Health Care Spending after Adopting a Full-Replacement, High-Deductible Health Plan With a Health Savings Account: A Five-Year Study

A high-deductible health plan linked with health savings accounts reduced health spending initially, and over a four year period, according to new research from EBRI.

This study reports experience over five years from a single large employer in the Midwestern United States that adopted a high-deductible health plan with a health savings account (HSA) for all employees. This study represents one of the longest observation periods reported with a full-replacement CDHP, and it is one of the few studies with a matched control group.
In the first year of the HSA, the employer’s aggregate health care spending was reduced by $527 per person.
Results show that spending was reduced significantly in the inaugural year of the HSA plan in medical, pharmacy, and total-claims categories. Further, the magnitude of the cost savings was greatest in this first year but the cost savings continued over the succeeding three years albeit at a slower pace.
The introduction of the full-replacement HSA plan reduced total spending by 25 percent in the first year. Each category of health spending experienced statistically significant reductions in the first year of the HSA plan with the exception of spending on inpatient hospital stays. Spending on laboratory services and prescription drugs had the largest statistically significant declines (36 percent and 32 percent, respectively).
When examining the spending components separately, only pharmacy and laboratory spending were statistically significantly lower throughout the entire four years after the HSA plan was adopted.
Reductions in pharmacy spending were large and mostly sustained over the four years after the HSA was adopted. In the first year of the HSA, pharmacy-spending reductions were 40–47 percent for individuals in all but the highest quintile of spending.
When spending by pre-HSA health spending quintile was examined, the largest spending effects in the first year of the HSA were seen in the third and fourth quintiles. The highest pre-HSA quintile group experienced spending reductions in the first year of the HSA that were not sustained. The second HSA-plan year showed total spending was reduced only in the second and fourth quintiles. By the fourth year, the HSA plan reduced pharmacy spending in the fourth quintile only while the third quintile continued to have reduced spending as compared with the year before the HSA plan was adopted.

Mullahy Reduct consumption the fo income a Appendix – Sample Matching Impact of HSA on He References Figures llowi ions 6–8 sh , J nd ch ng . ‘‘Much Ado in of year at t health ronically ow th health e summ -service hse e ill. O ab rvi employer’s out Two: alth Spending ces s n ary results e of utilizat at th the stre e ion disc Rec en fro a retion a d o fngths o and Statis tnsidering m of er separat tintro he nf20 d res th ductio 06 e equation Re e t data is th (either transf rict F ntical Analysis i gur of ions can te 1 or hde e s based ma e lar layed high-deducti be placed on tio gn e sampl or and Tw on 2006 h acce b elerat le siz o-P all HSA e ee alt a o d and t rt Model in wed rollover ) hplan care spendi in antici h e m is like ipati x o H l y amounts. Employers fealth E on healt ng. a result ofThe result the hc plan on o impendi fom th types. e etrics s are large ng .’’ Paul Fronstin is director of the Health Education and Research Program at the Employee Benefit Research Health Care Spending after Adopting a Full-Replacement, Figure 6 Institute (EBRI). M. Christopher Roebuck Baselin is e Samp president le Ch and aract CEO erist of Rx ics ( Econ in 2006) omics, LLC. This Issue Brief was written are not shown change deductible One of Journ th by 2 re to e quir a tin we h l of H 0e 0ted to 6 his aknesses is th HSA spe epla alth Ec make n pla di n n ng qu design . onomi unintiles: Percentage and Dollar Impact of CDHP on Medical Spending, at only o used elected csbalan 17, 1998: Figure nby e y ces most e availa 6 shows m ar of 247–81. employe da ble ta to w was edi eFigure 3 s, ca exami orkers upon l spe namely n ndi ed a ng (such as in $2 jo ,150 fteb r the CDH disruptio for singl pati P e ns , ent, out enroll were ado but can ees p allo atie a pnd ted. w s nt $4, , emer uch acc 300 gency for e ss for families. - The Bundorf, M.K. “Consum Coarsened Exact Matchi er-Directed ng techniHeal que th Plans: Do was used to cThey Deliver? reate a compar ” ison grou Research Sy p (Ia nthesi cus, King, s Repor and Porro t No. 24, O 2011ctober ). One- Figure 4 shows the results for the DiD regression model presented in equation (1) in the Appendix. Separate estimates High-Deductible Health Plan With a Health Savings Account: with assistance from EBRI’s research and editorial staffs. This work was conducted through the EBRI Center for Average A Relative to 2006, by Year and Baseline Cost nnual Increases in Health Spending, 2007 ?2010 quali depa Wage frtment, ie ad me nd inc di aome cal e nd lab spen data xpens are es din a not nd g),availa Fi for c gure ertain ble 7 s fo h premiums r ows pharmac this analysis, . Givy sp ebut n th HSen Aprior at din most e g studies by Com quinti mployers paris have les, on shown Figu do nre 8 ot allow th shows tot at far funds smaller al spe to be use increas nding by d e s afin ter cost to-one 2012, Robert pairing betw Wood Johnson Foundati een study group subjects o and n. comparison group individuals was conducted based upon calendar are presented for the various components of health spending, as well as for total spending. Only the findings from the N A Five-Year Study Research Finally, air, K. this V., J on a.n Health Park, P alysis is Be . W una nefi olf ble, J ts e to Innov .Jdeterm . Saasti son (EBRI CRHBI). The f en in, R e the .R .cause Allen, of and R. Gangul the large ollowing organiza increase y. “Con in sumer health tions -spen Driven Health Plans: Impact on provided ding in th the e com funding parison for grou EBRI p Group Group 2006 Spending Quintile (Range, Mean) 2007 2008 2009 2010 2007 2008 2009 2010 job disru shifting quintiles. T hav ption, hee c more detaile workers o aused signi nly fic d ant get val spendi reductions ng esti ue from mates shown in the healt account h-ser i v if n ices they Figur utili e use th zation 3 are e not use (Ne funds whouse for d in t quali a hn ese fi d fied medical se the gur Ine sura s be nce cause in rvices. Experim divie du nt als in year 2006 (i.e., pre-period) values of the following variables with pre-specified cut points (in parentheses): gender; regression related to the impact of the HSA on health spending (the coefficient on the interaction term between the July 2013 • No. 388 CRHBI: Ameri 1 (0-$161, Mean=$52) can Express, Blue Cross Blue Shield Asso -7%(N=10, ciati509) on, Bo14% ei(N=10, ng, CVS 509)Care-6% mark, General 18% Mills, Healthways, in Current Study Data and Meth 20Utilizati 07. As a o n an result d Expenditures for of this increase, first-y Chronic ods ear savin Disease Sufferers.” gs as a result of th Journ e ado al of Occupati ption of the HSA onal & plan Environ may be mental overstated. Buntin, M.B., A.M. Haviland, R. McDevitt, and N. Sood. “Healthcare Spending and Preventive Care in High- the lo Group wer 1993 -spe ). ndin The g d a qui ta ntiles also su diggest d not oft that en th incur ere was expense an “is nex in e perie ach o nce f th factor” e more in det the aile cod cate nversion gories to a . ccount-based Additionally, oplans, nly the age (18, 25, 35, 45, 55); geographic region; and Charlson Comorbidity Index (0, 1, 