“Brand-Name and Generic Prescription Drug Use After Adoption of a Full-Replacement, Consumer-Directed Health Plan With a Health Savings Account,” and “How Would Defined Contribution Participants React to Lifetime Income Illustrations? Evidence from t ..

Prescription Drug Use: Do consumer-directed health plans (CDHPs) result in lower prescription drug use? New research from EBRI of one big employer finds that moving to an HSA-eligible plan reduced the number of both generic and brand-name prescriptions filled. Lifetime Income Illustrations: Do consumer-directed health plans (CDHPs) result in lower prescription drug use? New research from EBRI of one big employer finds that moving to an HSA-eligible plan reduced the number of both generic and brand-name prescriptions filled.

Brand-Name and Generic Prescription Drug Use After Adoption of a Full-Replacement, Consumer-Directed Health Plan With a Health Savings Account

A full-replacement HSA plan was associated with a 4.7 percentage-point rise in the generic-drug dispensing rate (GDR) after one year, and settled 3.4 percentage points higher after four years. The GDR for maintenance medications experienced a similar effect, while for nonmaintenance conditions the GDR rose by 4.1 percentage points after one year, but was just 1.7 percentage points higher after four years.
At the end of the four-year follow-up period, GDR was greater by 4.5 percentage points for hypertension, 15.4 per-centage points for dyslipidemia, and 7.8 percentage points for asthma/COPD. No significant effects were detected for diabetes GDR, but the measure for depression was lower by 8.4 percentage points after 2010.
GDR increases were due to individuals discontinuing use of brand-name drugs without substituting generic therapy. After one year under the full-replacement HSA plan, 0.43 fewer generic and 0.95 fewer brand-name prescriptions were filled, on average.

How Would Defined Contribution Participants React to Lifetime Income Illustrations? Evidence from the 2014 Retirement Confidence Survey

In May 2013, the U.S. Department of Labor's Employee Benefits Security Administration (EBSA) published an advance notice of proposed rulemaking (ANPRM) focusing on lifetime income illustrations. The 2014 Retirement Confidence Survey (RCS) included a series of questions concerning monthly income illustrations similar in many respects to those provided by the EBSA’s online Lifetime Income Calculator.
The vast majority of respondents said the retirement income projection was useful; more than 1 in 3 (36 percent) of the respondents thought that it was very useful to hear an estimate of the monthly retirement income they might expect from their plan, and another 49 percent thought it was somewhat useful.
A total of 17 percent of the respondents indicated that this information would lead them to increase the amount they were contributing. However, of those responding that their illustrated value was much less or somewhat less than expected, 35 percent indicated they would increase their contributions.

Figure 3 a Impact of CDHP on Generic Dispensing Rate Figure 1 6 level not The Data how Conc Endnotes Although Greene, Jessic have we mu thlusio at ight and ch acce re was ted du heMethods n ss ctions y a high medi et al. “T woul to e current an r in d by i(or presc n he Impact of Cons crea 3. 4 m account r perce ise dpoi iption their nnt) -dr tage balanc for contri ug po th u ut m es e butions. ints ilization er-Dir money and after ecte ther can The curre fo d H efo ur re aealth Pl re ntly v years. sult erage coul saved ind ans o pharm con These not in dn iti a Prescripti tccurately a he o results cnal y em ex in pen ployer crease were on Dr provi ditur -s niug Use.” d pons e n ee arly savings 2total 014 ored id enti contri projec Healt pl for cal an butions employer h ted Affairs for as re r the eported tirem . 27(4) for GDR pla all e n nt in of the Brand-Name How Would Defined and Generic Contribution Prescription Participants Drug Use React After to Deviation of the Illustrated Monthly Income from the Amount Expected, 9 CDHP Effect After CDHP Effect After CDHP Effect After CDHP Effect After 1 Figure 1 July/August 2008:1111 -9. maintena sp income. resp RCS onsors, ondents was n $increases ce 50, was medicatio 000,$927. an in d d n th o s. Wh wnstream e wei The en g the hte GDR stated dmedical me for d niin an onmaintena tention costs for th may e to tot n incr ec c ae l lipse ann e con ase u dtal itions h contri ose cont benef but aribution lso ions rose its. was w A by a lar s asso 4.1 $ g6 e ,90 perc bo ciated 0 dy . enta The ofwit lit ge me h er points th ature dian, e