Low-income workers who switch to high-deductible health savings accounts (HSAs) plans are more likely than their higher-paid colleagues to avoid certain types of health care, according to new research by EBRI.
This study examines whether there is variation by worker income on how an HSA-eligible health plan affects health care services use and spending. Does the typically flat-dollar gap between a health plan’s deductible and the employer contribution to a health savings account (HSA) have a bigger impact on the use of health care services among lower-income workers than it does for higher-income workers?
The data for this study come from a large employer that offered an HSA-eligible health plan alongside a preferred provider organization (PPO), includes between 150,000 and 200,000 individuals, and covers health care services use and spending over the six-year period from 2009-2014.
Here are the key findings:
- The HSA-eligible health plan was associated with a decline in (non-preventive) outpatient office visits for workers at all income levels, but the decline was over twice as large for workers and their dependents with incomes less than $50,000 as compared with those with incomes of at least $100,000. The decline in specialist visits accounted for most of the decline in outpatient office visits among the group of workers with less than $50,000 in income.
- There was an across-the-board decline in prescription drug fills regardless of worker income. However, unlike the results for outpatient physician office visits, there was not a clear relationship with income level.
- The HSA-eligible health plan was associated with a reduction in various preventive services by worker income. For example, lower-income workers reduced their use of influenza vaccinations more than higher-income workers. The usage of preventive office visits exhibited the same general pattern as influenza vaccinations.
- In contrast to the other findings, the HSA-eligible health plan was associated with an increase in emergency department visits and inpatient hospital admissions among lower-income individuals.
- The usage levels of certain health care services--inpatient hospital days, avoidable emergency department visits, pneumonia vaccinations, human papillomavirus (HPV) vaccinations, and glycated hemoglobin (HbA1c) testing for individuals with diabetes--were unaffected by enrollment in the HSA-eligible health plan both overall and by worker income.
9
Fi Fig gu urre e 1 53
The Finally, Quan, H., In contrast, H s with T for in these increases appear to References HSA-Eligib th udy ranke e stu dec dia divi severa d b line duals y al etes. et a in d so e le Health office S prev Men ll of the prior with . "Co As are savin xamin over a entive vis incomes dinied g Al tsPlans and Heal by the i off be short-term ggorithms for studies su e g hig 54 s accounts tha belo ice visits m enro h- pw $ a an ct of th 50, lle d lo ffered from (i.e., d i De 00 e w n e HS th Savi 0 w ffinin - the ftroutine priority, e are cts. Th eA g Comor HSA-e refully p shortc no l physic ngs an on bas e e port lon ligi ffe b e Accounts (HSAs) omings i g d on ienrolle b sp d er statistical al able, ct of t le ities i ex endi healt the Oregon State Priori ams for own es ng usin h nn e H IC h th an e pla D-9 SA-eligi e d, d 0.17 l a y measurement n g allo signi and - w dCM an ults an e4 for re mor fre b we fle icant. q d h d amo d ue ICD P ewel PO en alth e ntly l T - iof work tized List of H k l10 A h unts from fun ely than visits for chi plan e stu rollees. Ann d d di ministrat edd er i y e d thos by th not v clai n xco am le dr e in me. ms data, co e in u ive Data. alth ines data ary by en) ex al i the divi Wharam, n Services, come dual, PPO income hibited t " go Med nverted wa a ting o whi n et al. s have for a id cha back e l ch
Figures
Paul Fronstin is director of the Health Education and Research Program at the Employee Benefit Research Institute
The Impact of an HSA-Eligible Health Plan on Health Care
Figure 2
a
ImpaA cv t o erfag HS e U A-E se o ligib f H leeal Hte ha C ltar h P e S lan on S ervices pe , b ndi y Wo ng,r by ker E In m cpl om oy e, e 2 e I 0nc 13ome
Figure 3
to 20 received a same ge takes into acc unused amou number (20 into 2 0Care 709, 0 ) 1 and Ha of 4 but neral 42 doll pro he , n most enrollme alth car nts can ars. This o relationshi s viland, o unt th tate-speci . 11 (e e e et a N be servi ove was roll fp ffic e l to . m cnts in t annual (201 ces, such as ed ove tantige b income of a er 2 1) use me 0n (PSA spe he H r0 to t 5 as in dical ): n d SA-eli h di 11 ge in e ) test ng o fl next 30 interventio patient o-code uenza - g9 n ible . for prosta year primary vaccina hospit d healt inco an n o h pl Figure 1 d ev care m Figure 9 al days n high te tions. Among pr e an canc en data. eventio ph er taken took , avoi amon ysicia 1 er. Hib pl n fr d b o ace in g HSA-eli n visits; speci able ard, i om one jo f death, ndivi emer Gree 2duals 01 gcompl 3 ne ib gen b to le h w a wit a hc n list visits; othe en y de the e d h incom ications, or alth plan Tus fina ne partment lncial er (2008) used a xt. Furt es b e inr n fut e vis centives to r outpati lo o hermore, u llees ($ w $ re costs; its, pn 50,e 000, 112, eum nt funlik e lag se n t 7 o roll i h 81) f rvices; nia e HRA o ree o r n f
Figure (EBRI 1, H ). M. ealth Pla Christopher n Enrollment, by Roebuck is Yea preside r, 2009-2014 nt and CEO .......................................................................... of RxEconomics, LLC. This Issue Brief was written .................... 9 with
Brot-Goldberg, Zarek C., Amitabh Chandra, Benjamin R. Handel, and Jonathan T. Kolstad. "What Does a Deductible Do?
