Previous research demonstrates that payments from third-party payers for infused oncology medicines are higher when care is provided in hospital outpatient departments (HOPDs) compared with physician offices (POs). Some have speculated this is due to differences in patient characteristics and treatment regimens between the two sites of care. This study employed a novel analytical approach that distinguishes differences in the cost of drugs due to price alone from differences attributable to drug mix and treatment intensity for cancer patients. The analysis was based on 18,195 users of the top 37 infused oncology drugs prescribed to employment-based and commercially insured patients in 2016.
Key findings:
- Hospital prices for the top 37 infused cancer drugs averaged 86.2 percent more per unit than in physician offices.
- For every drug examined, HOPDs charged more on average with statistically significant relative differences ranging from 128.3 percent (nivolumab) to 428.0 percent (fluorouracil).
- The mean annual reimbursement to providers per user of infused cancer drugs was $13,128 in POs and $21,881 in HOPDs.
- Had hospital unit prices matched physician office prices, holding drug mix and treatment intensity constant, we estimate that commercial insurers would have saved $9,766 per user of these medicines in 2016, a savings of 45 percent.
Our findings have implications for private third-party payers, including employers and commercial insurers. To counter higher HOPD pricing, employers can aim to negotiate contracts with hospitals for site-neutral payments to ensure that costs for the same treatment are not higher in the HOPD relative to the PO. In the absence of countervailing market power, third-party payers can engage cancer patients through plan design to guide them to less costly sites that are clinically appropriate for their care. Insurers use both value-based insurance design (VBID) and reference pricing to vary patient cost-sharing based on the choices that they make regarding use of health care services. However, one thing to consider is whether cancer patients receiving oncology services will be sensitive to cost-sharing, since they are some of the highest-cost claimants. They not only are more likely than the average person to reach their deductible, they are also more likely to reach their out-of-pocket (OOP) maximum. Hence, higher patient cost-sharing may not be effective unless clinically appropriate VBID or reference pricing tools remain in force for patients who exceed OOP maximums.
Figure 1 Hypothetical Example Showing How Differences iF ni g Pa ure y m 2 ents for Infused Chemotherapy Drugs in Hospital Fi Fi Fi of Study Ref Medi g g goncolo ure ure ure reas car ere 1 4 3, e on, w s prese Li gy n Diffe Pay ho ces mitati pr w ment Adv s e rences i actices nts h co ow a nd hy ons uni ucte (n Trea pothet Alpert iso t p d ry Com ri the fo ces tm , Hei ical ent for t mis two llowi , a Int nd Jac he 37 s-ion. d nensit rug g val Rep obso drug example o y of idit ort y checks s n th to th var 2 e Top 018) ief d e C F ho (N 37 Infus . Firs b iF g Fy w ongress iu ig g ikpay, Ri re s u u th ite re re t, we 4 es 3 5 of e ed care. : Me e revie cha Che leme di r ds mo care we nts For , therapy d drug and combin e and He very Penson 2 - s D pro e pe alth rtug c o dific uct ex s Car 01 comm Adm pl , HOP 7) e ain D has s ini the eliv Ds c on ste ery Sys s idosages as red gn harge ource o ificantly to Employ tem. d f dif mo incr rfer e o ment eas ences n ed - Paul Fronstin is Director of the Health Research and Education Program at the Employee Benefit Research Institute Location, Location, Location: Cost Differences for Oncology Outpatient Departments and Physician Offices Can Be Partitioned by Drug Mix, Treatment Intensity, and Differences in Mix Among the Top 37 Infused Chemotherapy Drugs Administered to Employment-Based and Commercially Differences in Total Payments for the Top 37 Infused Chemotherapy Drugs Administered to Employment- average in the ov barg recom eral,a lW wit ini Bas pay mend as n h dif hin g ed ments p gton, ed and ow fere ier o i n pack Co nce nDC: Me the f mm sHO t ra aercially I wo ge P ng Ds dp in in sac, ites o g from vis serts -20 nsured à- to i f 1 vis car 7. 128.3 d th eentif Can . ird The u -pe party y p cer P rclausi ent pp atients paye er o bl f the pan ers ran i com n el mean P ge shows hys s pare for P ician O e d hac with ow charge O h d ffdic t rug rug hat o es a for paym ac nd f niv cor i Hos nolumab de ent ding pps en ital are to patie de O to nt calcu utpatie on 428. nt cologist lated. 0wei nt D percent gh ep T st or . he artm for b mo ix ents dy of , (EBRI). M. Christopher Roebuck is president and CEO of RxEconomics, LLC. Bruce C. Stuart is a Professor in the The Alpert, A., Dre iffe are re sev H. nce Hei, an eral s ins Pr tud dy ic Mir elimi p ell etat e r Jaco U ions that nit bson. for th " we E arrant me valuating Top 37 ntio I th nfu e Ro n. se d Fi le rs C ot, th h f e Paym m e os th tudy f ee nt Poli rapoycus cy Dr i e u n dg D s solely on ri A ving dmin Integ i s inte fuse rati red d on i to canc n Em the er pdru On loyg m cs ology e an nt- d Differences in Treatment Intensity of the Top 37 Infused Chemotherapy Drugs Administered to Employment- Medicines Based on Site of Treatment Insu red Cancer Patients in Physician OU ffi n cie t s Pr an ic de Hospital Outpatient Departments, 2016 Based and Commercially Insured Cancer Patients in Physician Offices and Hospital Outpatient Departments, drug fluoro s surface p urac rov 20 ili. de 1 a Ev 6 rea. d ................................ ery d in HOP App iffDs fa rox erenc imately vo e rs was the 7 stat 8 more percen ................................ istically sig ex t of pe i nsiv nf nused ifi eca drnt oncol ug at p<0. (a) ................................ ogy claims by 7 05 5 ex pecept for rcent with to positiv th 25 e ................................ p light e e i rcent nt ly p ege , resc whe rs in th ribe read s e ibren u n ni POs ...................... tstu f the xima ield drug h bad mi 9 x payment School o B Mark and asef d et." di P a har dn not H d mac ealt Cfocus oy at Unive h mAf m fairs e on rcpati i a 3lrs 6, n ly ent ity of Ino. 4 cha sur Mary e (ract 2 d018) C la erist a nd n : 680 c , ics e Baltimore r, which m Pa –688. tien .ts Thi eans th ins Ph Iss yue at the sicBr iain ef results O was w ffice cannot s ritten and w H be ith o s use ass pita d ist lto as O anc utp e sess a from tie qu n the t alitative Based and Commercially Insured Cancer Patients in Physician Offices and Hospital Outpatient Departments, Drug User Mix by Site of Care Percentage Schef Evidence fler, i R ndi ., cand ates D. Arno that ther ld.e "Insur are several er Mark drivin et Po g wer facto Lorw s ers behin Prices d this in tr Numero end of c us Conc onsolidation entrate . Firs d Prov t, until 20 ider Ma15 rket , M s." ediHealth care January 16, 2020 • No. 498 2016 By Paul Fronstin, Ph.D., Employee Benefit Research Institute; M. Christopher Roebuck, Ph.D., is 60 vedotin. pl pausib ercent Over le all, fva or d lu the es base ru g w(eight a) a d n on ed d t 4 av his criterio 0 erag perce ent un n fit p or d (rric esults ru e g was ( n b). ot s 8 Treat 6.2 ho wn). percent ment inte higher nsity in is also HOPDs hi. gher in HOPDs at 15 units vs. 13 Institute’s research and editorial staffs. Any views Depa ex rtm pre es n sts ed , i2n0 thi 16s report are those of the author and should not differences in cancer treatment between POs and HOPDs. 2016 Second, we report differences in payments for the drugs PO HOPD Difference in Panel 1. Accounting Cost Deconstruction Fi poli gure cy a 4, war Affairs Diffe ded rences i 36 hos , p no. ital n 9Pr ac (S ice qep uisition o p tem er Unit berf 2 f of or t 0f17 -cam he Top ): 1539 pus 3 ph –15 7ys Inf 4ic 6. used ian practices Chemoth with erapy extra Dru rev gs A enue in dminis the tere f d oto rm of facility Employment fees -Bas , ed Clough, J. D., M. A. Dinan, and K. A. Schulman. "Trends in Hospital-Physician Integration in Medical Oncology." uni RxEc ts fon or d omic rug (a) s, aLL ndC 9 ; vs an . d 8 for Br d uce C. rug (b). S tuart Drug (, a) Ph has .D., Uni a unit price versi of ty $ of 100 in Mar HOP yland, Bal Ds and $60 in timore POs . The prices per Percentage thems beelves an ascribed do n d to the ot of an fialy cers ze dif , trus fere tees nces , or oth in admini er sponso stratrs ion cos of EBRI, ts or Em other ploysee ervic Bene es p fit Res rovided earch I durin nst g the itute ra -py Educatio sessions n a. nd Drug Mix in Pe Pe rc re ce nn ta ta gg ee Model Variables Setting and Drug Products Fi whi gure ch Ho are 5 wever, a inte an una d grates Com vailabl bout 20 m the e erci to thre pe ally In indepen rce e el nt sements ure of de d C claims nt ancer prac of o repo tices ur Patients mo rted . Secon del a i sin sn Phy ho d, gl win there e unit dis sg ician actua has b Offi pe l m ee ces nse ean n an d despite littl adnnu e H ant os alpi itrus pa th tal Out ym e t en fa ent ct that o for pati s p cement ent er d nly rug of on user v e er dr tical by ug in s it our e of American Jour ®nal of Managed Care 23, no. 10 (2017): 624–627. unit for drug (b) are $50 and $30, respectively. In this example, the annual reimbursement paid per user o Diffefr e dr nug ce (a) in is Research Fund (EBRI-ERF), or their staffs. Neither EBRI nor EBRI-ERF lobbies or takes positions on specific policy Third, the MarketScan sample available for analysis was constrained by lack of dosage information on claims H u OsPD ing Mean Units of Drug Administered per User Difference in Difference in Price The Moran Company. "Cost Differences in HO P Cancer C D are Across Settings.PO " 2013. Location, Location, Location: Cost Differences for Oncology care tog integsrat ampl ion ether Dep e in t (BCG artm with he ihos ents ntrav counter p, ital esical) 20m 1 fa 6ark ctual ................................ had a etplac payment recomm e (Gaynor s in end HO ee t d dos P al. ................................ Ds20 — e 1 of what 4) . T 1 mill hir insurers d, igram. and would W pe ................................ rha e be ps mos hav lieve e s th p t ent important, at this had di t................................ hey screp the p e aid ancy xpan PO aros sprices ion e of fthe or 11 NumbeM r o efan Pa Pey rc m ee nn tat gp ee orf Use Nru b my b e Si r te of of C Pe arrc e entage of $1 Pr,5 oc 00 in edur HOP e Ds and $780 in POs. For drug (b), the respective payments per user are $450 and $240. Co C u W n oei m te gh pra fa ti re n cd g tu W ailth NDCproposals -based pa. EB yment RI invit metho es com ds. Fourth, ment on eve thi n s r aft esear er exclu (T ch. re a ding putativ tment Inten es N ity DC ) b -based y Site cl of aiC ms ar, the e sample contained non- Treatment Intensity in Procedure Mean Price per Unit by Site of Care per Unit in HOPD Drug (a) Drug (b) Drug (a) Drug (b) Conti, R. M., and P. B. Bach. "Cost Consequences of the 340B Drug Discount Program." Journal of the American Medical 34 the Introduc 0B e be xact sam Dcaus rug P e som tio ricing Program e mark n e prov et b ide asrs k und et of biller by d th ru N e gs a ational Affo nd rdab D treat rlug e Um s Car Co eent rs e deAc in (N tens t in DC) rather t Tity provi 2010 otal U p ser ed rmitt ed han CPT. s in edHO mor U Ps An Ds e erh sdos N uring D pi C tals spe 2 T 016 o to cifi tapu les the . UrO schas e ver the rs drug e me fu dill cin saes a mple t Code Generic Drug Name PO Medicines Based on Site of Treatment Compared With mean payment rates by the proportions of users of each drug in each site generates an average per user payment of pl Pr ausib ocele extr dure eme values that required Winsorization. These exclusion criteria may limit the gener Halizability o OPD Comp f ath ree ds tudy White, C., and C. Whaley. Prices Paid to Hospitals by Privat PO e Health Plans Are Hi HO gh PD Relative to Medicare and Vary Code Generic Drug Name Compared With PO Figure 5, Differences in Total Payments for the Top 37 Infused Chemotherapy Drugs Administered to Employment- Association 309, no. 19 (May 2013): 1995–1996. of sigdr nif Jmanufactu 9 ug ic 26 antly s7, actdis Pa ual rer c clounted paym ita( xlab el ents eler), pric ae verage dru s (Desai an g, and p d $13ack d Mc ,128 age W in POs ill size. T iams 801 and 20 hus, a 18 $2 )1,8 (n Conti 81 N 8DC . 6in an % -bi Hd OPDs ll B ed ach claim , a differ 2013 867 shoul ) (ence Gd AO 20 coo rres f $ 18, 8) pond to t 9753 .7 (Jung et a %. Ha hd payers e num l. 21 018) 1 be 2 .r5 of . % PO HOPD HOPD TPr here oc h ed as u rbe e en a marked change® in site of treatment for cancer patients in the U.S. over the past 15A ycears tual. Pa In 200 ymen 4, ts $1,238 in HOPDs vs. $564 in POs, a difference of $674. Nu C m ob d ee r (%) o G f e Un se err si c Drug Name PO HOPD With PO findings. Finally, the MarketScan database, while very large, is a convenience sample and not necessarily SugW ges idely: ted Fi Cinta diti ngs f on:rom Paul Frons an Employer tin M. Chr -Ledist Topher Roe ransparency bu I ck nitiative. , and Bru Sant ce C. St a Monica, uart, CA: Rand “Location, Corporat Location, Loca ion, 201 tion: 9. J9267 Paclitaxel $0.76 $1.40 184.3% By Paul Bas Fro ed nsti and n Co , Ph mm .D., ercially I Empl nsured oyee Ben Cancer P efit Rese atients iarch n PhysIn ician stitu Offte ices a ; M. ndC Hos hristo pital