quartile has we after ight one of e d, http:// www.dol.gov/ebsa/newsroom/fsanprm.html For As def this ine study, d contributio pharmacy n plan and s have medical continu claims ed and to evolve insura in nce to e the ligibil prim ity ar data y form were of obt emained ployer-sponsor from a large ed re employ tiremeer nt . plan The as a Function of the Quartile of Illustrated Income Lifetime Adoption Income of a Full-Replacement, Illustrations? One Year (2007) Two Years (2008) Evi Consumer-Directed dence Three Years (2009) from the Four Years (2010) 2014 Health 7 Baseline (2006) Variable Means year, document projected monthly but inco was ed mon m th just te e hly ill nustrated, e 1.7 retir t ecperc e ono men entage m 23 tic income percent value poi of n (i ts n of medicatio h cthose u igher rrent in aft dollars nbot e ar dfo herenc h ur ) ofwas years. the e $2,0 botto for The ch 12. m ronic a two d option disease quarti of lethe (e.g., s mentione HSA Roplan ebu d ctalso k hey et al., prompt would 2011 incr ed ). ease data in the cover private ed tsector, he five-y pu ear blicperiod policy of anJa alysis n. 1, has 2006 been thro inc ug rh eas De ingly c. 3 1 conce , 2010. rned This that Midw employee est manuf decision acturer making completely regar ding 2 100% March 2014 • Vol. 35, No. 3 Generic Dispensing Rate (GDR) According to EBSA, w hen projecting account balances, it is reasonable for a plan administrator to assume: Haviland, Amelia M. et al. “The effects of consumer-directed health plans on episodes of health care.” Forum for Health In Retirement Plan an attem With pt to provide a Confidence Health some addition Savings al evidence Survey with Account respect to potential defined-contribution-participant reaction, increases Moreover, their contribut in th m e io Congr ons. st condition This essional dec -r s eased p Bu ecif dget ic to GD Of 1Rs. 7 fice percent At (CB the O ) for en rece d th of ntl o sty e h Study Group e annou in fo th ur-y e n third ced ear t fquartil o hControl Group llow-up at in e scoring an pe d rio 7 perce d relevant , GDR nt was for polic th gr iose eeater s, it in no tby hw e highest ass u mes replac savings ed an its All prescriptions d pr decumulatio eferred provi n decisions der organizat may iobe n (PPO subject 0.047 ) with to *** errors an HSA associated plan, 0.051 which ***with comm an ina enc bility 0.046 ed oto *** n Jan. accura 1, tely 2000.034 7, determin for *** all e a chow tive Economics & Policy. 14 (2), September 2011:1–27. 90% Was the Employee the Monthly Bene fit Amount Research More, Institut Les e s, (EBRI) or about included What a (N=13,093) series Had Be of en questions Expected? (N=13,093) in the 201 % Bias 4 Retirem p-value ent Confidence Variable Contributions continue to normal retirement age at the current annual dollar amount, increased at a rate of 3 percent per year. 4.5 that quartil perc a e. 1 entag perc Maintenance prescriptions ee n t poi incr nts ease for ihy n Medicar pertensio e n, Part 15 D .4 pr pe escriptio 0.048 rcenta *** ge n -dru points g utiliz for 0.045 atio dysli *** n pwi idemia, ll result anid n 0.038 7.8 a 0.perc 2 ***perc entage ent reduction poi 0.035 nts for *** in employees much mont an hlyd retirem their de ent pen indcome ents. can There be for gene e, data rated we fro rem a v an ailable account from balanc one year e at b reti efor rement e and age. four Whil yeare s after today very few By By Paul Jack Fronstin, VanDerhei, Ph. Ph.D., D., Employee Employee Be Be nene fit fit Research Research Insti Intstitute ute, and M. Christopher Roebuck, Ph.D., RxEconomics 4 Age: Only Survey 8 percen (RCS) t concer of the ning respo m no dents nthly indicat income ed illustrations that the m similar onthly amou in many nt was respects much to letss hose than provi expect ded ed, by tthe hough EBSA ne ’s arly online Nonmaintenance prescriptions 0.041 *** 0.041 *** 0.039 *** 0.017 *** asthma/COPD. medical costs ( C No ongressional significant Bud effects get were Officedetected , 2012). Giv for edi n abetes the pot GD ential R, b for ut the the measur medical-cost e for dof epressi fsets, on CDHPs was lowe and r otby her 80% implementatio retirees ? co Investment retur nvernt of their the ent ns are 7 percen HSA ire plan. define d t contr per year ibution (nomina anld/or ). IRA balances to an annuity at retirement age, the monthly Iacus, S.M., G. King, and G. Porro. “Causal Inference without Balance Checking: Coarsened Exact Matching.” Political 5 Average (years) 30.74 30.48 1.40 0.25 Hypertension prescriptions 0.026 0.008 0.022 0.045 ** Does 1 Lifetime in 5 (19 Hea Inco per ring me cent) the Calc re Mon