2, 3)—a proxy measure of health By Paul Fronstin, Ph.D., Employee Benefit Research Institute, and M. Christopher Roebuck, HSA and Ye 35% ar indicators) are shown. The introduction of the full-replacement HSA plan reduced total spending by 2 ($162-$473, Mean=$303) -11% Mean -25%**Mean -5% p-value 2% IBM, John Deere & Co., JP Morgan Chase, Mercer, and Pfizer. Any views expressed in this report are those of the On the Medicine other hand, 51 (5), May 2009 an HSA is a tax :594-603. -exempt trust or custodial account that an individual can use to pay for health care 2 Deductible and Consumer-Directed Health Plans.” American Journal of Managed Care 17 (3), March Pharmacy- and medical-claims data and insurance-enrollment information were obtained from a large employer located findings as evidence from t 3 ($474-$1,064, Mean=$727) d by he re eros gression ion of fi rst-year related to t spen he ding impact o reductio -29%*** f thn e sHSA on over the health suc 8% ceedi spen ng din years. g (the -10% coefficient on -14%* the interaction status (See Charlson Pompei Age:, and MacKenzie 1987; Deyo, Cherkin, and Ciol 1992; and Quan et al., 2005). Ph.D., RxEconomics 25 percent in the first year (2007). Each category of health spending experienced statistically significant reductions in authors and should not be ascribed to the officers, trustees, or other sponsors of EBRI, Employee Benefit This study adds to the consumer-directed-health-plan literature by reporting changes in spending during the four years expenses. In 4 ($1,065-$2,569, Mean=$1,691) order to qualify for tax-free contributio -29%*** ns to an HSA, the -10% individual mus7% t be covered by a h -4% ealth plan that in th2011:222-30. e Midwest, covering the five-year period from January 1, 2006, through December 31, 2010. On January 1, 2007, terms between HSA and Year) ar Average (years e show)n. 29.94 30.91 0.00 Additionally, pre-period values of the eight variables on use of health care services (described below) were also Research In the first yearstitute-Education of the HSA plan wit and h the Resear excech ption of spen Fund (EBRI-ERF ding on ) in , or pati their ent hospital st staffs. Neither ays. Spen EBRI din nor g on la EBRI-ERF boratory lob bies after Newhouse, an HSA 30% 5 ($2,570+, Mean=$6,157) J. and the Insurance Experi plan was adopted and showiment Group. ng that, in-13%* onFree For All? e employer, s -2% p Lesson en 29% dings from reduc the RAND He tions 5% eroded ov alth Insur er -1% time an ad ncte hat has a mi Health innimu suran m ce , annua covera l, sta ge <18^ is tutory de complex ducti for b nle on-he ($1,2 alth 50 fo mar se intenanc 0.lf38 -only cov e orga erage nization 0.38 and (non-HMO) $2,500 0. for 96 fami plans ly because coverage th in eir 2 013), 3 this employer adopted an HSA plan for all active workers and their dependents. Both before and after 2007, the balanced across the study and comparison groups. Using this process, 10,509 members (79 percent) of the study or takes positions on specific policy proposals. EBRI invites comment on this research. services and prescription drugs had the largest statistically significant declines (36 percent and 32 percent, reductio Experimen ns var 1 (0-$161, Mean=$52) ied t. Cambridge (MA): Harvard by pre-replacement levels of Universi health -$38 t sy Pr pending. ess, 1993 The $74 . data suggest that -$32the highest users $90 were least 18-24^ 0.03 0.03 0.96 Charlson, M.E., P. Pompei, K.L. Ales, and C.R. MacKenzie. “A New Method of Classifying Prognostic Comorbidity in and Figure Health Care Spending after Adopting a Full-Replacement, premiums all health 6 sho do w care not s the im account servic pact o es except preven for f a nearly doptin all g a of f ull-r tive the e spending services mu placemen by t HSA o st be subject to the deductible. plan partici n spen pdin ants, g oas n all th h ey e alth c do in The HS are ser HMOs. v A is owned by Iices other n plans with than the choices employer offered PPOs as the network type. Prior to the adoption of the HSA plan, deductibles ranged from $500- Introduction cohort were successfully matched (1:1) and employed in the ensuing analysis. Given the potential for dual coverage 2 ($162-$473, Mean=$303) -$94 -$205** -$44 $18 respectively). 25-34^ 0.09 0.09 0.96 affected and that moderate users were most vulnerable. The data did not allow for distinguishing discretionary from Longitudinal Studies: Development and Validation.” Journal of Chronic Disease 40 (5), 1987:373–83. indivi pharmacy s of provid duale an 25 rs, he % pd enis complet din alth g i , n by 2 sure a 0ly n 0ce 6portabl pr co e-HSA verea. T g he eh ier sa lla e is no th-spe den w use ndi ithn -g quintile. mu it-or-lose-it ltiple cost-shari T rul he e bi ggest associated ng mef echanism fects witof the h an s withi HSA, as HSA n th we e any money re fo same und in plan; le its ftfirst in the $1,500 per person, and co-i 3 ($474-$1,064, Mean=$727) nsurance ranged from 80 to -$365*** 100 percent$98 for in-network s -$125 ervices, and fr -$183* om 60 to 90 percent High-Deductible Health Plan With a Health Savings Account: Health i (e.g., Me nsuran dicarece covera ), and conse ge with hi 35-44^ quently uno gh deduc bserved healt tibles (gener h-sal ervices uti ly 0.in 18exce lizat ss of ion in $1, 0.18 00 the 0 for avai sin lable g0. les a 96 claims nd twic data, all e that members wer for families) is e C Parente opyright Inf , S.T., R. Feldm ormation: an, This and S. Chen report is . copyright “Effects of ed by a Consumer the Employe -Dri e ven He Benefit alth Pl Research an on Ph Institut arm e (EBRI aceutic ). al Spe It mayn be ding necessary services utilization. Th 22% is is essential for understanding the value of account-based, high-deductible plans. 4 ($1,065-$2,569, Mean=$1,691) -$576*** -$207 $132 -$71 account at year: T differential he HSA the en cost reduce for d of th high d spen -per e year automatica formance ding by 29 vs. per standar llc y stays with ent amon d netwo g t in hrks dividuals in e HSA. of provider Both th the i e thir s; in-network ndivi d adual nd fo and t urth vs. out-of-net quinti he emplo les, y work and er are by cost allo 13 per shar wed to c ing; ent for out-of-network services. 45-54^ 0.26 0.26 0.96 In the secon d year of the HSA plan, total spending was 8 percent lower relative to the year before the HSA was gene requir rally asso ed to be less than ciated wit 65 y h savings-acco ears of age as unt options that of the last da can be used to y of the five-year study cover out-o period. f-pocket costs for health care and Utilization.” Health Services Research 43 (5), October 2008:1542–56. used A Five-Year Study without permission but citation of the source is required. Claxton, G., J. Gabel, B. DiJulio, J. Pickreign, H. Whitmore, B. Finder, P. Jacobs and S. Hawkins. “Health Benefits 5 ($2,570+, Mean=$6,157) -$515* -$99 $201 -$21 contribut among t reference hose i e to prican HSA, in the ng (fi hi xeghest d with pay 55-64^ annual m spendi ents contri for ng quin services butions subject tile. By wit the h ben foe urth to ficiaries 0. statutory li 06 year payi of th nmits ($ g e HSA, o th 0. e 06 dif 3,2 fnly th ere 50 nc foe middl e r self for 0.96 -only co any e quint high verage e ile r char was e an ges x d $ peri ); 6 a,encin nd 450 for g EBRI Employee Benefit Research Institute Issue Brief (ISSN 0887 ?137X) is published monthly by the Employee Benefit Research Institute, 20% adopted. However, when examining the spending components separately, only pharmacy and laboratory spending were Source: EBRI estimates based on administrative claims data. services. These are often referred to as consumer-directed health plans (CDHPs). Accounts are typically one of two In 2007: Premium Increases Fall To An Eight-Year Low, While Offer Rates And Enrollment Remain Stable.” 