pher Outpatie Roebu nt Dep ck artm , Ph. ents D., , J9045 Carboplatin 785 8.5% 858 9.6% 113.6% reimburse Since uni the ts i sd e nc d H luded iOPD scounts u in nit price that are n pack ot availab s pet revail (e.g., ing in le 1 to vial i the ndep (N PO endent DC) may market ph equa — ys ician wl hich 1 p 00 rac aver mtices g ag (unit ed , 3 s4 $ )0B 12 ). p ,Unfort 11 rov 5 id ov es an un er a ately, w ll extr drugs a i e h nce invest adntive no w igate fay or d to — the (Actual) (Actual) (Counterfactual) app Pro rC es o ximately cd rie bed Dr 9 ug 4* p (D ercent G rue gn e rof ic che Drumo g N th aerap me y infusions for employment-based or commercially insured patie innts HO wPD ere s J9267 Paclitaxel 1,164.0 1,105.8 95.0% * represe Cosntativ t Differe e of nc aes for O ll employ ncolog menty Medi -basedc a ines nd com Based mercial on Si t healt e of h plans. Treatment ,” EBRI Issue Brief, no. 498 (Employee Benefit J9045 Carboplatin $23.48 $32.98 140.5% Desai, S., and J. M. McWilliams. "Consequences of the 340B Drug Pricing Program." New England Journal of Medicine RxEconomic 2016 s, ................................ LLC; and Bruce C. 75 ................................ Stuart 25 , Ph.D. ® , Uni ................................ 60 versity of Mar 40yland ................................ , Baltimore .................... 12 J9070 Cyclophosphamide 700 7.5% 765 8.6% 113.6% Mix) oncologist mean The botto ind paymen epe m sn to de panel t p s ntly ell er verif the o u f s Fi ir er gy pract w ure thi ou s 1 ices to hos ld co uses hav nject e drop a cou ure as pinterfactual ta ped M ls. ark by ThetScan ese $9,7 simul f66 acto or 45 p do ation rs es not r com tec ercent. bihni necord e que to to Nform ote that the de a basis mo per the nst of fect d r pay ate ho reci rugment. -pe mw ix for - me wei W incr an gh e reim teas te hus ex d ing p mea burs cluded n ric eme e nt all admini J92 s6 ter 7 ed in phy Paclits aician xel offices (POs), but by 2014 that pe $884 rcentage had drop $1,54 pe 8d to 57 percent $840 with a correspon 18 di 4ng .3% Winn, A. N., N. L. Keating, J. G. Trogdon, E. M. Basch, and S. B. Dusetzina. "Spending by Commercial Insurers on J9045 Carboplatin 46.0 44.2 96.0% Research Institute, January 16, 2020). J90730 0 9, nC o. yc 19 ( loph 2o 01 s8): pha539 mid– e548. $60.85 $120.80 198.5% J9310 Rituximab -75% -2617 5% -6 60 .7 % % -40 577 % 6.5% 97.2% rate counter disps ariti patien wou fa es ctu ld b ts in c al with e rei ancer tr af infused fecte mburs eatme d eme ( oncolo 1) if nt nt b u to nit prices y gy site o HO claims PDs is f c fr are repo om jus , furt P t rting 1s in Os 7.3her were perce addin app th nt e g low li u to t ed nit er to th he gr f th ield (r an e owing cos drug mix the etaini actu ng patient al and t bur mean treat den paym s shoul p ment resent cri de inte be rm ed d nsity ade B by CG to th foun e POs Ud in .,S . health shift tow J904ard 5 hos Ca pi rb tal out oplatipati n ent departments (HOPDs) (W $inn 1,0 e 80 t al. 2018). $ Duri 1,45ng this s 6 ame pe $1rio ,03d7, the difference i 140n .5 % Chemotherapy Based on Site of Care." JAMA Oncology, February 2018: E-1, E-2. J9070 Cyclophosphamide 49.2 45.0 91.5% * M Je 9a 1n 9 0 UnitsF p lu eo r ro Us ue ra r cil 607 6.5% 482 5.4% 82.6% J9310 Rituximab $870.19 $1,463.65 168.2% A T A G L A N C E 15 9 13 8 care sy HO indicati PDs intraves , ng that stem. (2) if ical). the The tren modrug mix s t of d the is l fo o ikely to cont bserv und in ed POs diffin er w ue ence ere unless price was d vi du go ate rous HO to pric PD ly chall paymen e. Th ene r ge t rates emaind d. Disc with er i usssions H a OPD ttrabout ibuta treatmen bl h eea to lth t in hig care tens her o cost ity, bserv and (3) i drivers ed f average Conclusio J C 9o 0pyright I 7 cos 0 t tn o comm C yc nlfor ophma ercial osp ti ho a carriers n: mi d Te his re for port is c chemot opyright herapy ed admini $ b 2y ,9the 9 stere 2 Em d in ployee HOPDs $ Bene 5,43 vs 7fit Res . POs b ear alloon ch In $ed 2s,titut 7 fro 39m 25 e (EBRI) percen . You 1 t 9 may higher 8.5% J9310 Rituximab 32.2 29.0 89.9% * Fis(her, TreatM. me D nt ., In R. P tensiu tynekar, ) Y. M. Yim, A. Small, J. R. Singer et al. "Differences in Health Care Use and Cost Among J9171 Docetaxel 582 6.3% 558 6.3% 99.7% J9190 Fluorouracil $3.71 $15.87 428.0% treat s treat hould ment ment consid inte inteer the nsity nsity i f role n POs oun tha d in POs . t hospitals were pric play in ed at cost HOPD grow ra th f tesor w p ith hys dicia rug nmix -adm from inis HOP tered drug Ds. Tshe f . Poten irst tial counterfa solutions ctual includ e in 2004 to 42 percent higher in 2014 (Fitch, Pelizzari, and Pyenson 2016). Various studies have attempted to ascertain J J copy, pr 9 91 39 10 0 int, or F R lu ito uro x d im u ow ra ab c nloa il d this report solely for persona $28 l ,an 05d 9no 63 ncom .2 mer $42cial ,428 use, provi5 de 8.$ d 92 that 5,22 al 5 l hard copies r 93 1.6 3 etai 8 % .2% n Unit Price $100 $50 $60 $30 Our f J9i3 nd 55 iPatients ngs T dra emons s With tuzum trat a Cancer be that p Rece aym ivi eng nts I fo ntrav r inf enou used 538 s Ch canemot cer m her edi 5 a.cin py 8%es in P in hys the icia com ns Offi 523 mercial ces V an er d semplo us Hos 5.9% ym pital O ent-b utpatie as 10 ed 1. 1nt % Previous J After r 917 rese 1 emov aD rch in oc g d eemons s ta in xgl ele-tra unites t dos that es,paym we stil ents l observ fromed thi n rd on -party -plausib $ payers 8.7 le 6va for lues at infus both ext ed on $cology 24 remes .34 me ofdi the cine uni s are ts dis hi2 gh tributi 77 er .9 w % on hen repl greater aces HOP price D trans unit paren prices cy with and co thos untervai e found ling in m POs ark . et Th po e wer exer second a cis nd ed third co by thirunter d-party factuals payers are (Sch compute effler an d id n Ar a similar nold wheJ th any an 9er th 190 e hi d alF gh l lco uer opyrig ro pu aym raht ci and ot e l nts mad her e to app HO licab PDs are le not ass ices c ociated ontai $234 ned ther with mor ein, a e com nd $935 p you lex p may c atients ite or , greater qu $218 ote reso small p urce u ort se, ions 4 o 2r 8 of .0% JJ 9 9 01 07 01 DD ox o oc ru eb ta icxie nl 535 5 5.1 82 % .9 600 516.3 6.7% 11 10 60 .6.