HSA-eligible health plans were created under the Medicare Prescription Drug, Improvement, and Modernization Act of
Sample Demographics, by Plan, 2013
Services Use and Spending by Worker Income
Distributi aon of Annual a Earnings, 2014
Impact of HSA Impact of -EligibleHSA Health -Les Eligi s than Plan on ble Health Plan $50, Likelihood 000 ? on $75, of Receiv 000 ? $100,000 ing ? $125,000
that pl vaccinations, human pa effe 100 funds, prescript wheth di ctiveness e d HSA a an r a not h ion wer wor dr m e a of me ugs; in k introduce ve a ounts can er was dical preve patie pai pillo int d; e be use ntiv d n ervention; t mavirus hnrollment e hospital servic o of urly or salar dfice for no vis (HP an in iV n t after moving to t d wh - ) he HSA healt ivaccinatio ed. es; emer ether Fel h e -d x eli a co man pen genc g nis, glycate bs a lndition e es (al than amon y de he n the HSA d al Parent partments b th eit o is sel dpla h -eligible heal e e g nn 2009 mo f (2 -l increased from P a taxa imiting. PO enro 010 globi ; and total ?2013 ) ble n had actu (HbA To basi llees th pla det he ($98, 1c s), wh n 3 alth car e al incom ) ,rmine though this estimate was testin perc 21 ich 4). ent the e g for in furth e spen dat in im e 2012 to a r contri pact of ding. dividuals but for th butes 26 perc a sa e CD with mple o to the not HP by ent in f
assistance fro The Impact of m Cost-Shar EBRI’s research an ing on Healt d edito h Car rial staffs. Any e Prices, Quan vi tities, an ews expressed d Spendin in t ghis repo Dynamirt are t cs." NBER Wo hose ofrking Paper No. the authors and
2003 (MMA). To be an HSA-eligible health plan that satisfies the MMA, the plan must have had an annual deductible of
HSA-Eligible
Wharam, Impact of J. Fr Income on t ank, Bruce E. hLandon, Al e Use of Health Care ison A. Galbra Total ith, $50,000 Ken P. Kleinm $74,999 an, Stephe $99,999 n B. Soumera $124, i, an 999 d Dennis or M Ross- ore
Figure 2, Sample Demographics, by Prev Prescription enti Plan, ve 2013 Cancer ...................................................................................................... 10 Drug Fills, by Screenings, Employ by Employ ee Income ee Income
201 statistically sig diabetes, cervi worker i save ince only 4 3, 2 follow 9n CDH n com 30tive %ed e P . As a resu nificant c , the enro by a sli al canc study llees. ly d er scr ght lt, a i Br e i ffn e xe ot-Gol crease to r amine worker enin ent from z g, dberg, d t a has a h n 2 e d c 8 ee e ro perce o x tperi . One lorectal ca much stron al. ence n (2 t in of 015 of 2 10 ) al nc 0 h 0 i g 1er scree er 4 o so had n . urly ince divi Thus, 2013 duals an nnin tive actual d sala g. wit not i ried employee h was the first n incom to spen come es b data, d HSA e year of ts, as income but only w een funds $ HS 50, (r 6 A-e elative 000 perc data was lig and ent of ib to HRA le $ he 99, its sam anot a lth pla 999 fun v d did p ailable. s n le ) , not had but
should not be ascribed to the officers, trustees, or other sponsors of EBRI, Employee Benefit Research Institute-
21632 (National Bureau of Economic R a esearch), b October 2015.
at least $1,300 for employee-only coverage and $2,600 for family coverage in 2016, and the plan’s out-of-pocket
Variable Health Plan PPO Less than 29% $50,000 ? $75,000 ? $100,000 Figure ? 7 $125,000
August 2016 • No. 425
Outpatient Physician's Office Visits 3.8 3.5 3.7 3.8 3.8 3.9
Services Finally, Degnan. in addi "Emergency tion to the k Le Depart sse th y independent an $50ment Use a ,000 $50,000 variable nd ?$74,999 Subsequent indica Emp $75,000 ting plan H loye o ?spitalizatio $99,999 e s in choi th ce e $100,000 —HSA-el ns Among Members of a samp ?le $124,999 iw gie bl re e rhe ela $ alth pla ti 125 ve ,000 ly or h n Hi i M versus gh or gh-Deductibl in e come PPO—a . Ave e ll rage
Less than $50,000 $125,000 or Mor28% e
$50,000 ?$74,999 $75,000 ?$99,999 $100,000 ?$124,999
coverage for have a It foun rather to income pr d t belo save them even hat t w most enrollees, and by h tive of $10 e re 0,0 ductio for fice 00. visit. Among in possible fut The ns in bot EBRI stu h h u 2014, re igh duse. divi y c - and enroll o duals mbines a low ees with -prior c had some inc tua ity omes of at l worker offic ex e visits inco pe lea rien m were sce e data t $10 u gr sin 0,0 eater w g th 00, t ith e f medical o p hr hourly le HSA an. T claims h-e eli emp la gck of ib data lle oyees than he a stati alth and a pla sti b for call n o u had t one y no -
By Paul Fronstin, Ph.D., Employee Benefit Research Institute, and M. Christopher Roebuck,
Education and Research Fund (EBRI-ERF), or their staffs. Neither EBRI nor EBRI-ERF lobbies or takes positions on
Total $50,000 $74,999 $99,999 $124,999 or More
M maximum c ale an not exceed $6,55 52% 0 for self-only covera 54% ge or $13,100 for family coverage with the de a ductible counting
Primary Care Physician Visits 1.9 1.9 1.9 2.0 1.8 1.7
Figure 3, Distribution of Annual Earnings, 2014 ...................................................................................................... 11 Impact of HSA-Eligible Health Plan
0.04
Plan s models i Health Plan. ponsors ncluded t and " h Je insurers ournal o followifng covari may or may the American M ates: n age; ot be edical Associatio a concer ge-sqn ue ared; annual d with ngeo these 2 salary 97, no. 10 (March graphic re findings. was $ gion; 99,5 Sinc ho 56 20 e i usehold size; the n 2 07009, ): incre 10 incr 93 ase i -1e 102. union status; an s emer ing to ge $1 ncy 05, in 0dicators 90 in
Buchmueller, 0 Thomas C. "Consumer-Oriented Health Care Reform Strategies: A Review of the Evidence on Managed
Where the world turns for the facts on U.S. employee benefits.
signific effe Examinin salaried emplo Ph.D half o ct on f., RxE t ant h pr g th e sa fin eventive o e mple d cT impact yotal ing onomics ees. One wh has in en of ffic enr 20 limi come below e vi 14 o sits. t data ar ation of llment in a $ the e i 1n 0 n-tr study 0 $299 HSA-eli ,00 oduce 0*.* was that*dg may ible -$530 in 30 dicate perc that-$179 ent o th f ere is enrolle-$215 an es w ine* e xre perie in -$225 pla nce* n facto s with r associated w-$356 a $1,*00** 0 dedu ith ctible
Age specific policy proposals. EB 32. RI invit 4 es com35. ment on 7 this research.