1100 13th St. NW, Suite 878, Ma W le^ ashington, DC, 20005-4051, at $300 0 p.5 er 4year or is inclu0 de .5d 4 as part of a m 0.9 ember 6 ship subscription. Periodi- Notes: Medical spending includes: inpatient hospital, emergency department, outpatient hospital, outpatient physician, substance abuse and Claims data were available for analysis for one year prior to the adoption of the HSA plan and for four years post- family cover By Paul Fronstin, Ph.D., Employee Benefit Resea spendi tiered p ng re harma du agc ctions relativ ey in payments 2013, plu e wher sto the y catch-u e the ep ar pla co before th cn ement tributions o e HSA f a for me was ado idicatio ndivirch Institute, and M. duals a n pin ted: ag tier eSpen s 55 is din associat and g ol reder du ed ctions for t ).with Christopher Roebuck, a difhis ferent group w benee fici rear 14 y per cost. - Recommended Citation: In order to measure the over all effect Paul Fronsti of th n e switch and M. to a Christoph n HSA eplan r Roeb on uck, health “Health care spendi Care ng, Spendin this a gn after alysis bui Adopti lt ong n the a Parente, S.T., R. Feldman, and J.B. Christianson. “Evaluation of the Effect of a Consumer-Driven Health Plan on statistically significantly lower, and those were the only components of health spending that remained lower throughout types: health reimbursement arrangements (HRAs), which are offered, owned, funded, designed and controlled by cals postage rate paid in Washington, DC, and additional mailing offices. POSTMASTER: Send address changes to: EBRI Issue Brief, 1100 mental health, and laboratory. CDHP=Consumer-Directed Health Plan. Presented are coefficient estimates of the (difference-in-di Policyholder 0.40 0.39 0.27 fferences) Health Affairs 26 (5), September-October 16% 2007:1407-16. adoption. All workers were given a choice of two deductibles: 1) $1,250 per individual, $2,150 per family; or 2) $2,150 cent. In The addition dollar o terms, by f a large dedu the fourth ctible an year d an of account the HSA, th with e em middl ployer-provi e quintile ded was spen funds to din tg his $1 mix 83 less certai per nly year t increh ases an tth hee y Full-Repl Ph.D., RxE 13th Sace t. NW m c , ent, onomics Suite 8 High-D 78, W ashi eductible ngton, DC, He 200alth 05-405 Plan 1. CoWi pyright th a 20 He 13 alth by Em S plo avings yee BeAccoun nefit Reset: arcA h IFive-Year nstitute. All riS ght tudy,” s reservEB ed.RI No.Issu 38 8.e generalized lin effect of CDHP in each of four follow-up years from generalized linear models with gamma family and log link. Models included i ear model (GLM) with a gamma distribution and log-link presented in Lo Sasso, Shah ndicators for , and Frogner Medical Care Expenditures and Utilization.” Health Services Research 39 (4); August 2004, Part II:1189–1209. the entire four years after the HSA plan was adopted. Furthermore, the reduction in spending on laboratory services Household size 3.25 3.26 0.38 employers; or 15% health savings accounts (HSAs), which are offered by employers, owned by employees, designed and The earliest se male, policyholder, five age groups, three geographic regions, four years, CDHP, and four CDHP times year interaction terms (pr ries of studies examined a single employer that introduced an HRA in 2001 (see Pareesented), as nte, Feldman, and Brief, per individua no. 388 l, , $4, July 300 per fa 2013. mily. Workers overwhelmingly (89 percent) selected the higher-deductible option, most were in cognitive the year before demands for n the ew users HSA was a of accou dopted. nt-base d, high-deductible health plans. However, it is reasonable to expect (2010). As they note, this m Years odel of tis fr enure equently applied to healt 14.02 h care data be 13.33 cause of its tracta 0.01 bility and ease of appeared to decline as the significance levels dropped (though they remain statistically significant). By 2010 (the fourth effe Claxton, G., ctively co well as Charlson Comorbidity Index, years of tenure, and household size. Statistical significance based upon robust standard er n M. trolled by Rae, N. P the achal, A U.S. governm . Danico ent, an , H. Whitmore, K. Kenwar d tax advantaged. Employers hav d, and A.Ose ei now -Anto. “Health Benefits In been usin rors (clustered g CDHPs for over a 2012: Christianson 2004; Feldman, Parente, and Christianson 2007; Parente, Feldman, and Chen 2008). Feldman, Parente, likely the result of payroll de Nortduction heast^ for premiums of nearly $ 0.00 300 annually fo 0.00r employee-only 0.96 coverage in the lower that Parente, S behaviors .by policyholder) denoted as follows: *** p<0.01; ** p<0.05; * p<0.10. More information about the econometric model can be foun T., R. Feldm will changa en, asand Y. Xu users become . “Impact of Full more familiaRe r with plac ho ement w to use with Con the plan sum for er Driven Health Plans meeting health-car d in the e-servi on Health ces estimation as a result of advances in personal computing (Manning 1998; Mullahy 1998; Manning and Mullahy 2001). 10% year of the HSA plan), total spending was still 6 percent lower relative to the year before the HSA was adopted. Moderate Premium Increases For Employer-Sponsored Plans; Young Adults Gained Coverage Under ACA.” decade. In 2012, 22 percent of smaller employers, 36 percent of larger employers, and 59 percent of jumbo employers 9% AT A GLAN CE Report Availability: 10% appendix and in Fronstin This re , Se port pulveda and Roebuck is available on (2013 th ).e Internet at www.ebri.org Figure and Ch7 sho ristians ws the im on (2007) exa pact o Midw m f the HSA o es ined three y t^ n pharmacy s ears of data p afte endin 0. r the HRA g 97 . Reducti was adopt ons in 0.97pharma ed, acy spen nd 0. foun 96 din d th gat were la the HRA rge an result d mostly ed in deductible pla Care Cost an n, compar d Use ed with no worke of Preventive Servic The Emr plo contri yes.” ee Benefi bution Insur t Research Institut a toward the nce Markets premium an e (EB d Companies: RI) was founded in 1978. Its for the higher Analyses and Actuarial deductible plam ni. ssion is to The needs over time. 1 offere Health Aff d some fo arirs m 31 (10), October 2012: 2 of a CDHP, and nearly 1 in 324-33. 5 workers wer e enrolled in one (Claxton et al., 2012). South^ 0.03 0.03 0.96 7% contribute to, to encourage, and to enhance the development of sound employee benefit sustained ov higher spendi er ng o the four n hospital year and physician s after the HS services, an A was adopte d d. In had the first no impact year of the HS on prescriptio A, pharmacy n-drug spen-di spen ng, alt ding hou reduc gh ttions he employer The basic als model o contribut with a lo edg-l toi the nk is HSA and deposited $700 for employee-only coverage and $1,300 for family coverage, Computations 1 (1), 2010:4–14. 6% Figure 5 shows the total amounts saved as a result of introducing the HSA. In the first year of the HSA, spending was 5% West^ 0.00 0.00 0.96 Who we are Figure 7 programs and sound public policy through objective research and education. EBRI is the only Table of Contents were findings The data 40-47 m also ay have bee peshow rcent that for i n ndivi in hea fluence lth-services duals i d by n alits small l but spendi theng sample highes appear of t quintile ed HRA enroll to vary of spe by een s baselin ding, (n=42 an e 9d t l). evel he of high hee alth st-spe spen ndi ding ng qui befor ntile e the HSA 5% regardless of ? This which study reports experience over five deductible was chosen. years from a single large employer in the Midwestern United reduced by $527 annually per person. Because second-year spending increased faster in the HSA group than in the Davis, K. “Consumer-DirCharls ected Health Care: W on Comorbidity Index ill It Improve Health ^ 0.09 System P 0.08 erformance?” 