% 7% markets are nearly two times higher, on average, in the HOPD relative to the PO for the same drug and that these cost Settings." Journal of Oncology Practice 13, no. 1 (2017): e37–e46. care is provided in hospital outpatient departments (HOPDs) compared with physician offices (POs). Some have J including sm 9355 Tra alsl tinte uzuge ma rs b (e.g., 2s and 3s) and large volumes $10 many tim 4.39 es the norm $17 al 4 .dosage 98 range. We remove 167.6% d Mean Payment per User 20 fashion 17). with drug mix and treatment intensity substitutions. high Jer 92-6cos 3 t dr Oug xals ip . The latin consensus is that patient 447 mix and treatm 4.ent 8% modalities are 361 basically similar 4.0% across sites o 84.0f %care J J the 9 93 15 7re 5 1port T D pro ra oc svided that te utz au xe m lab you do so verbatim and with pr $4,4 oper 9140 c7 itation. .1 $1 Any use 2,565 beyond 36 t 3.he 8$4s ,5 co 2pe 1 of the foreg 89 2oi .7 47 ng % .9 % * differences are due to pricing decisions of hospitals, not differences in modality. On a drug-by-drug basis, HOPDs st th $1,500 $450 $780 $240 Endnotes speculate these de txtreme hi s is d value ue to sdiff by erences Winsorizing the in patient clch aims aract data eris at tics the 1 and tr an eat d 99 ment pe rerc gientil mene s b of etw the drug een the -s two pecifs ic ites of units c ’ are. Disc J9ussio 000 n D oxorubicin $8.75 $29.40 335.9% J9035 Bevacizumab 367 4.0% 380 4.3% 107.7% Fitch, K., P. M. Pelizzari, and B. Pyenson. Cost Drivers of Cancer Care: A Retrospective Analysis of Medicare and but th J J requ 9 9at 0 30 50 payment rat 5 ires EBRI D To ra xs o ’stru p uz es are b rior u icm ina ex b much h press permis igher in hospita sion. For pe l srmis etting $s 4io 2 sns, , 4 whe 9 p 8 lease th 40 er .4care is con $tact 63de ,6 EBRI at 5 liver 9 ed o pn ermis 3 hos 7.$ 13 p s7 ions@ ital ,97 -base 6 ebri.org d cam . pus 91 es or 1.6 87 % .6% * charged 1.3 to 4.3 times more than POs for cancer medicines. Over one year, employers and insurers could save This study employed a novel analytical approach that distinguishes differences in the cost of drugs due to price alone The res distributio ults frons. m the As f w irst ith co siu ng nterfactu le-unit val al ues, simulation su we exclugges ded al t th l patien at insure ts with rs could re any Winso duce riz reim ed cla buim. T rsement his e by xercise $495 per drug J9263 Oxaliplatin $1.90 $4.08 214.9% Our an J906 a0 lysis of Cisp playm atin ents for infused oncology dr355 ugs by third-party 3.8% payers in 20431 16 demonstrates 4.8 % that large dif 12 fer 6.2 ences % freeJ -s 9tand 263 Com ing mercially O hos xapi lip tal lat-io Insured nwned cli Pnics opul ation Cla (Fisher et im D al. 2 a017) ta 2004 (Fi-tch, P 2014. 1,8e 6 liz Milli 2.zar 4ma i, n, and 2 P 016. yenso n 1 2,01 666) 9.2 (Gordon et al. 2018 89 ) (W .6%inn * J9000 Doxorubicin $353 $1,090 $324 335.9% Mean Payment Across All $9,766 per covered cancer patient if they paid PO prices rather than HOPD prices for infused cancer therapy. from 1 user p redu der yea ifferenc ced the r ( es a sample 73 att rib peutabl rc b ent y e an s to avings a dr ddug itio ) i mix f PO nal 7 and prices r pe tr rcent. eatment ep F laice nall intens d y, HO to confi Pity D prices forne ca,t ncer he analys assuming patients is to p no. change Th atients e an i alys n eit treat is her dru was b ed exclus as g e mi d ively x o on r in Imp JRepo 92li 1ca 7 rt tio A Lev ns f uai pro la lid bil or E e ity:mployers This report isand availaInsurers ble324 on the inte rnet at 3.5% www.ebri.org 60 0.7% 19.3% J9035 Bevacizumab $82.36 $208.32 252.9% in Sre ee T imbursem ruven Health An ent rates alytics. to HOPDs MarketS and P can Database Os were primaril s. httpsy :// du truv e to enh diealth fferences .com/ mark in price ets/ sl rather ife-sciences/pro than difducts/data ferences in - Drugs et al. 20 J90318 5 ). Ho Bewever vacizu,m data ab limitations constrain researchers’ abi 52 lity 9.7 to match cancer pati 42 ent 6.7characteristics acr8 os 0.s 5 % sites * J9263 Oxaliplatin $3,534 $6,805 $3,167 214.9% 18,1 J95 920 users 1 G of emthe cita top bine 37 infused oncology drug323 s prescribed to 3 .emp 5% loyment-bas 317 ed and comm 3erc .6%ially insured 10 patients 2.0% treath ment ospitals inte and nsity p.hys Simi icians’ larly, s offetting ices, w HO e d Pro D pped patie drug mix and nts who rec intensity eiv toed PO l oncology t evels reshera ults py in proj in other ected ou savings o tpatient settin f 23 gs Fronstin, Paul, and M. Christopher Roebu $1c ,2 k, 38“ Persistency in High-Cost $564 Health Care Claims: ‘It’s Where the Spending Is, treat Our tools J fi /marketscan ment mo 9 nd 06 ing 0 s hdal ave imp C- iity s dat p. la aba tW ilicatio ne ses fou ns for nd that th private e av third erage -party emplpa oyyers ment , - includin based $6.5 2 or g employ commercial ers and i ly $insured 1nsur 5.87ers pat . E ien mployers t receivi can ng onc 24 3 exert .4ology % Given that nearly half of oncology therapy takes place in HOPDs, employers could cut their drug costs nearly in half of care J90.6 0 NoneC of is t plhe atin secondary data sources used in these prior 4 stu 1.8 dies included informat 42 ion .7 on cancer stage, 102.1% J9035 Bevacizumab $43,630 $88,888 $35,143 252.9% J9031 BCG (intravesical) 279 3.0% 33 0.4% 12.3% in perce 2016. ( Thi e.g., nts an study in d 2 -home 4 w pa ercent s) cond or who , res uct rec ed pe t ctiv ei hrough the ved ely.tr eatment Note EB that RI in both the Center se es a phy for timat Researc sicians’ es are n hof on ot ad fice H ealth Ben and ditive a h be osefi caus pital ts e Innovat dany uring the cha ion ng ye (es i EBRI ar (less n on CRH e el than BI) ement , 2 with Stupid,’” EBRI Issue Brief, no. 493 (Employee Benefit Research Institute, October 24, 2019). medici press J9 ure nes 217 o n b cosoth hea t L$ e9, up 766 rolth pla li d mo e rens and hos per year wh pita en ls to shift treated f in rom HOPDs disc ount $vs 36 . 8 POs ed .2 ch 9. arge co Our find ntract ings$ ind 8 s b 17 icat as .5e 1d e that on a h m igultip her p le aym of Me 2ent 22dic .0 s % for are 2 simply J92by 17 shiftLing eu ppati rolid ents e to PO settings without necessarily affectin 8.0g quality of care. They 8 .co 0 uld also negotiate 10 site 0.4% - tolerance for alternative treatments, survival prognosis, and other patient-specific factors that could influence drug See J9 http:// 060 vbidcenter.o Cisplatin rg/frequently-asked-questions/ $273 $678 $279 243.4% J9206 Irinotecan 227 2.4% 221 2.5% 101.2% the funding support of the following organizations: Aon Hewitt, Blue Cross Blue Shield Association, ICUBA, JP will af pe fect t rcent of he es the timat sample). ed imp act of changes in the others. Our simple counterfactual model does not include any or so infuse Diffme othe ed reonco nce in lo r M p gy e ros adr n p ug ective c s in HOPD ase rates s relat . Suc ive to PO h a co s alition are a rof esult of employ hiers gher is p al ric ready es pa tryin id tog hos thispi (tals Koller relat and ive to Khul prices lar 201 p9) aid . to In J9201 Gemcitabine $38.54 $69.89 181.4% neutral J92 p 01 ricing fo Gem r m cited abic inines e . 