ulator. plied thl th y Th at Ae it mmajor was ount som Figure differ ewhat ences Ch less ange are: tha n Expecte expected. d Retiremen More than ha t Ag lf (58 e? percent) thought that it was 8.4 plan Paul perc Fronstin desientag gns tie h s at poi director raise nts aft pati of e r tent h 201 e Hea cost-sharin 0. lth Education g for pr an escripti d Research on drugs Prowoul gram d at seem the Empl to beoyee counter-pro Benefit du Rective, search part Institute icularly in income that could be purchased provides a very convenient method of illustrating the approximate amounts that the <18 0.32 0.34 -2.30 0.07 Analysis 20 (1), 2011:1–24. ? A discount rate o Dyslipidemia prescriptions f 3 percent per year, in order to 0.077 show *** the projected account balance in toda 0.127 *** 0.147 y's dollars. *** 0.154 *** Brand-Name and Generic Prescription Drug Use After Introduction 70% Less about than what 1 th percent ey had oexpect f the respondents ed, while anrep other lied 7 that percent they repli would ed reti that re it swas ooner somewh as a res at ult more of the than neexpect w informat ed, and ion, but Members were allowed to choose between two deductible categories: 1) $1,250 per individual, $2,500 per family or light (EBRI of ). tM. he Chr resiults stopher 18–24 prese Roe nteb d uck in tis his preside analysis. nt an Instead, d CEO of va lRxEconomics, ue-based 0.03 insu rance LLC. This 0.02 desi arti gns cle (VBI was 5.40 D) wr that itte re 0.00 n duce with assistance or eliminate account ? Rather balan Diabetes prescriptions thces an simply could “saf usin ely” g no ge rmal nerate retir in e ment terms 0.006 age of mo for nthly the calculation, reti0.040 rement incom the RCS e. T respo h0.030 e pro ndent posed was me aske thodolo 0.051 d his g y or inh the er 8 10 In May 2013, the U.S. bDepartment of Labor's Employee Benefit s Security Administration (EBSA) published an advance As Adoption of a Full-Replacement 5 9 percent percent a relativ in said edi measure, cated it was thmuch at GDR they more can woul increase than d re tire expected. either later as throu a resu gh hig lt. her The , Consumer-Directed Health gevast neric majority use or low (89 er percent total ut ) ilizatio did non. t believe Plan sponso this rs When converting current and 25–34 projected account balances into lifetime income stream 0.16 s, it is reasonable for a plan ad 0.16 1.50 ministrator 0.21 to assume: 2) $2,150 per individual, $4,300 per family. Most enrollees (89 percent) selected the higher deductibles—probably pres from Nair, Kavita V. et al. “Consum cript EBRI ion- Asthma/COPD ’s 60research d %rug copays an prescriptions d in edi o er-Drive trorial der to sta n He bols ffs.alt te Trh his Plans: Impac adh work 0.073 erenc was *** e mconducte t on Utilizati ay be a mor 0.112 d ton her o *** effe and E ughctive the xpe EBRI n ad nd itur 0.104 effic es for Chron Cente ient ***r for strate Ric Diseas eg search y (0.078 Cher e Suffer on new, *** Hea le th rs.” advance notice of proposed rulemaking (ANPRM) by the U.S. Department of Labor's Employee Benefits Security expected retirement age. 35–44 0.17 0.19 -4.80 0.00 would notice information like of proposed ly w co onsider uld impact rul membe emaki thng eir rs (A choosing exNPRM pected ) fge ro ecusing tirement neric eq on uivale alg ife e.ti nts me A total /alternat incom of e 19 ive illpercent us strations over bra of in tho nds periodic, s e to who be a thought defin desira ed-contr ble that and their ibutio effici ne -nt Depression prescriptions -0.071 *** -0.087 *** -0.095 *** -0.083 *** because the lower-deductible plan had annual premiums of almost $300 per worker (for individual coverage), Journal of Occupational & Environmental Medicine 51 (5), May 2009:594–603. Rosen Plan With a Health Savings Account Benefits anIn d novation Fendrick, (EBRI 2007). CRHBI). Alternatively, CDHP plan sponsors may push , p. 2 to have maintenance medications defined as ? A rate of interest equal to the 10-year constant maturity Treasury securities rate. Administration (EBSA) would take this approach and expand it to include the prospective accumulation period for EBRI Employee Benefit Research Institute Notes (ISSN 1085 ?4452) is published monthly by the Employee Benefit Research Figure 1 sho 50w %s how 45–54 the illustrated monthly income deviated from 0.25 the amount 0.22 expected as 5.20 a functi0.00 on of the quartile Source: EBRI estimates based on administrative claims data. ? Since the age of the spouse was not known for married respondents, only the single life annuity income pension-plan benefit statements. The ANPRM sets forth certain language and