Specialist Visits 1.9 1.5 1.7 1.9 2.0 2.1
toward this limit. These minimum allowable deductibles and maximum out-of-pocket limits are indexed to inflation.
on Use of Health Care Services
0.02
depa HSA-eli for eac rtment h g ca ible len vi he s dits alth ar yea and i plans r; an npat sh d iould ent the Ch patient also be e arlson C 0.02 adx mis amine omorbidity sions d se dis par In ap201 p dex (C aear ov tely 4. Me from CI) score, w er tim dia HR n e A-relate income ra , some p hich d is a lan s plan ngeproxy measure d s p becaus from onsors $ and 93, e an 00 ins employer 0 in of he urers 20alth statu 0 may 9 off to e dec ring t s ide h e
25% Inpatient Hospital 11 91 73 71 44 -61
Competition and Consumer-Directed Health Insurance." Milbank Quarterly 87, no. 4 (December 2009): 820-841.
G HS and a eogr A- aphi e $ lig 7c i5 b R 0 le egi ehe m on ployer co alth plans. ntri On bution ce ind to th ividue alHRA, making t s see the out-of he -pdeductible ocket costs gap of associated $250 r with an emerg elatively smale l.ncy d epartment visit
health pla Pren s on crip use o tion Drufg servic Fills es masks other important 12.1 8.8 11.3 13.1 12.4 12.5
0.02
Figure Certain 4, primar Percentage of Sam y preventive services—typic ple Using Health ally those Care Servic deem es, by ed to pr Type of Servic event the one, 2013 set of disease—can ........................................... be and often are ... 11
-0.1
Use of
not to mo Study S former may derived from a dify mp be of E diag the le merferin nos g designs of their enc is g y cod D oepar nly e the st from m ment hi plans gh-de edic as da uctibl a resu l claims 1 e h lt of th e data alth 118 for 1 plan ese fi $10 *** wh 0,0 9ndi health 00 ile ngs. neit in -6 2 c However, 014. o her nditions (Ch fun About d -2 in if s g n on ponsors or insu o arlson, e-hal r facil -f4 o itatin et al. 1 f the g e worker co rers are m 987 ployees 9 ) (D conce eyo, ntributions had Crned herkin
Northeast 2% 2% Figure 8
Retirement Inpatient Hospi and health benefits are at the heart of w tal Admissions 0.03 0.03 orker 0.03 s’, emplo 0.03yers’, and our nation’s 0.03 0.02
and a differe nnces ensuin amg in ong vario patient a us d groups mission, they . The study i may be dent less li ifies kely to use the emergency 22%department for health care services
Copyright Information: This report is copyrighted by the Employee Benefit Research Institute (EBRI). It may be
0
exempted from the deductible and covered in full (Fendrick, et al. 2014). (These preventive services are in addition to
Introduction Health Service Utilization Services
Primary Care Physician Visits -20 *** -13 a-18 *** -18 *** -11 ** -11 **
Bundorf, M. Kate. "Consumer-Directed Health Plans: Do They Deliver?" Research Synthesis Report No. 24 (Robert
0
a Fel to an HSA. and C bou M diman dw t th iol es Inpati e 19 and t i In m 92 ent pa Parente ) (Qua many HSA-eligible H ct os of pi HS tn, al (20 D et al. A- ay1 el s0 ig ) 200 e ib23% x leamin he pla 5). Import an ed da lth pla s, wor Impa ta on over antly, gender n c ks on ers are t6% o 0.f l 11 H ow Sleft all s e A r--in o E incomes belo p and c li n en 0. o g their o me 12 ib din policyho l e g worker on w H n to he eal lw $ der status we 0. s, they can alth car fin 10 th 1 P 0d0 a ,00 lan n e HSA s 0 co ein 0. rvi re 20 13 provider nsider c impl es 14 fr (Fi modi icitly i om gand open an ure a sin 0.fic n 10 cl 3). Ov ations to all ude gle larg d i erall, n0.10 accou th e empl 7 e m evi perce ate any n o oyer tdels . nt
economic security. Founded in 1978, EBRI is the most authoritative and objective source of
-0.2
in the future.
Figure For th important used e EB 5, Aver without di RI s fferences age Use of Health Car tudy, permission data amon wer g in bu e use t citation come d f er S grou om a e of t rvi ps c large h es, by wit e source h res manu Worker Income, p is re ec fat ctquire uring em d. 2013 ployer, wit ................................................................. 12 h employees located throughout the
Outpatient physician's office visits -0.250 ***
The Impact of an HSA-Eligible those preventive services that the Affordable Care Act (ACA) man Health Plan on Health Care dates be covered in full.) Otherwise, all health care
Specialist Visits -7 *** -21 -13 *** -14 *** -8 ** -8 **
Since 2001, more employers have begun offering “consumer-directed health plans” (CDHPs), which are a combination
SouthEmergency Department Vis18% its 15% 0.15 0.20 0.17 0.16 0.14 0.13
Wood Johnson Foundation), October on O 2 ff 0i1 ce V 2. isits, by Employee Income
since they did Endnotes negativ that added a e im 20% pa n HRA to ct on th not vary over is its arr grou ay of time and p. For health exam wer pl pl ean choic therefor e, employers ee abs s. Th c had o e HRA o orbed in uld incomes provi p the in tion de below h was dividual-level igher ad $5 HSA co de 0,0 d i 00, n fixed effects. 200 15 ntributions percent 1, and me to lo ha dical d in wer- comes bet claims income da ta ween
information on these critical, complex issues.
to the United States. impact of the HSA-eligib The study used health le health insu plrance an. For enrollm example, ent informatio Priman, me ry care dical physi an cia d npres visits cription drug claims-, 0a .1n 8d 2 ***
-0.02 Other Outpatient Services -133 *** -269 -136 *** -133 -170 ** -68
services must be subject to the deductible.