0.70 Health Services Advocates of CDHPs contend that providing participants with an account and subjecting claims to high deductibles Percentage and Dollar Impact of CDHP on Pharmacy Spending, private, nonprofit, nonpartisan, Washington, DC-based organization committed exclusively to 1 Quan, H., V. Sundararajan, P. Halfon , A. Fong , B. B urnand, J.C . Lut hi, L.D . Saunders, C .A. Beck, T .E. Feasby, and experienc Introductio plan design ed a n w .................................................................................................................. a24 s introdu percent ced. de cline High in util pharm izers (the acy spen fifth din preg -. HSA By the quintile fourt ) h had year of the lowes the HS t reduction A, individua ........................................ in ls i use n the of medical four lo west 4 Source: EBRI estimates based on administrative claims data. States that adopted a high-deductible health plan with a health savings account (HSA) for all employees. public policy research and education on economic security and employee benefit issues. comparison group, spending was only $169 lower annually per person relative to the year before the HSA was adopted. befor Res e ins earch uran 39 (4), August 20 ce benefits are trigger 04, Part II:12 ed Relative to 2006, by Year and Baseline Cost will induce 19-34. enrollees to make more cost- and quality-conscious health care Parente, Feldman, and Xu No (201 tes: 0) were able to examine four employers that fully replaced their health plans with The study group consisted of 13,278 active workers and their dependents who were continuously covered by the spendi services W.A ng quin . in Ghali. the tiles e first “Coding Algorithms fo year xperia enc fter ed a the 2 HSA 1-41plan pe r Defi rcent r was ning Comorbi in etrodu ductio cn in ed, d pharmacy an ities in ICD-9-CM and IC d the sho spen rtest din durat g, ra ion nging from of D-10 Administrative D reductions $21 to $ in pharmacy 170. Ta hta.” e EBRI’s membership includes a cross-section of pension funds; businesses; trade associations; wher Background e Y is an This study anual s nd Prior p re en Lit pdin resents one era g for ture person ............................................................................................... of the i in year longest o t, Ye bar servation is an indic per aitor variabl ods reporte e d for with each year a full-rep from lacement .............................. 2007- CDH 2010 P(vs , an . 20 d it06 is ), 4 In the third year of the HSA, spending was $75 lower due to the HSA, but this amount was not statistically different 0% 2006 Spending Quintile (Range, Mean) CDHP=Consumer-Directed Health Pla 2007 n. 2008 2009 2010 decisions (Herzlinger 2002). Skeptics, on the other hand, caution that patients lack comprehensible, timely, and CDHPs and found that two firms experienced savings (4.11 percent and 0.94 percent), but that the other two firms labor unions; health care providers and insurers; government organizations; and service firms. HSA Plan Group Comparison Group employer Medical C during t are h43 (11), November 2005:1130–9. e five-year analysis period. Data from a second employer were used to create a comparison group. high spendi esng. t-spe Th ndi is ng is not quintile surp ex risi Vperie ang lues are proporti given nced a this 5 ons unl group’s perce ess denoted otherw nt re need duct fo ion in r medi ise. pcal harma serv cices y spen refdin lecte g id n the in th fourth eir poor ye ear of t r healthe h status HSA by Deyo, R.A., D 1 (0-$161, Mean=$52) .C. Cherkin, and M.A. Ciol. “Adapting a -47%*** Clinical Comor -20% bidity Index for Use with I -41%*** -41%*** CD-9-CM one of the few studies with a matched control group. HSA is an indicator for whether the person is in the full-replacement HSA plan cohort, and the coefficients on the Current from total spending Study Data and prior to Met h the a ods ................................................................................................ doption of the HSA. In the fourth year of the HSA (2010), spendin............................ g was $132 lower 6 trustworthy information that is critical for them to make informed decisions (Davis 2004). Skeptics are also concerned Variables used in coarsened exact matching process are denoted by ^. experienc ed in 2 ($162-$473, Mean=$303) creases in spending (3.57 percent and 1. -44%*** 28 percent)-36%*** . Unfortunately th -21% ere was no control gro -25%* up to This larger plan sponsor, with 82,644 continuously enrolled employees and dependents, maintained PPO coverage that relative to the Charlson theC year b omorbidity efore In th dex e HSA score. was a Alsdoopted, , while wh prescr ich amounted iption-dru to g sp $62, endi althou ng reducti gh the im ons wer pact o e pro f m the HSA inent ov pla er nt he was not four- Administrative Databases.” Journal of Clinical Epidemiology 45 (6), June 1992:613–9. interaction terms between HSA and Year reveals the difference-in-differences (DiD) estimate of the effect of HSA per person as a result of thRe e HSA relative ported p-values a t reo the year from Kruskal-before the HSA Wallis rank test of diffe was adopt rences in mee an d. s a cross groups. that higher cost sharing might lead to les s use of preventive care, primary care and other necessary health care services, Results ....................................................................................................................... 3 ($474-$1,064, Mean=$727) -40%*** -17%** -31%** .......................................... -21%* 6 EBRI’s work advances knowledge and understanding of employee benefits and their determine the effect of the plan-design change. Source: EBRI estimates based on administrative claims data. was not H ? In the first year of the SA-eligible throughout th HSA e fi,ve-year study the employer’s aggr period. Two PPO egate health c options were are spending was offered: on reduced by $527 per e with no deductible statistically significant year observation period, for the individuals se were most in this spen consistent ding in qui gr noups tile. with pre-HSA spending patterns representing moderate enrollment on 4 ($1,065-$2,569, Mean=$1,691) health care spending. The coefficients o -42%*** f interest ( ß -36%*** ) can be interpr-40%*** eted directly as -30%*** a multiplicative effect 3 especially among individuals of lower s importance to ocioeconomic s the nation’s econo tatus (Havila my nd e among policy t al. 201 m 1a; akers, Newhouse the news an media, and d the Insur the public. It ance Descriptive Findings ........................................................................................................................................... 6 Endnotes Feldman, R., S.T. Parente, and J.B. Christianson. “Consumer-Directed Health Plans: New Evidence on Spending throughout th person. e study period, and one with a deductible that had increased nearly 50 percent throughout the study users of healt 5 ($2,570+, Mean=$6,157) h services (the third and fourth spending -24%*** quintiles). Mo -19%*** derate health-services -9%* consumers -5%may represent a does this by conducting and publishing policy research, analysis, and special reports on of the HSA-plan design on total costs, as full replacement eliminates individual-level selection as a potential source of What we do Figure 5 Experiment Group 1993). Lo Sasso, Helmchen, and Kaestner (2010) studied the impacts of CDHP features on health care spending, examining a Figure 8 shows the total impact of the HSA on health spending by pre-HSA spending quintile. Similar to the results Impact of HSA on Health Spending ..................................................................................................................... 