68.6 54.0 78.7% * choice and cost. J9217 Leuprolide $2,934 $6,540 $2,946 222.0% J9306 Pertuzumab 211 2.3% 243 2.7% 119.7% Key f Mor indin gangs: Chase , Mercer, Milliman, Pfizer, and PhRMA. Fronstin, Paul, and M. Christopher Roebuck. "Reference Pricing for Health Care Services: A New Twist on the Defined Payment by Site of Care be havioral assumptions, but they could be added in more sophisticated versions. POs the f a Jbsence 9 or t 03he 1 s o af me BC such G ag (i ents mark ntra . v et espowe ical) r, third-party payers can att$ em 15pt 3.5 to 7 engage patient $3s8 throug 3.90 h increased pric2 e 50.0% J9031 BCG (intravesical) 5.5 5.5 99.5% JJ 99 22 90 91 NiG voe lu m mc aib tabine 165 $2,6441.8% $3,775 197 $2 2,.0 2% 82 1241 .18 % 1.4% Contribution Concept in Employment-Based Health Benefits." EBRI Issue Brief, no. 398 (Em ployee Benefit $674 trans In this paren study cy, com we exami bined with ne how plap n desig ayments n c from thir hanges to s d-par teer the ty payers m to les for is nfu cost sed ly sites oncolo ogy f care me di for cin treat es diment that ffer by site i o s cl f ini care cally J9206 Irinotecan $23.77 $46.91 197.3% J9206 Irinotecan 96.4 95.3 98.9% J9181 Etoposide 153 1.6% 145 1.6% 98.5% • Hospital prices for the top 37 infused cancer drugs averaged 86.2 percent more per unit than in physician J9031 BCG (intravesical) $847 $2,106 $842 250.0% Our empirical analyses use the deconstruction/simulation approach with tables showing, for each site of care, counts of Research Institute, April 2014). Our Study finding Ds es bols ign ter prior research showing that payments for infused oncology drugs in HOPDs exceeded payments in Panel 2. Counterfactual Simulations for a market basket of identical medicines. Using a novel analytical approach, we distinguish between differences in appropriate. Employers and insurers can use both value-based insurance design (VBID) and reference pricing to vary JJ J 9 9 9 23 3 60 0 46 6 Pa Pe Pe clirt trt a u x u z ez u l u m pm ro ata b eb in-bound particles 146 2 $,1 1 9.2 3 65 % .4 .7 2 129 $2 20 ,6 .6 28 5.5 1.4% 1 96 1 8 6 .9 9 .5 % .4 % % * offices. J9206 Irinotecan $2,291 $4,470 $2,265 197.3% Table of Contents cancer patients by drug, mean units administered per patient per year, mean payment amounts per unit of each drug, POs by a factor of 1.71 (Gordon et al. 2018) to 2.59 (Winn et al. 2018) between 2004 and 20 HO 14 PD . W Similar dif ith PO Tre fer atm ences ent Unlike previous studies, we made no attempt to directly control for differences in characteristics of cancer patients HO patien PD J93 tan 0 cost 5d PO -s Pe haring ba oncology metrexedsed med on the icine choic costs es that t due to phey ma rice 145 differ ke re ences i gard 1n drug i.ng u 6% se mix of health and 165 annu care services al dosage 1 ..leve Under VBI 8% ls. The D, c m 1etho 18 os .3t% d - J9299 Nivolumab 1,704.4 1,397.9 82.0% J9299 Nivolumab HOPD With PO Unit Price H$ O3 P9 D . 1 W 0ith PO Drug Mix $50.16 128.3% J9306 Pertuzumab $36,455 $54,296 $32,603 166.5% Fulton, B. D. "Health Care Market Concentration Trends in The United States: Evidence and Policy Responses." Health Introduction .......................................................................................................................................................... 4 and mean actual/counterfactual payment amounts per patient per year. For clarity of presentation, Inw tee nshav ity e not shown have been reported in Medicare Part B payments for infused oncology products (The Moran Company 2013). These treated in HOPDs and POs. Rather, we started by building an accounting cost deconstruction model that isolates allows us J9041 to e Bo stimat rtezom e ib the aggregate extra spending 138in HOPDs that 1.5 i % s unrelated to 138 patient care with 1.5%out having 1to 04 .di 0% rectly s haring is alig J •9 181 For ev ned wi Et ery d opos ru tih dg eth ee xamin value edo,f health HOPDs char care serv ged more o ices receive n aver d. It i 1age with 98.5s bu stat ilt on isttically sig he princip 1nif 6l6 e ic .8 o af nt lorelat wering or remov ive differences 84i.ng 0% * J9181 Etoposide $1.94 $5.56 286.1% J9299 Nivolumab $66,643 $70,115 $54,660 128.3% Affairs 36, no. 9 (Septem Drube g (r a 2 ) 017): 1530 Drug– (15 b) 38. Drug (a) Drug (b) Drug (a) Drug (B) measures of variance for mean values. Statistically significant HOPD/PO differences at p<0.05 are indicated with studies and others measuring differences in cancer-related 2 medical costs by site of care (Fisher et al. 2017) (Fitch, elements Methods J9370 ................................ of o Vncology incristine claims pay ................................ ments in both sites o 131 ................................ f care. Us1 in .4 g %this model ................................ , we 219 isolated the2 e .5 xt % ............................... ent to whic1 h 73.8% 4 acco financi unt for al barri rangi pa ers ntient c g from to eharacter s s 1entia 28.3 perce l, ist hiics gh. nt -va W (lue nivolumab) e pro clini vid cal sol e est to imates utions 428.0. of pe To potent rce the de nt (ial flu gre oro savings to e uthat raci the l). t hird qua -party lity of payers health and care services patients does using a J9264 Paclitaxel protein-bound particles 2,051.9 1,492.9 72.8% * J9264 Paclitaxel protein-bound particles $11.69 $21.60 184.7% Un Ji9 t 1 P8 ri1 ce Etoposide $60 $30 $386 $100 $927 $50 $100 $324 $50 286.1% asterisks (*). Peliz J9 zari, 395 andF u Pyens lvestra on nt 2016) (Kalidindi, Jung, and Fe 118 ldman 2018) 1.3 all % use some var 103iant of regres 1.2 sion % modeling 90 to .8 contr % ol differences are driven by price vs. other factors. The model comprises three variables, each computed separately by combination of accounting cost deconstruction and counterfactual simulation. not vary J930 5 with s Pe ite mof etre care, xed VBID could be used to encourage pati3 ents 94.1 to seek treatment i3 n 51 POs .0 by lowering cost8 -s 9haring .1% , Gaynor Data Sou , M., K. rce Ho, and St and ud R. y Sam Town. The ple ................................ Industrial Organizati ................................ on of Health Care Markets ................................ . NBER working p ........................... aper no. 19800, 4 J9305 Pemetrexed $77.34 $130.99 169.4% J9264 Paclitaxel protein-bound particles $23,994 $32,250 $17,458 184.7% J9033 Bendamustine 84 0.9% 62 0.7% 76.7% for p •ot ential The m disean similarities a annual remo imburse ng cancer ment to patien provi ts tr de eate rs pe d r in user HOP of Ds an infuse d POs d can . cer d As we rug noted s was e $ arli 13,1 er, 28 none in Po Os f th an e d s ite of care: relative to J9041 HOBo PDs, rte w zo hen mib services are received at that site of care. 52 3.8 443.5 84.7% Number (%) of Users Cambridge MA: National Bureau of Economic Research, 2014. Meas J904 uring Do 1 Bo se rte for I zom nfu ib sed Oncology Agents ................................ $54.10 ................................ $98.09 ................................ 181.3...... % 4 J9305 Pemetrexed $30,478 $45,985 $27,149 169.4% J9040 Bleomycin Sulfate 73 0.8% 55 0.6% 78.3% secondary $21 dat ,881 a sources in HOPDs. use d in these prior studies included information on cancer stage, tolerance for alternative Prescribed Drug* (Drug J9370 Vincristine 8.5 10.5 124.6% Result J9370 s Vincristine $13.46 $34.87 259.0% JJ 99 10 34 01 Da Bo cart rb eazzo in m eib 75 25 60 $28,3340 60 .6% $43,5040 6 53 $23 0 75 ,.9 6% 94 9 25 1.1 8% 81.3% Study Design ......................................................................................................................................................... 5 1. Drug mix: the proportion of users receiving each drug (note that patients taking more than one drug are Mix) treat Similarly, r mentsefe , su rence prici rvival progn ng r osaises is, and cost ot- her sha patient ring wh -en pati specific ents factor seek s that c careould at cert influenc ain hea e drug lth care choic prov e a id nd ers cost wh.ere the Methods Government Accountability Office. Drug Discount Program: Characteristics of Hospitals Participating and Not J9395 Fulvestrant 132.7 125.5 94.6% J9055 Cetuximab 58 0.6% 43 0.5% 77.1% J J9 93 37 90 5 V Fiu nlc vri es sttira nent -75% -25% $114 $1 -60 09 % .86 $367 -40% $185.72 -7$142 5% -1 25 6% 92 .1 5% 9.0% The final • Had stud hos y sam pital u ple nit inclpr ude iceds 1 mat 8,19 ched 5 users phys of ici tahe n of tofice p 3 p 7 ric infe use s, hold d oning cology drug drug mix s an prescrib d treated ment in to employment tensity const -bas ant ed , we counted as users of each drug taken). quality or outcome of treatment is not dependent on the site of care. Under reference pricing, employers or insurers Results ................................................................................................................................................................. 7 J9033 Participat Bend ing i amu n sthe tine 340B Program. GAO-18-521R Drug 1,Disco 168.0unt Program, June 1, 201 164.8. 1 99.7% J9179 Eribulin 52 0.6% 53 0.6% 106.0% J J9 93 09 35 3 F Be ulv ne ds atm rau ns t tine $14,580 $29.41 $23,314 $54.98$13,791 1871 .0 6% 9.1% and The data comm est uerci simate that edally i for tnsur his st com ed ud p mercial y a tients also lacked i in ns 201 urers i6, nf w orm of oul whom ation on d have sav 51 t p hese varia ercent ed $9 ,766 wer ble es ptr er u . eated So s i er nst i of nea Pthes Os d, a we focus end me 4dic 9 pe ines ed rcent ein xclusiv 201 in HO 6ely on , P a sav Ds. ing Figsure of 2. Treatment intensity: the mean number of units of each drug administered annually to each user (which in pay a D f at ixed a a Source mount and or limit St the udy ir contri Sampl but e ions toward the cost of a specific health care service, and health plan J9040 Bleomycin Sulfate 11.8 11.3 95.3% J9271 Pembrolizumab 41 0.4% 52 0.6% 131.9% Discussion ........................................................................................................................................................... 10 J9040 Bleomycin Sulfate $63.55 $94.57 148.8% MJ e9 a0 n 3 U 3 nits pBe er U ns de a rm ustine $34,351 $64,008 $34,236 187.0% observ 2 shows t ed45 d he r pe ug treat mix rcent. of u ments sers b a y site of nd payments care. . Th On e drugs the treat are ment sid listed in de e, we creas expl ing orde icitly cons r of iduse ered in differ POs.en Th ces e dist in dru ribu g tsiel ons ection were turn can be decomposed into number of therapy sessions and units infused per session). Gordon, L., M. Blazer, V. Soundankar, D. Kazzaz, S. Weidner, and M. Eaddy. "Cost Differences Associated With members must pay the difference in price if a more costly health care provider or service is selected (Fronstin and JJ 9 9 31 03 80 RD am ac ua crb irua m za in be 37 0.6 40 % .1 20 56.0 0.2% 56 9.3 2% .2% This study analyzes data from a sample of over 1.7 million members under age 65 enrolled in employment-based (Treatment Intensity) J J9 90 14 30 0 Bl De ao cm arb yc aiz ni n Su e lfate $750$6.81 $1,064 $20.39 $715 2991 .5 4% 8.8% and Implications heavily co treatment nce fo ntrated, i r Employers ntensity w acro ith tand he ss the to Insurers p s two even s ................................ et drug tings s accounti , which together ng for ................................ a serve pproxi mat as surrogate ely half o f all u meas ................................ ures for sers in ea dich ffere site nces i o ................ f car n patient e. By 13 3. Oncology C Unit price:are the D mean elivere pd aym in a e Community nt amount (pl San etting + memb Versus er) a Hos per u pi ntal it of Sett eaing ch : A drug Match . ed-Claims Analysis of Roebu J93ck 54 2014 Ad )o . -tIn ras t tu he zu co mant b ext emta on f sthis ine study, a refe 34rence price co 0.4 uld % be set at ®a 36 level that corres 0.4p % onds closes1t t 10o the .1% J9055 Cetuximab 590.4 442.7 75.0% health coverage and commercial insurance plans in the 2016 MarketScan Commercial Claims and Encounters Our findings have implications for private third-party payers, including employers and commercial insurers. To counter 15 9 15 9 13 8 J J9 91 03 50 5 D Ca ec tu ax rb im az aib ne $409 $65.86 $1,142 $124.62 $381 1892 .2 9% 9.5% charact contraserist t, the ics b an ottd om p 1 ractice 4 dru g sty s accou le. nted for just 6 percent of all users in each site of care. However, there were J9047 Carfilzomib 33 0.4% 40 0.4% 126.0% Patients With Breast, Colorectal, and Lung Cancers." Journal of Oncology Practice, October 2018. cost of services in a PO. If patients choose to receive services in an HOPD, they would pay the difference in costs. The Study Limitat J9179 ions Eri 1 b ................................ ulin ................................................................ 204.3 ................................ 136.5 ................. 66.8% 14 * database. We created a dataset of outpatient infused oncology claims as follows. First, from the outpatient higher HOPD pricing, employers can aim to negotiate contracts with hospitals for site-neutral payments to ensure that The unit price variable is established via negotiation between third-party payers and providers before care is delivered. J9055 Cetuximab $38,882 $55,176 $29,158 189.2% notable JJ 99 31 0 diff 7 39 ere PaEri nces by nitu bm ulu in m site o ab f care. The top three33 most prescrib $10 2 ed .1 4% . 