Prev -0.3 entive Services
Because West employers have been of 57% fering HRA-base 76%d plans longer -0.02 than they have been using HSA-eligible health plans,
of health insurance coverage with a deductible, generally in excess of $1,000 for employee-only coverage, and an
-0.02
-0.29** -0.02**
Figure This workers. were examined from Issue Brief 6, T Total Average Annual S his would does not 2000 re Leduc ss t pr h ? e an the 200 esent f $50,3. 0typically fl 00 pThe ending indings stud on Healt at-doll related to t y included ac ar gap betw h Car hose varia e tual in $50, Services, by een 00 come 0 an b the h les th d data $Type of e at 74,9 alth w for e 99, plan re employees, but fo un Servic 29 deductibl a p ffe ee rcent h ctand Worker ed $1 e b an 2ya 5, - en 0.02* d 00 d inco und that income 0 or remployer co ol Income, 2013 lM m m ore es bet ent in tw he ntri ee HS n butions was not ...... A- 12
Services Use and Spending by Worker Income Recommended Citation: -0 .0 2Paul Fronstin $50,000 ? $7 and M. 4,999 Christo $75,000p ?$9 her 9,99 Roe S 9 pec biuck, a $1 lis 00 t v , “T 0i00 sihe ts ?$12 Impact 4,999 of an HSA-Eligible Health -0.067 Pl*a **n
while worker incom outpatie P ent off r data escriipt for ce ion D visits the rugs perio fell as a r d -0 2009 .03 esult ?2 -014. 151 of enrollm * Th ** e empl -ent 435 in oyer ha ** d of -79fer ***ed an -119 HSA **-*eligibl-e76 health plan since -216 *** at least
-0.03**
-0.03***
Physical exam\w ell child visit 39% 28% 34% 37% 43% 44%
Buntin, Unknow Meli n region nda Beeuwkes , Amelia M. 0.3% Haviland, 0.5% Roland McDevitt, and Neeraj Sood. "Healthcare Spending and
indivi Descriptive Statistics 1 most of the ea dual savings or r rly research eimbu orsement acco n CDHPs focus un et to d onpay out HRA-ba -of-poc sed plans. Un ket costfs for h ortunately, ma ealth care services. ny studies onl There ar y noted that e two ty th pes of ey
EBRI focus -0.04 es solely on employee benefits research — no lobbying or advocacy.
eligi to the H a pre bd le h ictor SA, so ealth of total spen that lo plan by wer worker income an din -inc g. A main ome worke limita rs d overall. A stati would hav tion of thee stu ad $75, s d ditional ti y was cally signi 000 an the funds Pre d sample si sfc ic $r99,9 ant e ip to pay tion99, d ffect rze: ufor g 22 f was ills p Th heal eercent h not rth care services. One e we foun re ad fe inco d on wer m inpati than es bet e 3,00 nt w-ee hospital 0 0 .7 n 57 ***
on Health Care Services Use and Spending by Worker Income,” EBRI Issue Brief, no. 425, (Employee Benefit
See EB http://www.mercer.com/newsroom/nati RI E -0 m .4 ployee Benefit Research Institute Issue B ornal-survey ief (ISSN 08 -of-employer-spon 87 ?137X) is publi 2009 shed s ?ored-health-pl 2014 by the Employee B ans-2015.html enefit Researc h Institute, 1100
0 -0.04
the HSA An em 2009 aployer n- d in elig 2 ib may off 013 le h intro ealth er an pla duce n, HSA-eli d f the inanci dec gible a line l in h w ceealth a ntives to s more tha plan, but enrn o is not ll in the require plan,d whi to co ch ntribut boostee d to a partici n HSA. pation Whe . n an employer
Breast cancer screening (females,
Relationship to Policyholder
-0.04**
15%
By Paul Fro Preventiven Car stin, Ph.D e in High-Deductibl ., Emplo e an yee d Co Benefit nsumer-D Rese irect arch Institute, and M. ed Health Plans." American Christopher Roebuck, Journal of Managed Care,
were examining “CDHPs,” without specifying whether the CDHP was an HRA-based plan, an HSA-eligible health plan, or
Figure accounts: heal 13th S 7, Im t. NW pact of H EBRI , S th savings uite 8 stand 78, SA-Eligible Health Ws a acc shi alone in em n ogunts (HSAs ton, DC, 200 pl Plan on Use of 05 ) oyee and -4051, ben healt at $3 efits h 00 re pe Health Ca irese m r ybursement ear o arrch isre in Inpat acl s S ud e a an ient hos rvices ed rrangem inde as par ................................................... pi pend t o tale f adm a nts memb en is (HRAs). Both t, sions nonprofit, and no ership subscr t ipt yipes may r on. Pnpa resor-rtisan te 0. e d002 ......... 13 ceive **
15%
days Fig consideratio indivi ure , avoida duals 1 shin t on w blwith t s e eme h the e sample, a his a dis rge trib ncy ppr uti n oach involv depa on of d only rtment in 4divid 29e we vi s us making sur als re en its by ty , pneumo rolle pe o d i e t n ni f h $10 plan a the HRA. e lo vaccinatio 0,0 over th wer 00 -inco and n e m s, HPV 200 $12 e work 94,9 ? vaccinat 299, a 014 ers u pnd see the additio riod. B i 28 ons, or HbA pee rcent tween n ha 1 al HSA c200 testing for d in 9 a comes of contri nd 20 butions 1 2,
Research Institute, August 2016).
age>=50) 50% 40% 47% 48% 52% 54%
does contri twice Poli as lar cyholder but ge e for in to adivi work duals er’s HSA, those c with 38%income le oss than ntributio 44% ns are excluded from the worker’s taxable income. Employer
-0.06
standard postage rate paid in Dulles, VA. POSTMASTER: Send address changes to: EBRI Issue Brief, 1100 13th St. NW, Suite 878,
2 Less than $50,000 ? $75,000 ? $100,000 ? $125,000
March 2 -0.50orga 11: 2 nization. 22-230. It analyzes and reports research data Inpat without ient hos spin o pital day r unde s rlying agenda. All fin 0. d 002 ings,
Ph.D., RxEconomics
tax-favor both. Also, stu ed contributio dies that e ns an xam d make ined CDHPs tax-fa may have o vored reimb nly ur exami sements. In ned a h 2015, ealth 2plan 9 perce witn ht of a high emp dedu loyerctible; s with in 10 other or m wor ore ds,
Literature reviews can be found in Buchmueller (2009), Bundorf (2012), and Buntin, et al. (2011).
indivi toward healt duals wit Cer hv h care servic ic di al c abet ances. er ses t creeni hat are hi ng (femal gh es value, , because $12 once th 5,000 or e empl more. oyer contributes to an HSA a worker 3 can use the
most enrolle The study Washingto exa es wer n, D mined C, 200 e i the 05 n a -4 set PP 051. C O. Only a of em opyright 2 ployees f 0e 1w 6 by workers ha who Employ were ee B fd si e ull nef -g t it R ime active ned esea up rc hf I onr wo s the titute rkers, und HS . All ri A-e ghlts igre ier a bse lerv g h ed. e ea 65 N lth p o as of . 425. lan De befc. 31, ore 2 2 01 014, 3. as well
-0.06
Spouse 21% 21%
$50, contributio 000 comp ns must be ared with the thos same do e with i llar am ncome ount or t s of at least he same percentage of the deductible for all employees.