9 and Utilization.” Inquiry 44 (1), Spring 2 employee benef 007:26–40. its issues; holding educational briefings for EBRI members, congressional and period. Both 1 (0-$161, Mean=$52) plans covered preventive services in ful -$24*** l and used 20 pe -$10 rcent coinsura-$21*** nce for most other covered serv -$21*** ices. locus where high-deductible plans and an HSA account will have a greater impact. The data do not permit any bias. Annual Reduction in Health Care Costs Per Person Due to a sample of individuals that include d multiple employers offering mostly HRA plans on a full replacement basis (about 1 2 ($162-$473, Mean=$303) ? Results show that spending federal agen was re cy duced staff, and th significantly i n/a e news media; n the inaugur n/aand sponsoring public opinion survey al year n/a of the HSA plan in n/a s on emplo medical, yee shown in Figure 6, the HSA had stronger effects on spending in its first year. The largest spending effects in the first See Figure 6 in www.mercer.com/pressrelease/details.htm? Figur idContent=1491670 e 2 This an Discussion alysis c ............................................................................................................................... onstructed a comparison group by matching on pre-period values of a rich set of cova ............................ riates to the study 10 assessment of the necessary vs. discretio Adoption of nary need HSAfor Plan, prescription Relative -dr to ug 2006 spendi , by ng. Y ear This Issue Brief examines the trajectory of spending on health care services over four years after an employer 3 ($474-$1,064, Mean=$727)benefit issues. EBRI’s Ed -$109***ucation and Re -$46** search Fund -$185** (EBRI-ERF) performs -$58*the charitable, 2 percent offered an HSA). This study included data on beginning-of-year account balances and rollover amounts, and Fronstin, P., M.J. Sepulveda, and M.C. Roebuck. “Consumer-Directed Health Plans Reduce The Long-Term Use Of pharmacy, and total-claims categories. Further, the magnitude of the cost savings was greatest in this year of the HSA were seen in the third and fourth quintiles, which exhibited moderate levels of spending. Spending was The model als group. More inoformation controls for on Me this a a num n H pr eber ocess i alt of h C enrollee charac sa provided in t re Spendin hte e A gristics. Th A ppe mo ndi un xe . ts Charlson Com among CDo Hrbidity P and In dex and demo graphic Study Limitations and Conclusion .......................................................................................................................... 12 eliminated its 4 ($1,065-$2,569, Mean=$1,691) traditional health insura educa nce tiona plans a l, and scientif nd re -$240*** plic func aced tti hons of the Instit em wit -$206*** h high-de ute. EBRI ducti -$231*** ble -ERF healt is a tax-ex h pla-$170*** ns with empt organizatio health n 2 $600 thus was able to examine the role of the account. Although the study did not include a control group or pre-CDHP data, Because this analysis impose continuous eligibility in the HSA plan cohort and comparison group, the sample is one year Outpatient Physician Visits And Prescription Drugs.” Health Affairs 32, no. 6, June 2013:1126 -34. first year but the cost savings continued over the succeeding three years albeit at a slower pace. 31 perc Prior work ent lo has wer in found thte fou hat CrD th HPs quintil reduce e an spen d 29 perc ding ent l prima ower in rily amon the g tlo hiw- rd quinti and mediu le. Inm comparison, -risk people s(see pendin Bun g was dorf 201 15 per- 2 for 5 ($2,570+, Mean=$6,157) Comparison-Group Members, b -$331*** y Year and Ty -$271*** pe of Healt-$129* h Care Servic-$62 e characteristics (age, gender, and geogra supported b phic region y contributions ) were and g included, rants. and three other variables were constructed for use savings accounts (HSA plans). Prior analysis of the impact on health services utilization was recently published in Appendix – Sample Matching and Statistical Analysis .............................................................................................. 14 the investigators executed an instrumental-variables analysis (and included firm-fixed effects), and concluded selection older each year Source: EBRI estimates based on administrative claims data. , and increases in use of health care services will naturally increase over time. As a result, the Charlson cent lo a summary). wer in t This he fifth suggests quintilthat e, 14 in perce dividual nt l s owho wer exceed in the se their cond qu deduintil ctibele , and 1 s will n 2o perce t chan ng t low e the eir i r se n th rve ic lowest es cons qu umintil ption e as as covariates i Group/Cos n the multivar t Me527 asu *** re iate models: 1) an indicator of whether or not the mem Yearber was the policyholder (or Fronstin, Sepu Notes: CDHP=Consumer-Directed Health Plan. Presented are coefficient estimates of the (difference-in-differences) effect of CDH lveda and Roebuck (2013). Spending on inpatient and outpatient services, office visits, sub P in each of stance abuse Haviland, A.M., N. Sood, R.D. McDevitt, and M.S. Marquis. “How Do Consumer-Directed Health Plans Affect References ? The intro ............................................................................................................................... duction of the full-replacement HSA plan reduced total spending by 25 percent in th ........................... e first year. Each 15 bias was of minimal concern. It is worth noting, however, that the study examined a non-traditional CDHP design. Comorbidity Index will increase over time as well. Annual average Charlson Comorbidity Index scores for both the HSA plan Results 2006 2007 2008 2009 2010 (thou a result gh of not st c four follow-up years from generalized linear models with gamma family and log link. Models included indicators for male, policy han atistically signi ges to health finsuranc icant) EBRI Issue Briefs . e designs that is a monthly add large periodi deducti cal b with les in incom -depth binat evalu ion atiw on of em ith spen holder, five age plo ding yee benefi or savin t iss gus es dependent), 2) household size (i.e., the number of individuals on the policy), and 3) the number of years of tenure with and mental health, prescription drugs, laboratory services, and emergency services are examined separately over a $500 Vulnerable P groups, three geographic regions, four years, CDHP, and four CDHP times year interaction terms (presented), as well as Charlson opulations?” Forum for Health Economic Figure 4 s & Policy. 