." Health Affairs 30 ???????? ?????????????? = ???????? - • Ia n th nd e p la ag ng -c rost ega t ce om , ep m one ploy ntesr.s and workers would collectively save $10.3 billion annually if price differentials 40% ?? ?? ???????? 1 + ???????????? Howe50 ve %r, increasing consolidation of health care providers makes it harder for employers and insurers to exert any kind ?? Figure(s 1) : 91-99. Estimated Savings Presc • r iption This a drugs llowcead n b for e tche ov e dre erd iv unde ation o r tf th he e p hm ar am rkaecty sh be ane re34% fit enj or oy te he d b m ye H dOP icaD l b s a ene t tfit he . M drost ug -or lev ae l a l: nd other patient- between HOPDs and POs were eliminated for the 72 PAODs examined in this paper. If we extended the 15 Fr onstin, Paul, and M. Christopher Roebuck, "Location, Location, Location: Cost Differences in Health Care Services by of prP ichys ing icia press nur -Adm e on ini hospit stea re lsd . Out In th pat e ie ab nt se nc Dreug of ssu ch market power, employers and insurers can attempt to administe sa re vd ings to a medicall P tions AOD (es, a .g., gins gre uli gn) ate a sa re vfil ings w led atould reta b il ep ha $1r 4.1 mab cie ills ion or e ta hr ch ye ough ma ar. il-order pharmacies, and those We then calculatedT to he p 7 tot 2 a Ph l p yot sie cnt iaia n-A l sa dm vings a inistes t red he O pu rtod pauc tie t nof t D the rug e sxpe (PA ctO ed D num s) ber of $H 1OP 0.3D b p ila lio tie nnts in the Figur 30 e % 1, Sit Sa em of ple Tr C eha atm raecnt te r— ist ic A s, 2 Closer 019 L ................................ ook at Lab, Imaging, ................................ and Spec????????ialty Me ???????? dicat................................ ions," EBRI Issue Brie.................... f, no. 525 5 engage patients through increased price transparency. However, price transparency by itself has been found to be Roebuck, Mark C., Robert J. Kaestner, and Julia S. Dougherty. 2018. "Impact of Medication Adherence on Health 40% (?? ?? ) ?? ?? ???????? c laim •s are On ad aj udic per-a m te ed m und ber,e p r etr he -y e paha r b rm asi as, sa cy be vne ings w fit. Aould n inc b re ea $8 sing 0.21 num forb e the r of 72 m d ed rugs icatie ons a xamine re b de in t ing his de v pe alop pere d a nd as either population for each drug and the potential PPPY savings: ???????????? ?????????? 16 = We us (Ee m dp tloy hee 20 e B 19 ene CC fit A E Re out sea practh I ient ns d tit au ta te fil , e Fe tb o rua extrrya 20 ct 21 all c ).la ims with Healthcare Common Procedure Coding ?? All Physician-Administered Outpa????????tient D????????rugs ???? ???? ?????????? ?????????? insufficient in reducing hospital prices unless combined with plan design changes intende $14 d. 1 t o bist llie oe nr patients to less Services Utilization in Medicaid." Medical Care 266-273. doi:10.1097/MLR.0000000000000870. Figure 2, Distribution of Various Physician-Adminis(t??ered Out ?? patient ) + D (r?? ugs, ?? by Sit ) + e ( of ?? Servic??e ................................ ) ...... 7 ?? ?? ?? ?? ?? ?? inj ectables or intravenous drugs. Because a physician often administers these medications, they are largely paid for via $110.03 for all PAODs if price differentials between HOPDs and POs were eliminated. System (HCPCS) codes in the “J,” “Q,” and “S” ranges for our cohort. We subsequently selected the top 100 costl20 y % sites of treatment. Furthermore, recent public policy efforts to address pricing transparency found that 34 ^ ???????? 