Total $50,000 $74,999 $99,999 $124,999 or More
Emergency department visits -0.1 002
Figure 8, Impact of H whether S on fi A-Eligible Health nancial data, Pl option an on Office s, or tre Visits, by Emplo nds, are revealing yee Incom and reli e ............................................. able — the reason EBRI informatio ..... 14 n is
a healt Haviland, h plan et a tl. (2 hat was 011) e neit xam heined data r HSA-eli fro giblm e nor 43 em associated ployers that of with an fere HR d eith A. Not k er an nowi HSA ng- what ty eligible healt pe ofh CDHP plan or the stu HRA-d base y d
workers an Report Availability: -0.1 d 59 percent of employers This report is with availa 500 or ble on th more e I workers offered nternet at www CDH .ebriP .org s based on either HRAs or HSAs. Also,
-0.0 age> 8 =21) 32% 31% 32% 30% 31% 33%
account not However, as their Child dein pe ju 2 n013, d st to pay ents enro . To al llme for low for h nit in gh-va 41% the cros lue HSA- d ser s-seeligi ctional tim vices, but 35% ble health also e se plan ries for mode incre low-value ased ling, the d h to 2 e s alth 6a mple perc servic ent focus ,es and eve upe from d on i 3 nn div pe nircent i o duals n-he alth serv n who 201 we 2. re ices. The
-0.6 -0.57***
Total -$326***-$309 -$186 -0.09***-$253**-$248**-$442***
-0.09***
$10 3 0,000. Mor the e spec gold sta ificaln ly, dard there for was a priva 0. te 48 r analysts a eduction nd in decisi Pron event make ive Srs, ervgove ices rnment policymake -0.10 r*s, ** the media, and
Charlson, M. E., P. Pompei, K. L. Ales, and C. R. MacKenzie. "A New Method of Classifying Prognostic Comorbi -0.10*** dity in
With respect to the use of health care services, most (87 per-
There are exceptions to the comparability rule. For instance, employers may make matching contributions that are
Colorectal cancer screening -0.11***
Findings are actually exami CDHPs health pla covere 10 n. %The stu mixed with r d ned about 26 makes dy exa e mil itspect to m dif lion people in 201 ined the first ficult medic to draation a -w year AT A GLAN conc 5, experi dlusions about prior rher eprese een nce ce ntfor ing abo of in CDHP dividu u t 13 perc en wals with o rollees duri CE rk. ent o highng the perio f blood the privately pressure, d 2004 insured market, dyslipidem ?2007. Beca ia, and use
-0.12 -0.12***
Household size 3.5 3.2
Inpatient Hospital 36 121 59 66 8 -14
increase in en Regar enrolle dd less of in at l rollment appears wheth east two er aconsecutiv n employer co to e y be direc ears ntributes to covering the period 2009 tly attributabl an HS eA, an to an incr individual e e ?ase in 2014. n fi Wh rolle nanc ile d the ia in l i an HSA nstudy di centives to -el d igi not i b join th le he mpose a alth e HSA pla multi -n eli is g y ibl ear e
Physical exam/w ell child visit -0.005 *
the public.
Figure office visits per person 9, Impact of HS amon A-Eligible g in Health dividuals Plan on Prescription Dr with less than ug Fills, by Employee Income .................................... 15
Longitudinal Studies: Development and Validation." Journal of Chronic Disease 40, no. 5 (1987): 373-83.
-0.1 cent) of the individuals in the sample used some type of service
(age>=50) 19% 15% 18% 17% 20% 22%
conditional o -0.7n a contribution by the employee if done through a cafeteria plan. Furthermore, employers may contribute more
diab In additio actual worker income data were etes, CH n, so F,me of the reduced usage re and depression. For not the availa molble, famili ates to service st part, thes e HS we s A plan re cl not subject assified had no t as low oimpact o the de income ducti n me if bdication le, such the median ad as in her fl income o e uenza nce, but vaccina f th ther e e tions
Umost CDHP nion Status enrollees were in HSA-eligible versions of such plans (Fronstin and Elmlinger 2015).
Emergency Department -3 83*** -7 -9 -7 1
The Employee Benefit Research Institut Breast c e anc (EBer RI) was founded in 1978. Its screening (females, age>=50)mission is to 0.001
health eligi continuous ble pla to o n, en pw erollment r n h an ere t d contr he empl equi iburoyer dro te to a ement, n it did p HSA ped o premiums use only n a tax- complete pre rela ferr tive to th ed bas eligiibi se . For lity PPO. While years. After 2016, a the worke a epplyi arliest data i r with ng these indivin du criteria, t this study ar al covera he sa ge e is mple from
7% -0.17***
-0.17***
$50,000 in Vac inc cinati ome, a ons & S0.29 creeni rngs eduction among persons with
-0.2
Prior Research in 2013 (Figure 4). As expected, most of that usage was
to the HSAs of non-highly compensated empl Member 2% oyees. 50%
Primary Care Physician Vcontribute to isits -19, to encourag *** -18 ** e, and to enhanc -16 *** e the developmen -19 *** -1t of sound empl 0 *** -8 *** oyee benefit
were and families breast some specifics in tca hencer screenin employ worth ee’s fi n gs. Lower-i ve-di oting. -0.1g 9*** itFor ZIP -0.1n 9instance, a *co co ** me work de area ders m was herebelow nay cut ce for20 d ba 0 epr ck on th perc essent of ion w ese se a the s ne rvi fgati e cderal es b ve overall, ecau poverty l se th dr ey may n eivel base ven by in od o t re divi nali t dh uals ze e th at
Cervical cancer screening (females, age>=21) -0.016 ***
-0.12
Figure 10, Impact of HSA-Eligible Health Plan on Likelihood of Receiving Preventive Services, by Employee Income ... 15
200 allowe containe 9, most H d to d -0 ma bet .8 S w ke an A-eli een 1 g ann i50, ble he u 000 al alth HSA contri and 2 plan 00, enrol 000 buti i lon of ees joine ndividuals, $3,d 35in 0, de201 pen while 3 a din a n g on worker d had thonly one e year, with fam and in or t ily co wclude o years of verage d both can cov u co nion a entribut ragen in d non e the as m pl -union u an. ch as
Claxton, This study EBRI explo G., ex et a al. mines w r "Healt es the breadth of emplo heth h Be er ther nefits I e is vari n 2007 a:t Premi ion by y worker u ee ben m Increases Fal incom efits and related issues. e on l To hoAn Eight-Y w an HSA-eli ear giLow, While ble healt h pla Ofn fer affects h Rates An ealt d h care
Tetanus vaccination 8% 8% 8% 8% 8% 7%
$50,000 ?$74,999 in income, a 0.28 reduction among
Who we are
physician office visits and use of prescription drugs. Over
Proponents of Non-memberSCDHPs contend that pecialist Visits 98%programs and so providing -6 50% partici ***und public policy pants -13 with a savin through objective -15 *** gs or reimbursement accou -11 resear *** ch and -7 educati ** non. EBR t and su -8 **I is the only bjecting