14 (2), April 2011a:1-12. category of health spending experienced statistically significant reductions in the first year of the HSA plan with Hospital in Endnotes HSA ...................................................................................................................... G pati roup ent and outpatient sur and trends geries wer , ase well as exclud e crit d from ical anal the ys de es of ducti em bplo le, yand ee be patient nefit polici cost-s ...................................... es and pro haring for posals pr.e EBRI scription 16 cohort and control groups are available from the corresponding author upon request. accounts. Our The observations in this analysis are not inconsistent with that point of view. For example, high utilizers, the employer. The e term is an idiosyncratic error. Comorbidity Index, years of tenure, and household size. Statistical significance based upon robust standard errors (clustered by policyholder) Percentage Impact of CDHP on Health Care Costs, five-year period that includes both pre- and post-HSA-plan experience for active workers and their dependents. The Total $1,756 $1,872 $2,288 $2,648 $2,792 Notes is a monthly periodical providing current information on a variety of employee benefit the exception of spending on inpatient hospital stays. Spending on laboratory services and prescription drugs T drugs he se coul cond HS d not be A-pl pa an year sh id from the owed account total sp . Overall, ending was r the study concl educed only i uded t n the hat e second a ach $1 n incr d fourth ease quin in thtiles. By t e HRA led to he fo a urth Descriptiv denoted as follows: *** p<0.01; ** p<0.05; * p<0.10. More information about the econometric model can be found in the appendix e Findings and in defined as pre-HSA high spenders (fifth quintile), had the lowest reductions in medical-services spending and the 3 Inpatient hospital 118 291 307 410 399 Haviland, A.M., N. Sood, R.D. McDevi Relative to 2006, by Year and Type of Health Care Service topics. tt, EBRIef and M.S. is a week Marqly uis. rou “ndup of EBRI r The Effects Of C esearch onsu and mer-Directed insights, as well Health as updates on Plans On analysis makes use of a matched control group for active workers and their dependents from a second large employer Five health- publications Fronstin, Sepulveda and Roebuck (2013). services utilization measures and three preventive screening indicators were derived annually for each individual: had the largest statistically significant declines (36 percent and 32 percent, respectively). year, the HSA 92 cent increas plan e in tota had no l spe effect on nding on spendi health ng amo care service ng thes highest- spen . ding 2006 quintile. The HSA plan reduced A number o shortest durati Em afer hg e o enc alth-s n of y de pha pen par rmac dtin men g measur yt benefits es w reductions ere derive tha d n an onually t62 her ut for ilizer eac 6 gr h 3 in oups. dividual. 82They included spen 98 din 101 g on 1) surveys, studies, litigation, legislation and regulation affecting employee benefit plans, while Figure 1 contains the baseline (2006) sample characteristics of the 10,509 individuals in both the full-replacement HSA Marginal effects could not be estimated because of the non-linear function being non-differentiable at the mean of the regressors. Figures Episodes Of Health Care.” Forum for Health Economics & Policy 14 (2), September 2011b:1-27. that continued to insure most of its workforce through traditional network plans without high deductibles or accounts Cost Measure (Dependent Variable) 2007 2008 2009 2010 1) number of inpatient hospital admissions, 2) number of inpatient hospital days, 3) number of emergency-department visits, $400 pharmacy s Out pp eat ndin ientg h ios n th pite al,fo am urth bula quinti tory, a le, nd but surgi not e cal nough to r 532 educe to 58tal spe 9 ndin 741 g. Total spendin 840 g also 8 was not 84 lower EBRI’s Blog supplements our regular publications, offering commentary on questions inpatient hospital admissions, 2) outpatient surgery, 3) emergency-department visits, 4) outpatient physician’s plan and the comparison group. Few statistically significant differences were detected between the population enrolled Total -25%*** -8%* -4% -6%* (preferr ?ed pro Whev nider or examining ganizattion (PPO he spend) ing placom ns). T ponent his study s sep repr arat esents o ely, only ne of th pharmacy and e longest observatio laboratory spen n periods ding repo wer rted e Lo Sasso, Sha Figure 1, Baseline h, and Sample Frogner (2010) fo Characteristic und that HSA s (in 2006) enr ........................................................................... ollees spent 5–7 percent less compared with traditional ............. -plan 6 4) number o Out f o pat uitpatient physi ent physician'c sian’s o officeffice/clinic visits, and 5) 4 number of p 86 48r6 escriptions filled, on a 30-day adjusted basi 611 663 695 s. The received from news reporters, policymakers, and others. EBRI Fundamentals of Employee in the first and second quintiles. Only the third quintile continued to have reduced spending in the fourth year of the office/clinic visits, 5) substance-abuse and mental-health treatments, 6) outpatient pharmacy, and 7) laboratory Herzlinger, R. “Let’s Put Consumers In Charge Of Health Care.” Harvard Business Review 80 (7), July 2002:44-55. in the full-replacement Inpatient hospital HSA plan and the comparison gro Figure 8 up. This -33%was an int 6% ended eff30% ect as a result 11% of matching. Substance abuse and mental health 21 20 25 30 33 Study Limitations and Conclu Benefit Programs sion offers a straightforward, basic explanation of employee benefit programs in with a fullstatistically -replacement si CDH gnific Pantly lower , and it is on thr e of ou the ghout t few st he udies entire with four a matched contr years after the H ol grou SA pla p. n was adopted. enrollees, with most of the relative reduction in spending occurring during the first year enrolled in the HSA, and that preventive screening indicators included three yearly cancer screenings recommended by the National Committee for Quality HSA plan as compared with the year before the HSA plan was adopted. In the third quintile, spending was 15 percent Figure 2, Mean Health Care Spending Amounts among CDHP and Comparison-Group Members, by Year and Type services. This analysis exami Emergency department nPercentage and Dollar Impact of CDHP on Total Spending, ed spending in total and for each -17%* type of service separa -8% tely. 8% 5% Laboratory 38 27 37 40 45 Average ages were about the same, a the priv lthoat ue and gh statisticall public se y ctors differ . The ent: EBRI Data 29.9 for the book on Emplo HSA plan group yee Benefits and 30.9 is a for stat the istic al lower This study spen din hasg several on preslimitat cription dr ions. ugs First, drov it draws e moston of the the e sx aperi vings. Th ence of is study e a singlex, amin Midw ee ds over tern, 7 manu 6,000 fa indivi cturing duals emfro plom yer 709 Assurance (NCQA) as Healthcare Effectiveness Data and Information Set (HEDIS) quality measures. These included: 1) a of Health Care Service........................................................................................................ .............. 6 lower, $300amount Pharm ing to ac Outpatient hospital, ambulatory, and surgical y a reduction of $232 as compared wi4 th the year 49 -13%* before th 368 2% e HSA 434 was ado -7% 507 pted. -12% 556 Iacus, S.M., G. King, and G. Porro. “Causal Inferen Relative to 2006, by Year and Baseline Cost ce without Balance Checking: Coarsened Exact Matching.” reference work on employee benefit programs and work force-related issues. comparis ? oReducti n group, tho ons in ugh pharmacy the perce spending were l ntages of individual arge and mo s within ea stch ly sus age tained group over the f were not so tatis ur years tically di after th fferent. e HSA Just employers with a Clar om g p a e, ar n ic s d us aptive on G ed roup data o workfor ne year ce with pr re ior t gioo n and al, socio-demo two yearsgraph after the ic, a HS nd A ado company ption, a cultur nd us al charact ed a diff eristi erence cs potentia -in-differ lly ence mammogram (for females, ages 40 and olde Outpatient physician's office r); 2) cervical cancer screenin -14%*** g (for females, ages 21 a 1% 4% nd older); and 3) a 5% Another specification check involved estimating separate regressions based on total spending on health care in 2006. Political An 2006 Spending Quintile (Range, Mean) alysis 20 (1), 2011:1-24. 