30% ?????????? ?????????????? = ?? • ???????? ?????????????? t Fr he onst me in, P dicaa l b ul, a ene nd fitPo . M. Muc te Cn hr h tia is of lt Sa op the he vi n ut rg Roe ili s za Fb ro tion o uc m k . H f "o P ts he ep rsi ise ta stl e PA O nc ?? u O y tD p in Hi a s t ?? is ie n c gtonc h D -C e e os p na ttr rt H am te ea e dlt n in h C t re alra e ?? t iv Ce laly im ine s: xpe 'It’s Wh nsive e, re g e the ne rSp ic-e dnding omina t Ie s, d Figure 3, Average and Median Percentage Price Differential per Unit Based on 72 Physician-Administered Outpatient medications in terms of total expenditures to model further. Saline and Ringer's lactate solutions were excluded Shrank, WIlliam H., Teresa L. Rogstad, and Natasha Parekh. 2019. "Waste in the US Health Care System: Estimated percent of As w hospit ell a als s, ha in a ve g g not reg p aost te: ed usable pricing data, and another 12 percent posted$ d 8a 0t.a 2, 1but those data fell well 24% Employers could cut spending by $14.1 billion by shifting patients away from more costly HOPD settings or by therapeut St icupid class '," eE s su (H BRI O c PD h a Isss a ue ) M nt B arrk ibioti ieu f, pno c os, a f. 7 49 2 nt 3 (E PA iem O m eD tp ic s lo s, y (p e ce e or r B tm ic ene ost em fit ebr e oid Re r, sea s, b per re c ynzod h I eans r)ia titze utp eine , Oc s,t ob and er n20 onst 19e ). roidal anti- from D the rugs ana ................................ lysis. Next 17 , we examine ................................ d raw counts of uni................................ que patients and claim ................................ s for each medication ..................... by place of 8 Costs and Potential for Savings." JAMA 322 (15): 1501-1509. doi:10.1001/jama.2019.13978. short of the requirements. 20% 9% Summing across all 72 drugs yields the aggregate potential savings in dollars. 10% 19% negotiating site-neutral pricing for specialty medications. The $14.1 b 72 illion r ???????? epre ???????? sents 1.5 percent of total health care inflam 20 m %aht tor tpie s:/ s./ww How we.e ve brr, i.or ovg e/do rall sp cs/e dnding efault -on sour PA ce OD /es is bri- is csu onc e- eb nt rie ? raft /e ed b rin i_ib the _4 93 mor _hi e gchc ost oly st, cla yeim t le s-ss common, subset of [(?? ?? )] 17% service: HOPD, PO, and other. For sufficient power in subseque ?? =1nt st ?? eps, we ?? required at least 10 patients in each ???????? Percentage of Aggregate Spending 1.1% Fi gure 4, Estimated Sa????????????vings ................................ ?????????? = ................................................................................................ ... 8 spending on workers and their families. ???????? ???? 72 ???????? 72 ???? ?????????? ?????????? PAODs known as specialty medications. Specialty medications are different from traditional prescription drugs and are White, C., 24a oc nd t19 C. .p W df? ha sle fvy r. sn 20 =19 ced . 83 Pric c2 ef_ s P 6.a id to Hospitals by Private Health Plans Are High Relative to Medicare and Vary ? Our of find bings oth HOP also Dsu ag nd geP st O. tha Thi t s r atte ed m uc ptes dt o the cha set nge of p m [a (e t ?? ie dic nta tbions e ?? ha vtior o 72 )m + . aTog ( y?? be e tfr he ?? uit r, le t )he ss +,se (b ?? em ca eus dic ea ?? p tions atie nt ac s )c ]a ount re su ed b jfor ect e 49 d to ?? =1 ?? ?? ?? ?? ?? ?? EBRI Issue Brief is registered in the U.S. Patent and Trademark Office. ISSN: 0887 -137X/90 0887 -137X/90 $ .50+.50 0% ?????????????????? ?????????????? = ? ?????????? ?????????????? increasingly us Widelye: dFi tnding Po o trre ea n s fr tt ic aa om l nc Sa ea v rn Em i n ag nd s p F ot loy ro he m err - d H Lis O ee dPD a Tr ses aM ns su ap rk ca h a u re pnc s om fy A ult Ilnit l PA iple iatO iv sc e D.le s Sa r (p osis nt ea r ( Monic MS), w a, hic CA h ha : Ra s nd a p Cror evpaor lea nc tion. e rate of Figure 5, Distribution of Health Care Spending, 2019................................................................ ?? ................................ 9 p10 er% cent of all drug claims and 73 percent of total spending on drugs adjudicated via the medical benefit. very little cost sharing for PAODs. In other words, employers and insurers who would normally use a combination of Hospital Outpatient Departments Physician Offices $110.0 O 3 ther This study was conducted through the EBRI Center for Research on Health Benefits Innovation (EBRI CRHBI), Fronstin, Paul, and M. Christopher Roebuck, "Reference Pricing for Health Care Services: A New Twist on the Defined • Prevalence in the HOPD sett18 ing was also calculated 19 for ?? = 1 each drug: a bout 0.1 percent in t mehe mb Unit er, p ee dr St ye aa te r) s. These specialty medications frequently require specific handling and/or storage. doi:10.7 © 20224 1, E 9/m RR3 plo03 yee 3. Benefit Research Institute -Education and Research Fund. All rights reserved. value-based insurance design (VBID) and reference pricing to vary patient cost sharing based on the choices that Figure 6, Distribution of Out-of-Pocket Health Care Spending, 2019 ........................................................................ 10 with the funding support of the following organizations: Aon, Blue Cross Blue Shield Association, ICUBA, JP Contribution Concept in Employment-Based Health Benefits," EBRI Issue Brief, no. 398 (Employee Benefit Source: Authors’ analysis of IBM MarketScan administrative enrollment and claims data. they make regarding choice of health care provider may find that those ????????efforts do not change patient behavior with Dividing 0% the aggregate savings by the estimated 128.5 million adults with employment-based health benefits gives the Percentage of Aggregate Spending ?? 1.5% Morgan Chase, Pfizer, and PhRMA. Specialty medications provide a highly sophisticated treatment — genera ?? lly when there are few or no other treatment Winn, A. Re N., sea N. rc Lh I . Kns eattit ing, ute, J. AG pr. il 201 Trogd 4on, ). ht E tp . s: M. //B waw sc wh, .eb ar nd i.or S. g/h B ???????? . ea Dlt us h/ ep tzin ublic a. a20 tio18 ns. /i"ss Sp ue ending -briefs b /c yont Com ent m /r ee rc fe ia rl I enc ns eur - ers Outpatient Inpatient Pharmacy ???????????????????? = respect to use of specialty medications for the reasons given ab ?? ove. Employers and insurers would need to think of per-member, per-year (PMPY) potential savings — that is, the savings spread across all plan enrollees. ?? Source: Authors’ analysis of IBM MarketScan administrative enrollment and claims data. options available. Some of the benefits of specialty medications include the reduction of the number of relapses; on Chemotherapy Based on Site of Care." JAMA Oncology E-1, E-2. Source: Authors’ analysis of IBM MarketScan administrative enrollment and claims data. e e e e e e e e e e e eb b b b b b b b b b b br r r r r r r r r r r ri. i. i. i. i. i. i. i. i. i. i. i.o o o o o o o o o o o or r r r r r r r r r r rg g g g g g g g g g g g IIIIIIIIIIIIs s s s s s s s s s s ss s s s s s s s s s s su u u u u u u u u u u ue e e e e e e e e e e e B B B B B B B B B B B Br r r r r r r r r r r rief ief ief ief ief ief ief ief ief ief ief ief • • • • • • • • • • • • A A A A A A A A A A A Au u u u u u u u u u u ug g g g g g g g g g g gu u u u u u u u u u u us s s s s s s s s s s stttttttttttt 19 19 19 19 19 19 19 19 19 19 19 19,,,,,,,,,,,, 2 2 2 2 2 2 2 2 2 2 2 20 0 0 0 0 0 0 0 0 0 0 02 2 2 2 2 2 2 2 2 2 2 21 1 1 1 1 1 1 1 1 1 1 1 • • • • • • • • • • • • N N N N N N N N N N N No o o o o o o o o o o o............ 5 5 5 5 5 5 5 5 5 5 5 53 3 3 3 3 3 3 3 3 3 3 36 6 6 6 6 6 6 6 6 6 6 6 13 10 12 11 8 9 3 2 6 5 4 7