5%
Colorectal cancer screening (age>=50) -0.024 ***
with these servic 200This study 0 c incomes ensus. es are wa betw As noted abov s conduct covere een $100 d i en d t ,000 e, because fu hll. rough an This is d $12 the of i4 EB n the use ,direct 999. RI Ce Not ly suggest of nter e th geo-code for at this e Rese ed by th d arch ffe incom ct w e on findin e, the aH s present ealth Ben g th results at so w eme hi h fits Innovation (E should en data gher befor -inc inte 20 ome worke r09 prete BRI ?20 d13 as CRH w rs di B aalso cut ff I), s u erences with sed,
The literature on the impact of CDHPs on health care services use and spending has been summarized at length by
-0.83***
4 $6,750. employees. Indivi P neum EBRI duals onia studie who vacchav inati s the worl on ( e reac indiv hed a id dual of health and retirement ben sg e 55 and are not yet enrolled in M efits — issue edicare s such a may make a s 401(k)s, n additio IRAs, retire nal $1,000 ment
services Enrollment use an Re d sp main Sta ending. b Does le." Health A the typic ffairs ally flat 26,-do nollar . 5 (Septe gap bet mb weeern -Oa healt ctoberh 20 plan 07)’s : 14 de07 ducti -16b.le and the employer
private, nonprofit, nonpartisan, Washington, DC-based organization committed exclusively to
persons w When an HSA ith $75, Otis used to her00 Out 0 ? pati $9 pay f ent 9,999 Ser o in r no vic incom es n-health expenses, e,- and 145 a ***0.19 the di -48 stribution is subject to regu -108 * -130 * lar income tax and a 20 percent -106 * -170 ***
Charlson Com -0.1or 4 bidity Index 0.108 0.174
80 percent Vaccinat of ions the & sample Screenihad ngs at least one physician office visit
claims to hi -0.9 gh deductibles requirin g out-of-pocket payment before they are covered by insurance will induce enrollees
-0.87**
the support of the following organizations: American Express Co., Aon Hewitt, Blue Cross and Blue Shield
but Figure back on Buchmu betw not w een 11, pr e residents of ller h Im eventive servic en (200 pact of 2019), 4 was inc Bu HSA low ndor e - es, such as El r- lu igible H f (20 and ded. A higher 12d e ), a alth her cervi n -e inc d P nce cal canc Buntin, lan on ome ne for asthm Likeli et er a igh al. b n ho or a/COPD d (201 od of hoods. T colorectal 1). Rec w Howev as a he can e ivin pa lso negative, a e cer screenin g Pr per co r, only eventive nclu a few des gs that C n d a studies hav a that CDHPs ncer p -0pears to .88*** were not Screee af n t ibe n r subject to fecte ied to e gs, by drive d lo Emplo nx wer prim amin the -iarily by y n e the ee come
w ith incom CHF -0.14* , C e O ade PD, or quacy, diabetes co)nsumer-drive 4% n benefits, 3%Social Security, 5% tax treatment of both retireme 4% 4% 4%nt and health
-0.3 public policy research and education on economic security and employee benefit issues.
catch-up contribution. Incom -0.28*e ** and gains on HSA funds are tax free.
contribution to a health savings account (HSA) have a bigger impact on the use of health care services among lower-
Prescription Drugs -189 *** -435 *** -99 *** -152 *** -125 ** -244 ***
Areductio nnual Ear n amo nings ng persons wi$112, th $100, 781000 or more $98,214 in iFig ncou me re 1 Tetanus vaccination -0.009 ***
excise tax for individuals u -0.29***nder age 65. Individuals who have reached age 65 are exempt from paying the excise tax.
in 2013, with 68 percent having visiting their primary care
to make more cost- and quality-conscious health care decisions (Herzlinger 2002). Skeptics caution that patients lack
Influenza vaccination 16% 13% 13% 15% 18% 17%
lower deductible. Al and Association, B higher -income Income ........................................................................................................................ -inc benefits, l workers may adults. Int ome po oeing Co., He pulatio cost eres manage bene tningly, a s to the althways, Int fit from ment d sam here e , nce e d worker a e e ucation rnational xfor h tent. Ho an igh nd Busi d bl w employe co e ood pr ver, on mmunicatio ness Machin essure re lim attitude it n campaigns was hi es Corp., J ation of s, policy reform p gher the P in study Mor t othe HSA promote ganChase, was the ropo pla the n sals, Kai u , use o s but .......................... 16 e of ser Perma and p does not fgeo high e -coded n v n s a ent ion alued ppea easset , r s
impact o Employees f CD were able to HPs on health cho co are services se from amo EBRI’s member un sg e and spendi health ship includes a main ng tenanc by cross-section of worker e organizations income. These studies pension funds; businesses; trade associations (HMOs), exclusive are summarized h provider ere. ;
Davis, Karen. -0.16 "Consumer-Directed Health Care: Will It Improve Health System Performance?" Health Services Research
Source: EBRI analysis of administrative claims data.
Source: EBRI anal -1 ysis of administrative claims data.
income workers than it does for higher-income workers?
Pneumonia vaccination (individuals w ith CHF,
Health Plan Enrollment, by Year, 2009 ?2014
(Figure 8). Not surprisingly, the impact by income was
0% Breast Cancer Screening (Females, Age>=50) Cervical Cancerphysici Screening ( an F em an ale d 57 s, Age> perc =21)ent Covi lorsi ecting tal Caa speci ncer Screeal ning ( ist. O Age> ver =50)70 percent
comprehensibl HPV and e v, timely, acfundi cination ( ng. an fem There is wi d trustworthy in ales, age< labor un = de 26) spread ions; h formation t e 7% alth recog carh e prov n at ition is cr 6% iders and insur that if employee be itical for t 6% he em to m rs; government org a nke i e 7% fits nforme data exist, EBRI kno anizations; d 7% decisi and service firms. ons (Dav 9%w is 2 s it. 004)
*** p<0.01, ** p<0.05, * p<0.10 Prescription Drug Fills
a Mercer, and Pfizer Inc.
to be 5 services not s Distributio organiz income data. The relat ations (EPOs), pre n es from an HSA d tuobject to income. use o the f geo-coded inco f c er deductible. ared provider n be taken at m o any tim e rganiz dataations is enot a . They are (PPOs) a perfect m trn eated d HS easA-eli ure of fami as taxable inc gible he ly alincome a th pl ome un ans. Appr less th nd family charact oximately t ey are used to w eo ri- pay stics thirds for
Health 39, savings no. 4 account. (August 2004): Part II: 1219-34.