2007 2008 2009 2010 Figure 3, Average Annual Increases in Health Spending, 2007–2010 .............................................................. .. 7 over one Background and Pri -was half (54 adopted. In the first y percent) oo f bot r Literature h groups ear of th were e HS male. A, ph Polic armacy-spending red yholders made up about uction 40 percent s were 40–47 percen of each grout for p. Average Total 2,128 2,745 3,000 3,179 3,494 Substance abuse and mental health -22%** -12% 2% 0% (Di influ Dencin ) model t g beh havior at took i unn der to a th ccount e HSA prior plan h ch eange. alth care Conse use as quent a ly, way to these c o fin ntrol for dings are adv ne orse select t necessarily ion for in genedivi raliza duals ble twho o colorectal exam (for males and females, ages 50 and older). Inpatient hospital admissions, and the three cancer screening The sample was divided into 2006 spen Contact EBRI ding quinti Publications, (2 les. Spending02) 659-0670; in 2006 was used as fax publication a proxy orders to for health status. (202) 775-6312. This 1 (0-$161, Mean=$52) -12% 10% -11% 9% Inpatient hospital 172 474 505 438 489 household size was individuals in about all but the hig 3.3 people. Te hest quintile nure with the of spending employers was about . the same between the two groups, 13 to Laboratory -36%*** -21%** -20%** -19%* The Figure literat 4, Perc ure on CDHPs entage Impac and their t of CDHP impact o on n Health health c Ca are re C spendi osts, ng is s Relative mall, but to 2006, grow by ing. M Year uand ch oTyp f the e e of arly r Health eseaC rch are Subscriptions to EBRI Issue Briefs are included as part of EBRI membership, or as part of a had a Discussion broade choi r popul ce o 2 ($162-$473, Mean=$303) aftions health . plans. Nair et al., (2009) also-14%* found that lower -25%** pharmacy spen -7% ding drove all of the savin -1% gs in the Lo Sa indicators were sso, A.T., L exactly matched on all of .A. Helmchen, and R. K their values. aestner. “Th The r e E emaining fo ffects of Co ur nsumer-Dir health-services utilizatio ected Health Pl n measures were co ans on Healtharsened Care permitted testing on whether the effects of the HSA plan design were symmetric across different health statuses of $200 Emergency department 35 41 51 51 55 Orders/ Pharmacy -32%*** -26%*** -26%*** -20%*** 14 years, but statistically different. As noted previously, nearly the entire sample was located in the Midwest. 3 ($474-$1,064, Mean=$727)$199 annual sub 169s*cription -29%*** to EBRI Note6% s and EBRI Issue -12% Briefs. Change of -15%** Address: EBRI, focused on Servic CDHPs that used e ....................................................................................................................... HRAs, and only a few studies have been published recently with data on HSAs. ..................... While 7 first year of a full-replacement CDHP in 2005, and noted that lower pharmacy spending was not large enough to reduce and employed i Spending.” Outpat ni the matching eJournal of Ri nt hospital, am process using sk buland Insuranc atory, and s Sturge’s r urgi ec 77 (1), al ule, th Jan 588 e defa uary 2010:85 ult binning algorithm of 698 –103. 823 the user- 935written Stata comm 1,059 and Source: EBRI estimates based on administrative claims data. This study enrollees. In reports experienc these specificate ions, three over five years from a equations we sin re g esti le larg mated: e empl total s oyer in pen the din Mi g, outpatie dwestern nt Unit phaed rmacy s States that pendin g, and ? When spending by pre-HSA health spending quintile was examined, the largest spending effects in the first 4 ($1,065-$2,569, Mean=$1,691) 1100 13th St. NW, Suite 878, Washington, -31%*** -15%** DC, 20005-4051, (202) 659-0670; fax number, -2% -9% Second, although Notes: N=10,509. CDHP=Consumer-Directed Health Plan. Presented are coefficient estimates of the (difference-in-differences) eff estimation of the impact of HSA plans four years post-adoption is a key stren ect of CDHP in gth of this analysis, it Outpatient physician's office 714 798 856 893 966 HRAs and HSAs both fall under the CDHP umbrella from a design standpoint, they are at opposite extremes of what is Subscriptions 132* overall spending. “–cem-“ that was implemented. This automation resulted in 18 cut points for each of these four count “measures. See adopted a Figure 5, hi Ann 5 ($2,570+, Mean=$6,157) gh-dedu ual Reduc ctiblte h ione in alth Health plan wit Cah r an e Costs HSA -15%*** Per for all em Person ployees Due to -7% in 2 Adoption 007. Theof an HSA 2% alysis exami Plan, Rel na etive s healt -1% to h2006 care, by all other medical spendi each of four follow-up years from generalized linear models with gamma family and log link. Models included indicators for male ng. Separate equations were not estimated for the various types of medical , policyholder, five spending because (202) 775-6312; e-mail: subscriptions@ebri.org Membership Information: Inquiries Figure 2 shows mean spend year of the HSA wer ing e seen in the thir amounts over time for d and fo the HSA-pl urth quintiles. Th an enrollees a e highen st pre d the compar -HSA quintile ison group. group Th experie e table nced Substance abuse and mental health 53 62 68 76 86 also requires plan age groups, three geographic regions, four years, CDHP, and four CDHP times year interaction terms (presented), as well as Char participants to have maintained continuous eligibility throughout the study peri lson Comorbidity od. This eligibility considere Lo Sasso, A d a. CDHP, with no T., M. Shah, and B.K. Frogn additional types of C er. “Healt DHPs in between. h Savings Acco unts and Health Care Spending.” Health Services http://gking.harvard.edu/cem/ Year ............................................................................................................................... for a complet regardinge EBRI description membership an of this meth d/or con odoltributio ogy. ns to EBRI-ERF should be directed ................. to EBRI 8 1 (0-$161, Mean=$52) -$69 $57 -$64 $53 spendi individuals ng inin t the plan ov he lower spen er a perio ding quinti d of foles ur years after did not often thi is n change. Results show that sp cur certain types of health care servic ending was es, red such as in uced spati ignient ficantly Laborat Index, years of tenure, and household size. Statistical significance based upon robust standard errors (clustered by policyhold ory 21 26 25 er) denoted as 29 31 shows total h $100 spendi ealth spendi ng reductions ng, as w in the ell fi as rs its t year of various compo the HSA that nents: weinpatie re not s nt, outpat ustained. T ient, he s emer econd gency HSA- -ro plan om, offic year s e-visi howe t, d 75 More rec criterion e may ntly,also a seri limit es of pa the gene pers exam ralizabil in ity ed 28 em of the rployers esults. that o ffered combinations of HRAs, HSAs, and health plans Research 45 (4), August 20 follows: *** p<0.01; ** p<0.05; * p<0.10. More information about the econometric model can be found in the appendix and in Fro 10:1041–60. President Dallas Salisbury at the above address, (202) 659-0670; e-mail: nstin, Sepulveda salisbury@ebri.org 2 ($162-$473, Mean=$303) -$136* -$239** -$66 -$13 Pharmacy 507 573 611 681 733 in th admissions an e inaugura d out l year patient s of the u HSA pla rgery. n in medical, pharmacy, and total-claims categories. Further, the magnitude of the and Roebuck (2013). Figure and pharmacy 6, total spen Perc spen entage din ding and g was r , etc D.o Overall e llar ducImpact ed avera only in of ge a the CDHP nnua secon o l spe nd a Me nnd dical din fourt g was $ Spendi h quintiles. By 1,ng, 756 for Rel the ative the fourth HS to A plan 2006, year, the HSA group by Year and and $2, plan reduced 128 Baseli fone r the An HRA is an 3 ($474-$1,064, Mean=$727) employer-funded health plan that reimburses -$453*** employees $89 for qualified m -$180 edical expens -$232** es. An HRA is So urce: EBRI estimates based on administrative claims data. with high deductibles, but no associated accounts (see Buntin, Haviland, McDevitt, and Sood 2011; Haviland, et al. spending decline was greatest in this first year and eroded over the succeeding three years. It is observed that Cost .......................................................................................................................... .................... 