concepts EBSA is considering as part of outcome illustrated amo fternthly adopitin ncome g an was HSA much plan. Ho less wever, than ex if pect indivi eduals d indicat dise conti d thnue at they use woul of bra d n retire d-name later dru as gs a without result, whi le th whereas the (otherwise comparable) higher deductible plan’s premium was $0. The employer deposited the same preventive, and (like cancer screenings) exempted from the deductible. a someone Institute, not 1100 yet 13 at rSt. etirem NW, ent Suit ag e e.878, Washington, DC 20005-4051, at $300 per year or is included as part of a membership 55–64 0.06 0.07 -3.00 0.02 CDHP=consumer-directed health plan. ? Mortality as reflected in the applicable mortality table under Sec. 417(e)(3) of the Internal Revenue Code. of illustrated income. While 18 percent of those in the lowest-income quartile, when ranked by the value of illustrated Parente, Step illustration hen T wa.s , Roger F used. eldman, and Song Chen. “Effects of a Consumer-Driven Health Plan on Pharmaceutical substituting The 13 future perc follo pro ent win posed og generic f those organiz regulations who thera ations tho py.,provide u The that ght la their may ngua the illustrated not fge undin repr and g esent co for amount ncept EB a RI ne s s woul CRHBI: t were econom d be s o Ameri mew ic pabenef rt c hat of an tit—partic les h Ex e press, sr e tha guln atory uA lexp arly meriprise, ected framewo if the in re dic Aon duced rk a ted under Hewitt, t utili hey Sec. zatio Bl wou ue 105 nld of subscription. b Periodicals postage rate paid in Washingt on, DC, and additional mailing offices. POSTMASTER: Send address amount into 40t %he enrollee’s HSA regardless of which deductible level was chosen, although this contribution was higher How Would Defined Contribution Participants React to COPD=chronic obstructive pulmonary disease. Male 0.51 0.51 0.30 0.79 th 1 monthly income, thought the value was much less than expected, only 12 percent of those in the second quartile had changes to: EBRI Notes, 1100 13 St. NW, Suite 878, Washington, DC 20005-4051. Copyright 2014 by Employee Benefit ? If married, the Spending an pa d Utilizati rticipant's spouse on.” Health Servic is the same age as the participant. es Research 43(5), 2008:1542–56. the Employee Retirement Income Security Act of 1974 (ERISA), and under which: leads Cross take this All prescription drug measures are adjusted using 30-days' supply equivalents. to Blue ill act ness Shi ion e. exac ld Only Association, erbat 9 perc ions ent and Boei of subse ng, those Des qwho ue eret nt thought emer Mutuge al, ncy-depa th Fe eir dera illustrated l rtment Reserv e amounts use Employe and iw nepati e re Be en nef equal t its hospitali System to thez iratio , ex Ge pectatio n. neral Thus, Mills ns the , HSA Policyholder 0.43 0.42 0.80 0.49 Of for course, workers any with such fami projecti ly coverage. on is necessarily required to make a number of critical assumptions—including future Lifetime Income Illustrations? Evidence from the 2014 Research Institute. All rights reserved, Vol. 35, no. 3. References ? Additio 30 n % ally, given that the information was being provided to the RCS respondent during a phone interview, a similar Presented are coefficient estimates of the CDHP effect in each of four follow-up years (compared to baseline) from ordinary, le assessment. The weighted percentage of respondents that thought the illustrated values were ast squares much lower 12 Spouse 0.23 0.23 0.20 0.90 plans’ Healthways, indicatimpact ed they I B on woul M, the JP d num Mor retire g be a n lr a ter, Chase, of bran but d Merc no an ne d er, ge of an neric those d P fdr izer. with ug f An illustrated ills y must views be expressed am ex ounts amine more d. in R this etsults h an report w from hat are th ththo ey ese expected sanalyses e of the a in are udic th ors ated contributio ? Pan ym aents comm ctivity, futur ence immediately once the p e rates of return, futur art eicipant reaches no asset allocation, rma al retirement a nd future age, nnuit if the participant is younge y purchase prices. r than normal Re tirement Confidence Survey, p. 9 ? A models, which included indicators for male, policyholder, three geographic regions, five age groups, four years, CDHP, and four participant's pension benefit statement (including his or her 401(k) statement) would show his CDHP times year or her current Parente, Stephen T., Roger Feldman, and Yi Xu. “Impact of Full Replacement with Consumer Driven Health Plans on Health only the projected monthly income (based on the projected account balance) was provided to the respondent. than