Under $50,000 COPD, or diabetes) $125,000 or More -0.002
Pre-tax dollars can be used to pay for th $50,000 ese ? out-of-p $74,999 ocket $75,000 expenses if the ?$99,999 employee $100,000 ?$124,999 has funds in either a flexible spending
-0.36***
a
lower in the higher-income groups.
b PSA He tes alt t h( sa mal vin es g, age> s accou = n55) ts. 39% 28% 32% 36% 40% 44%
of the asample had at least one prescription drug fill in 2013.
P Figure T and referred provider organization. he R worry t A12, ND Im Heal hapact of HSA t the th In higher suran- potential c El e Ex igible H perime e cost sharing w alth nt (N Plan on ewho Number usill e 1 lea 99 d3 of I to ) was th less use o npatie earl ent fAdmis pr b iest stu eventiv sions dy to e e an care, pr d Emer xamin imary c gency e the iare an m Departm pact of p d ot eher nt lan Vi de sits si, gn
-0.4
more generally, which may lead to a weak estimate HSA-Eligible for t Heah lth e Pe lafn fect ofPP fa Omily income on use of health care services.
quali of em fie ployees d medienrolle cal expe d ns in es eit a hser th define e Pd u PO o ndre HSA r Inte plan. rnal Re Other th venuean t Code Inf he luenz (I dRC) Sec. 21 educti a vacb cle inati an on 3(d), d maximum o in which u ct-of-pock ase they are e et amo -0.009 xcluded unts, ***
The data for t Source: Source: his study com EBR EBRI I analy analys sis is of of e administrativ administrativ from a l e claims data. e claims data. arge employer that offered an HSA-eligible health plan alongside a preferred
account or HSA.
Lipid test (males, age>=35 or females,
*** p<0.01, ** p<0.05, * p<0.10 EBRI’s work advances knowledge and understanding of employee benefits and their
EBRI delive *** p<0.01,r ** s a stead p<0.05, * p<0.10y stream of invaluable research and analysis.
Impact by Employee I of Income on Spending ncome ............................................................................................................ 4 Very few individuals (2 percent) had an inpatient ................... 16 admission,
on use o necessary hea f heallth care services by wo th care services, especi rker i ally n among lo come. It wer was -in conducte come indivi d fro du m al 19 s (Havi 74 throu land, et al. gh 1982 2011) (Newhouse 1993). in six geographic regions
Deyo, R. A., D. C. Cherkin, and M. A. Ciol. "Adapting a Clinical Comorbi HPV vac dity In cination ( dex for femal Use with es, age<= ICD 26) -9-CM Admin -0.004 istrative
Source: a a EBRI analysis of administrative claims data.
from an the PPO a inn divi d HSA pla dual’s taxa ns w ble er e income. very simi An in lar: (a dividual ) both nee plan ds not be cover required thee same l d by an e HSA-e vel of coinsurance ligible healt af h plan to ter the deduct withdraible w
H Health sav ealth savings account. ings accounts.
provider organization (PPO), includes between 150,000 and 200,000 individuals, and covers health care services use
age>=45) 44% 37% 42% 43% 45% 47%
The decline in specialist visits accounted for most of the
importance to the nation’s economy among policymakers, the news media, and the public. It
EBRI publications include in-depth coverage of key issues and trends; summaries of research
but 11 percent had used an emergency department.
of the 6 lower And Brot-Go whil wh Unite ldber e suc en d in g, et al. States an h cludi re 98% duce ng (20 20 d d e 14 1 utili 5 x c )amined util zation mi used laims data, data 98%ght izat fro co save costs in t m ion rates mpared w a large, for self ith people h $ -ins e short t 215 ured w ran hen empl e P doml rm, i S 2 A0 toyer that 1 es y assigne t4 might t data (males also r were , age> d ado to e pe dif = ted an x sult in 55) cluded. feren HSA-eligi larger l t coins Amourance nog i ng-t ble healt ndivi erm cost pla dua n 0. hs. 024 lplan for s Whil with s. ** e
100% Databases." Journal of Clinical Epidemiology 45, no. 97% 6 (June 1992): 613-9.
97%
Overall, i money from was met Depending on n for m divi tduals i ho e HSA. how workers view the HRA, there may not be st typ n th e s of health care services; (b e HSA-eligible health plan used $2 ) both pla 99 mu n less in s covered prev ch differe annual nce between a high-deductible health plan wit hea entive ca lth care services, re without any relative to cost sharin indiv g; iduals (c) h an
and spendin HbA g over the six-y 1c test (individual ear perio s w ith does this b did abet fro es m ) 2009 y co 85% nducting ?2014. and p 86% ublishing policy re 87% sear 86% ch, analysis, 83% and special reports on 85%
decline in outpatient office visits among the group of
What we do
-0.4finding 8*** s and policy developments; timely factsheets on hot topics; regular updates on legislative and
Figure -0 13, .5 Impact of HSA-Eligible Health Plan on Spending, by Employee Income .................................................... 18
Table of Contents Lipid test (males, age>=35 or females, age>=45) -0.011 ***
incomes of $100,000 all of its employees. This ?$12 study 4,999,was a avera bg le t e spendi o combin ng was e me $24 dical 8 lo claims wer w da hen ta with inclu work ding er 201 income. 4 claims d The ata, compar study indicate ed with d that
dated, the RAND study continues to be an important study. It is the only study related to plan design and income level
Source: EBRI analysis of administratiemplo ve claim ysee ben data. efit issues; holding educational briefings for EBRI members, congressional and
not on t neither plan he HS re A plan quired a r (Fige ur ferral from e 13). Most o a pri f t m hary care at savings phy ca sme from ician to see a redu specia ctions in list; an presc d ription (d) both drupla g fill ns used the sa s ($151) and me
HRA and a lower-deductible health plan without an HRA. Similarly, depending on how workers view HSAs, there may not be
Use of services varies by worker income. In general, lower-income workers used fewer health care services than
regulatory developments; comprehensive reference resources on benefit programs and workforce
workers with less than $50,000 in income. These lower-
HbA1c test (individuals w ith diabetes) -0.012
Since 2001, there have been numerous studies examining the impact of CDHPs on use of health care services and
Feldman, Roger D., and Stephen T. Parente. "Enrollee Incentives in Consumer Directed Health Plans: Spend Now or
federal agency staff, and the news media; and sponsoring public opinion surveys on employee
that use $22 Any HSA amo the HS 5 whe A- de a n 20 lig ra ibl unts un ndomiz 14 data wer e health p used a ed cont lan e exclu tw the rolled trial as ien d ne tr d of do. dAn for u th ced sometime after 2009 an d e year its evaluatio among can indivi be nduals with . rolle Also, in th d over inc e st in t d t omes h h udy, some at dat e acco of $1 aunt a were 25, fa000 n milies w d use availabl or d mor i ee for n re e futu assigned to , average are years. Also, s many as spe plan nd 200,000 ing was s that even
Here are the key findings:
Introduction ........................................................................................................................................................ 4
much differenc outpatie network o nt ser f providers an v e between a hi ices other d t th han physician o e same gh-deductible health plan wit pharmffice acy be visits ($1 nefit m33). a hnager an HSA Whil (PBM and e th). ea lower- re w Based on eredeductible health plan reduct thesions in e similarities, office v the without an HSA. isits to primar sample way care s
higher-income workers. issues; Wor and kers with major survey less than s of $50, public 000 i attitudes. n annual income averaged 1.5 visits to specialists, whereas