9 me 2 dicines in our s 12 $2ample 92.94 (pa 0clit .1%axel, carbo2 pl 34 7 a 0 .tin, and 8 .3 % % expect J92 ation 71 is Pe that p mbro at liients zumawould choos b e POs over HOPDs. Ho 1,w 20 ever, 6.3 POs below the re8 fere 04.nce pri 3 ce may increa 66s.e 7% Me medical an Paym fe ile, w nt pee r U esxtract er ed all claims with Current Procedural Terminology (CPT) codes in the J9XXX range as well Conclusion .......................................................................................................................................................... 14 cos We ts mak for the e no s came treatm laims abouten potential t are not h qua ig lit hative d er in the ifferenc HOPD es in relative to cancer c the are PO. or o I utcom n the ab es in POs sence ovs f co . HO untervaili PDs. Th ng m ere ark may et J9228 Ipilimumab 32 0.3% 53 0.6% 172.2% Jung For thi , J., s reaso W.Y. n, Xu, it ran epd reYs.ent Kali s the dindi. be"Impact st meas of ure the 34 of diff0B D erences i rug Pricin n infu g Pr sedogram on cancer dr Can ug prices c cer Site harg and S edpe by n di HO ng PD in s Me and di care. POs. J9179 Eribulin $24,903 $39,978 $16,639 240.3% cyclophospha J9271 mide Pem ) b acco rolizunt umed ab for 27.9 percent of all users in HO $5PD 6.7s3 but only 24.6 pe $9 rce 2.8 nt 4 in POs. On the oth 16 er 3 h .6an %d, prices to J930 at 8 orR cl aos me ucto iru the ma r befere$900 nce price, offsetti $270 ng some of the $16 ,5 6 savings. This 06 0. 4 $450 may hap 72 pe 8.n 9 $ to t 1,30he degr 0 ee tha $400 1t there 09.4% is as other identified oncological agents in the J85XX range and those with temporary CPT codes (i.e., “C” and “Q”). J9025 Azacitidine 21 0.2% 24 0.3% 118.8% wel powe l br, e dif thirfere d-pnces in arty payers quali cty an of en ca ga re ge by ca ty nc per e of pat settin ients t g. hr Ho ough spital cam plan desi puses gn to cl early guide hav them to l e greater ess re cos sources tly sites availa thatbl ae re i n Health Services Research 53(5) Part 1, October 2018: 3528–3548. 6.5 perc RefeJr9 ences 27ent 1 ................................ of P Pe O mpatie broliz nts um were ab prescri ................................ bed either leupro ................................ $li6 de 8, 4or 39 BCG intraves $74ical ,6 ................................ 6, comp 8 ared with $45, 6 jus 31.......................... t 1.1 percent 1in 63.6 15 % J9308 Ramucirumab $63.99 $102.46 160.1% J9354 Ado-trastuzumab emtansine 2,496.6 1,959.6 78.5% a competitive market and transparent pricing information. J9202 Goserelin acetate implant 15 0.2% 31 0.3% 214.9% The other Next, we sel two variables ected the 37 in our mo mo st p del res are bot cribed h oncology m influenced by treatm edicationsen that t patt toer gens. ther Dr caug ptur mi ed x is 92 p stro ercent ngly i nfluenc of total ed by clini the cally event appr of tr opriate eatment for t failur heir e care. or ot her Insure advrs ers ue se react both ions value from -based thera insu py. However, rance design it ( is i VBID) a mporta nd ref nt to erence note that pricin not g to all MJ e9 a3 n 0 P8 aymeR nta A m cru oc sis ru A m ll ab $42,642 $74,686 $46,643 160.1% HOPDs. J J9 90 34 57 4 C Aa drf oi-t lzra om stib uzumab emtansine 1 $,3 48 15 .0 .8 4 $669 .0 44 0.5 176 36 .6 .4 % % Endnotes ............................................................................................................................................................ 16 J9042 Brentuximab vedotin 8 0.1% 14 0.2% 182.0% chemotherapy utilization and spending. Claims volume for the remaining agents was too small to analyze Kalid cancer ty indi, Y., pe aJnd . Jun disg, eas and R e severity . Feld as man. wel "Dif l asferen by pcatient es in S to plerance ending On for Pro cancer vidert- hA erapy. dminist T ered reatm Ch ent emot inth ensity erapy varies by Site by a of var hos D y p pi rutal gatie s-based nt cos oncology t-sharing b prac ased tices on resid the ch e on oices hos that pitalt camp hey mak uses e . In fact, t regardinghe use o recent f hea glt ro h wth care in servic hospies. tal ow Honw ers ever, hip of one such J9354 Ado-trastuzumab emtansine $94,974 $129,410 $74,548 173.6% One thing to consider is whether cancer patients receiving oncology treatment services will be sensitive to the J9303 Panitumumab 343.7 344.1 100.1% J9047 Carfilzomib $40.03 $56.35 140.8% Total 9,275 100.0% 8,920 100.0% -- $743 $1,080 $1,075 patien ind t’s iviw du eight or al drugs, b bodut y e sur ve face n in athe rea aggre , type ga ofte this omiss cancer, and io dn uratio would un n of the like ra ly af py. fect o In our ur acco main uconc nting mo lusion de sl, . tot Patie al nts annu al Care in Medicare." American Journal of Managed Care 24, no. 7 (2018): 328–333. Fi practices thing to cons gure 3 prese has ide concentra nts r d is wh ata on ether ted on tr cancer eatmen purchas p t inten atin ien g e s ts ity. xistin recei Am g co vin ong g mm o tn he uni cology mo ty-s based t spresc ervice practices ribed s will b drug e that ss ensit , aretain nnu ive to cost al d their osag of -e r sfha -cam ates ring pu p , er s si nce l u ocat ser tihe ons were y are J9047 Carfilzomib $56,668 $53,001 $37,645 140.8% J9228 Ipilimumab 819.0 863.7 105.5% com bination of price transparency and cost-sharing changes. Cancer patients are some of the highest-cost claimants. NoteJ s9 : 3 H0 O 3PD=hPa osp niitta u l m ouu tp m ata ie b nt departments; PO=physician offices$117.67 $185.29 157.5% ($495) ($158) ($163) with claims that had missing, negative, or zero payments or zeros in the drug units field were excluded. The final spending on oncology medicines (S) is equal to the unit price (P) times number of patients receiving each drug (Q) s large (N om epe ras lyof compa h the hi et alrable . 