10 Editorial Bo 4 ($1,065-$2,569, Mean=$1,691) ard: Dallas L. Salisbury, publisher; Stephen Blakely -$805*** , editor. Any views -$382** expressed in this publicat -$43 ion and those o -$224 f the authors should comparison g Notepharmacy s s: CDHP roup =C in 2006. Fig o np se ume ndin r-Dg ir i en th cu te re d He e 3 fo shows the alth urth P lanquinti s; comme sub le s w neq chile o ed uent in 2nly th 00chan 7. e t ges i hird n qui spen ntile din c g. Duri ontinue ng t d to he ha first year of t ve reduced sp he ful endi l-ng as Third, deductible size matters in utilization, and in this study most plan participants chose the HSA plan with the higher typically com Manning,W. b G. ‘‘The Logg ined with a hied Dependen gh-deductiblet Vari health able, planHete , thou roscedastici gh there is ty, and th no statutory e Retran minimum-de sformad tiuctibl on Problem e requir.e ’’ m Journ ent al 2011a; Haviland, et al. 2011b). Researchers also had information on the amounts that employers contributed to these Standard errors were clustered at the policyholder level. All analyses were conducted using Stata/MP 12.0. not be ascribed to the officers, trustees, members, or other sponsors of the Employee Benefit Research Institute, the EBRI Education and 5 ($2,570+, Mean=$6,157) -$831*** -$366 $92 -$62 spending declines affected medical claims transiently, whereas prescription-drug claims were affected over a sustained replacement comp HS ar A, spen ed with din thge ye increased 7 percent ar before the HSA for th plan was e HSA pla adopn ted. grou p and 29 percent for the comparison group. In deductible amounts, which were higher than national averages for HSA plans at the time (Claxton et al., 2007). of He $0 alth Economics 17, June 1998:283–95. associated Figure 7, Perc withentage an HRA. I and n f Dact, employers have ollar Impact of CDHP a trem on Ph enarmacy dous amo Spen untding, of flex Relative ibility in desi to 2006, gning by a h Year ealth and planBaseline that accounts. T Research Fund, hese studies or their staffs. had a n Nothin ug her mbe er of in is to be co findings: nstrued as an attem First, moving to pt to aid or higher hinder de the adoption ductibles an of ad CDHPs ny pending le was associated gislation, regulation, with Source: EBRI estimates based on administrative claims data. 2007 2008 2009 2010 period of years. These results are consistent with the findings of Lo Sasso, Shah, and Frogner (2010), who reported Notes: CDHP=Consumer-Directed Health Plan. Presented are coefficient estimates of the (difference-in-differences) effect of CDHP in each of the next two years, spending increased 22 percent and 16 percent for the HSA plan group, while it increased less than or interpretative Cost .......................................................................................................................... rule, or as legal, accounting, actuarial, or other such professional advice. www.ebri.org .................... 10 incorporates an HRA. The amount of money placed in the account, level of the deductible, and comprehensiveness of a 14 percent savings in the first year, and the highest savings were associated with high-deductible health plans that four follow-up years from generalized linear models with gamma family and log link. Models included indicators for male, policyholder, five age reduced health care spending after the introduction of an HSA- based, high-deductible plan and attributed spending Fourth, although the HSA-plan cohort was matched to comparison members from a second employer, the possibility Manning,W. G., and J. Mullahy. ‘‘Estimating Log Models: To Transform or Not to Transform?’’ Journal of Health 10 perc Sourc ent in each e: EBRI estimayear tes based for the on adminis com tratp ivariso e claimn s dat gro a. up. coverage are groups, three geographic regions, four years, CDHP, and four CDHP times year interaction terms (presented), as well as Charlson all subject to variation. Employers often cover certain preventive services in full, not subjecting the m to had low or no employer contribution to the account. Second, higher deductibles resulted in fewer episodes of care as Figure 8, Percentage and Dollar Impact of CDHP on Total Spending, Relative to 2006, by Year and Baseline EBRI Issue Brief is registered in the U.S. Patent and Trademark Office. ISSN: 0887 ?137X/90 0887 ?137X/90 $ .50+.50 *** p<0.01; ** p<0.05; * p<0.10. remains reductions that to r Comorbidity Index, years of tenure, and household size. Statistical significance based upon robust standard errors (clustered b derived educedef pres fects crimight ption-d stil rul g cl be ai biased ms. if unobserved variables associated with the HSA y policyholder) plan were also Economics 20 (4), July 2001: 461–94. the deductible, and can carve out specific services from the deductible, such as prescription drugs, if they choose to do well as less spendi Cost .......................................................................................................................... ng per episode. Third, CDHPs affected lower-income populations and the chronically ill .................... to the same 10 denoted as follows: *** p<0.01; ** p<0.05; * p<0.10. More information about the econometric model can be found in the appendix and in correlated with health-services utilization. There is also the possibility that HSA enrollees may have altered the timing of Fronstin, Sepulveda and Roebuck (2013). so. An HRA is typically set up as a notional arrangement: Leftover funds at the end of each year can be carried over to extent as non-vulnerable populations, which for obvious reasons may have greater health consequences for the lower- © 2013, Employee Benefit Research Institute ?Education and Research Fund. All rights reserved. ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org Issue Bri Issue Bri Issue Bri Issue Bri Issue Bri Issue Bri Issue Bri Issue Bri Issue Bri Issue Bri Issue Brie e e e e e e e e e efffffffffff • July • July • July • July • July • July • July • July • July • July • July 20 20 20 20 20 20 20 20 20 20 2013 • 13 • 13 • 13 • 13 • 13 • 13 • 13 • 13 • 13 • 13 • No. 388 No. 388 No. 388 No. 388 No. 388 No. 388 No. 388 No. 388 No. 388 No. 388 No. 388 14 11 12 15 13 2 8 5 3 4 9 A monthl ebri.org ebri.org Issue Brief • July 2013 • No. 388 ebri.org Issue Brief • July 2013 • No. 388 y resea Issue rch report f Brief • r July om th 2013 e EBRI Education and Re • No. 388 search Fund © 2013 Employee Benefit Research Institute 10 6 7