2 Source: EB RI analysis of administrative claims data.
income workers reduced their use of specialist visits by
benefit issues. EBRI’s Education and Research Fund (EBRI-ERF) performs the charitable,
overall spending. The initial studies tended to focus on broad topics like who enrolls in a CDHP, how enrollees differ
Save for Later?" Forum for Health Economics & Policy 13, no. 2 (April 2010): 1-20.
$44 employees 2 lower an when d dep inclu enddin ents g .201 4 claims data, compared with Figur $ e 6 356 when 2014 data were excluded. In the case of the
requir if an em ed them ployer to pay contributes to $1,000 for o the HSA, it ut-of-poc is indivi ket expenses, dually ow w ne hich d a was nd ful a lymuch more portable if t si hgnif e worker icant dolla leaves his or r amounther than job. it
EBRI meetings present and explore issues *** with p<0.01, thou ** p< ght 0.05,leade * p<0.10 rs from all sectors.
physicians a further 80% limited to in nd specia divi lists, th duals enrolle e reductions d in eit in h spe er th ne P ding, PO o which r HSA wer plan. e stati stically significant, amounted to only $20 for
-0.6
those with $125,000 or more in incom educa e av tiona erage l, andd scientif 2.1 visits (Fi ic functi gons of the Instit ure 5). Similarly, ute. EBRI workers with -ERF is a tax-ex less than empt organizatio $50,000 in n
0.36 per person, whereas individuals with income of
Background on CDHPs ........................................................................................................... Figure 4 .............................. 4
Figure 10 Figure 12
from non 7 -enrollees, risk selection, and the impact of CDHPs on overa a ll use of services and spen 74% ding. More recent
? The HSA-eligible health plan was associated with a decline in (non-preventive) outpatient office visits for
Outpatient Physic T iao nt O al ffic A e v Vis er itsage AnnuaP l S rima pe ry C ndi are n Phg Health ys o icia nn H V sa is e vings ita s lth a C ccount. are Service Ss p, ecialist Visits
$75, would be to 000 ?$9da 9,999 y, and hi and $10 gher 0, than th 000 ?$1e mi 24,9ni 99 mum deduct income groible l ups, statistical signif evels in CDHPs today on icance was ce adjuste stronger d w for heinflatio n the 2 n0 . T 14h cla e RAND ims
Linear fixed ef fects regression models were estimated to examin EBRI regularly provides congressional testimony e the eff , an ects of the HSA-eligible d briefs policymakerhealth plan o s, member 72% o n use o rganizations, f health
supported by contributions and grants.
physician office visits and $7 for specialist visits per member per year. There was no statistically significant change in
annual income averaged Percentage of 8.8 prescript Sample Usi ion drug fills, wher ng Health Care Serv a ea as those wices, ith $12 by5,0 Ty00 or pe of more in Service, incom 2013 e averaged 12.5 fills.
$50,000 or more reduced their use of specialist visits by
Fendrick, A. Mark, et al. Hea Impact of lth Savi Impact of ngs Ac HSAcount-Eli -HSA Eligi-b Eligi le gible Health Plan ble High Health Plan Deducti on blNumber of e Healt on Likelihood h Plans Inpatient : Updating the Definition of
studies hav HRA-based The study fou e en Health Plans and xd t amine hat h d mor ighere-income targetHealt ed topic group hs Reimb s reduc , such as me ed u s rsement Arrangements (HRAs) pen di din cation g mor adher e than lo ence for indivi wer-incomde gro uals with ups i n chr thoe nic firs con t year dition ofs the
workers at all income levels, by but Ty the pe of declin See was rvice a over nd W twor ice as lar ker Inc gom e for e, workers a 2013 nd their dependents with
and the media on employer benefits.
data were included. Among all three groups, lower spending on prescription drugs appears to be driving the savings,
care services, spending, and experiment Finally, HSA-Eligible H the e revealed that mployer co ealth Plans a ntri hibute medication adherence. These gher nd He d a cost shar little alth S le aing ss than vings Accounts had a 50 grea percent models ter impact on (HSAs) of t coh .................................................................. ntrol fo e deductible the r time- poor tha iint nvariant perso o each n on the nonpoo HSA. n al char. racteristics—whether ......... 5
Source: EBRI analysis of administrativ e claims data.
spending on inpatient hospital services or emergency departments.
Preventive services increased with income as did some of the vaccinations and screening tests that were examined. In
100% Admiss of ioReceiv ns and ing Prev Emergency enti Department ve Services, by Visits, Employ by Employ ee Income ee Income
0.11-0. Preventio 12 per nperson. Lower . Center for Value-Based I -income work ners reduced th surance Design, eir use of University of Mic primary care offi higan, 20 ce visits 14. by 0.12 per person, but the
HSA-eligible *** p< h0. e0alth 1, ** p< plan. 0.05, * T p