2 gh0 est 15 b -)cos y site o (Ft c ulto lai n f mants care. 2017) . But t (T Med hey hPAC 201 ere was not only muc 7). are Ih n m more vari thes ore li eke situati ly tha abions lin ty the among , the qual avera less ge ity of p prescrib ercar son to e ed shou r me each lddi more the cations ir cl dedu os . On elct y ible, J9303 Panitumumab $40,447 $63,756 $40,489 157.5% Es J9 tim 02 a5 ted Sa A vz ina gc si*t* idine 4,944.8 2,256.0 45.6% * The Figure se pati sents are not only more likely than the average person to reach their deductible, they are also more likely to J9228 Ipilimumab $145.29 $241.11 166.0% Koller, Christopher F., and Dhruv Khullar. (-7 "The 3%) Commercial Differentia(l for Ho -23%) spital Prices: Responses (-2 F 4rom %) States and times study sampl mean units e was dispense derive d (U) d afpe ter an r user inves in both tigation of the HO the PD va and PO lidity o . f the dose field for infused oncology agents, as average res they are emble J922 , al 8 i ann n sde o u mo pe Ial piln re l treat im de uike nt m me phys a ly to bnt int ician practices reach t ensity heir was ou 2 t1.2 rather t -of- pe pocket rchan ent (OOP) m lo $1 wer ore in 18, maxi 9 in 8t8 eg HO rat mum PDs ed . campu com $Henc 20pare 8,2 e, s4 f d 1 hig acil wit h ith er ies. POs pat .ient $ 1D 2eco 5cos ,48 mpos t1 -sharin ing g ann may ual 1 not be 66.0% J9202 Goserelin acetate implant 8.2 7.6 93.2% reach their out-of-pocket (OOP) maximum (Fronstin and Roebuck 2019). And price transparency by itself has been J9025 Azacitidine $4.22 $10.67 252.9% Figure 1, Em Hypl pothetical oyers." J Exa ournal mple Show of the American M ing How Differe edical nAs ces sociatio in Payn men , Jults y f2019: or Infuse E1–d Chemot 2. herapy Drugs in Hospital described below. *D Jr9 u0 g 2 u5 sers m Aa zya e cx itc id eie nd e the number of patients in cases wh$ e2 re0 s ,8 o6 m 7 e individual$ s 2 ta 4k ,0 e7 m 3ore than one d$ ru 9g ,5 . 2 In0 such cases, th2 e5 2.9% eff dosage ectiv J90 i e unles 4nto 2 number Bre s clini ntu of cal xim t ly app a her b vapy ses eropriate dotin sions a VBID o ndr referenc drug units e p pe ricing r sess too 7 ion 62 ls re h .5elp main i s expla n fin orc the e f or source o 8 pati 90.ent 1 f sth wh ese o observ exceed ed OOP 1 16. 7% found to be insufficient in reducing hospital prices (White and Whaley 2019). Hence, the potential for price J9202 Goserelin acetate implant $426.10 $787.10 184.7% Annual Spending (S) = Price per Unit (P) x Number of Patients (Q) x Number of Dispensed Units (U) Outpatient Departments and Physician Offices Can Be Partitioned by Drug Mix, Treatment Intensity, and Unit Com pro pe potritive tion m of uark seret s tsa require king eac h numerous drug is de se nom llers inat e(d and by b touy tal e urs se )r as w s, notell to as tal pprice atient transpar s. ency. The market for outpatient J9202 Goserelin acetate implant $3,494 $6,017 $3,258 184.7% maximum di Mfe fere annces i s. n treatment intensity by site of care. For all but two 7 4 dru 5.3gs (leuprolide and vincr 587.0 istine), patients treate 78.8 d % in trans Jparen 9042 cy and Bre higher ntuxim co abs v t- esd haring ma otin y not be effective un$ less 129 b .6 en 5efit designs re $qu 21ire 5.7 that 3 the select services 166. 4% ^ Neprash, H.T., M.E. Chernew, L. Hicks, T. Gibson, and J.M. McWilliams. "Association of Financial Integration Between Measuring Dose for Infused Oncology Agents Price ....................................................................................................................................................... 6 oncology in the U.S. today meets neither criterion. Our results indicate that the proportion of infused oncology drugs Note J9 s0 : H 42 OPD= Bre hon stp u itx a im l oa ub tp v ae tid eo ntti n departments; PO=phys $i9 c8 ia,n 8 5 o7 ffices $192,018 $115,396 166.4% POs received more sessions per year than patients treated in HOPDs. However, the reverse was true regarding units impacte Mean d by VBID or reference pricing are subject to patient cos $9 t1 -s.hari 49 ng and not ful $1 ly cov 70.34 ered by insurers re 1gardles 86.2%s By standardizing drug mix and treatment intensity across the two sites of care, we removed variation in cost due to Physicians and Hospitals with Commercial Health Care Prices. JAMA Internal Medicine, 2015: 1932-1939. EBRI Issue Brief is registered in the U.S. Patent and Trademark Office. ISSN: 0887 –137X/90 0887 –137X/90 $ .50+.50 provide **Diffed in rence H OPDs betwee h na asc tco uant l ainu nd e cd outo nterise rfactu from al for 4 e3 a cp hercen eleme tn report t holdin ed g o in th 2014 er value (s W cinn ons et tanal. 2018 t. ) to 49 percent in 2016. D Stra u tig s tM ica ix l -W sige niifg ic h ate nc de M ofe d ain fference in means by site of ca $re 13 d ,1 e2 n8 oted as fol$ lo 2w 1s ,8 : 8 * 1 p<0.05. $12,115 180.6% ® infused per session, which were higher in HOPDs for 31 of the 37 drugs. The MarketScan outpatient medical claims file contains a field named “units”, which is meant to capture the of No pat tesients : HOex PD ceeding =hospi an an tal outnual patieOO nt d P ema part xm imum ents.; PO=physician offices treatment differences and isolated the fraction due to price alone. This in turn permitted us to estimate the Figure 2, Differences in Mix Among the Top 37 Infused Chemotherapy Drugs Administered to Employment-Based and © 2020, Employee Benefit Research Institute –Education and Research Fund. All rights reserved. Independent oncology practices are becoming less common and have disappeared altogether in some markets (Clough, Notes: HOPD=hospital outpatient departments; PO=physician offices ® Nikpay, S. S., M. R. Richards, and D. Penson. "Hospital Physician Consolidation Accelerated in The Past Decade in number of milligrams of each drug infused per claim. This field has only recently been included in MarketScan, All differences in means by site of care are statistically significant at the p<0.05 level except where denoted by ^. counterfactual cost of oncology drugs if prices observed in POs had applied in HOPDs. In other words, how much could Commercially Insured Cancer Patients in Physician Offices and Hospital Outpatient Departments, 2016 ......... 8 Notes: HOPD=hospital outpatient departments; PO=physician offices Dinan, and Schulman 2017). The combination of hospital mergers (MedPAC 2017) together with widespread purchases Cardiology and Oncology." Health Affairs 37, no. 7 (July 2017): 1123–1127. which means that there are no published reports to verify the validity of values contained in this field. For this potentially be saved if HOPD pricing were brought in line with what physician practices receive for identical products? ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org A res ebri.org ebri.org earch rep Is Is Is Is Is Is Is Is Is Is Is Is Is Is Iss s s s s s s s s s s s s s sue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B ue B or ri ri ri ri ri ri ri ri ri ri ri ri ri ri rit from the E ef ef ef ef ef ef ef ef ef ef ef ef ef ef ef • • • • • • • • • • • • • • • J J J J J J J J J J J J J J January anuary anuary anuary anuary anuary anuary anuary anuary anuary anuary anuary anuary anuary anuaryBRI 16 16 16 16 16 16 16 16 16 16 16 16 16 16 16, , , , , , , , , , , , , , , Educ 20 20 20 20 20 20 20 20 20 20 20 20 20 20 2020 20 20 20 20 20 20 20 20 20 20 20 20 20 20ati • • • • • • • • • • • • • • • No. No. No. No. No. No. No. No. No. No. No. No. No. No. No. on and R 498 498 498 498 498 498 498 498 498 498 498 498 498 498 498 esearch Fund © 2020 Employee Benefit Research Institute 14 12 13 11 10 15 16 8 9 2 5 6 7 3 4