Austin, P.C. “An Introductio expected Child were even n sm to Prope aller for nsit thos y Se core Metho with larger d s mo for Red nthly 0.34 uil cin lustrations: g the Effects of Confoun 0.35 3 percent for -1.00 di those ng in O in0.41 b the servati thirona d qu l artile Study and Control Groups reported and they sho would uld in 20 no ret Fi % gure t ire belat 4. ascribed e r. to the officers, trustees, or other sponsors of EBRI, Employee Benefit Research Institute- interaction terms (presented difference-in-differences estimates), as well as Charlson Comorbidity Index, years of tenure, and retirement age. household size. Consequently, a full cost-benefit analysis of imposing a regulatory mandate similar to that contained in the EBSA’s account balance and an estimated lifetime income stream of payments based on that balance. This lifetime Years of tenure 13.49 13.20 2.80 0.02 Care Cost and Use of Preventive Services.” Insurance Markets and Companies: Analyses and Actuarial Studies.” Multivariate Behavi The Em oral Rplo esearc yee Benefi h Mayt Research Institut 2011. 46(3): 399– e 42 (EB 4. RI) was founded in 1978. Its mission is to and 1 percent for those in the highest-income quartile. Combining the “much less” and “somewhat less” categories Statistical significance based upon robust standard errors denoted as follows: *** p<0.01; ** p<0.05; * p<0.10. This analysis began with a study group of 13,203 active workers (younger than 65 years of age) and their dependents Education and Research Fund (EBRI-ERF), or their staffs. Neither EBRI nor EBRI-ERF lobbies or takes positions on ANPRM would have to consider the impact of doing so in terms of standardization of disclosures, both in terms of 3 Household size 3.13 3.13 0.10 0.94 Respondents to the 2014 RCS who were currently contributing to an employer plan were asked: income illustration would assume contribute to the partic , to encourag ipant had e, an reacd to enhanc hed normal e the developmen retirement age t of sound empl as of the date ooyfee ben the bene efit fit The conditional Co10 mp % increase w utations 1(1), as ap 201 proximatel 0:4–14y $1,1 . 50 per year. After resulted Whenon the e inyear stat 41 ed perce unintent der nt th of ie on tho CDHP, tos de e in fer 0.43 t hretir e fe lowe e wer ment st generic quartil was associated e an for d 0.95 illustrated fewer with the mo bran nthly quartil d-nam ine come e of pr mo escr innthly dicati iptions ing ncome were that ithe llustrated, filled, valu oe n was average. 22 per- less A T A G L A N C E continuously covered by the employer throughout the five-year period of analysis. Next, a control group was created specific policy proposals. EBRI invites comment on this research. Who we are Charlson Comorbidity Index 0.14 0.14 -0.70 0.56 mitigating plan sponsors’ ERISA lia programs and so bility, as well und public policy as the potential through objective confusion that cou resear ld ch and arise educati among on. EBR workers I is the only who have Bundorf, M. Ka statement, te. regardl Consum ess er-of Directe his or d Hea her act lth Plans: Do T ual age or plann hey Deliv ed ret er? irement Research S date. ynthesis Report No. 24, Robert Wood 1 4 Both than cent of e mai x pecte those ntena d i. n nce This thea n bottom perc d non ent - maintena age quartil dec e line me nce d ntit medicatio o oned 36 percent they ns wo e x fuld o peri r the increase ence second d de th creases, eir quartil retirem e, altho 22 e npe u t gh age, rcent non but for -main only the tenance t13 hird percent quart utiliza ile of ta ion those nd “In recent Source: EBRI estimates based on administrative claims data. years, a number of organizations have developed calculators to estimate how much a retirement plan using The Rdata CS is cosponsored b fro 0%m another, y EB larRI, a private, non ger, national eprofit, nonpa mployer thrt at isan, public-policy maintained PP rO esearch or coverage ganiza during tion, and Green the full fiv wald & Associ e years. ates, a private, nonprofit, nonpartisan, Washington, DC-based organization committed exclusively to previously received voluntary disclosures of comparable information from their employers. A number of service Quan, H. et al. “Coding Algorithms for Defining Comorbidities in ICD-9-CM and ICD-10 Administrative Data.” Medical Care Johnson Foundation, 2012. Lowest-Income Quartile Second Third Highest-Income Quartile Brand-Name and Generic Prescription Drug Use After Adoption of a Full- did only in th not e 9 secon percent remain d quartil for signi thfe e icantly were highes lsimilarly otwer quart public afile. ter inc poli fou lined. r cyears. y r This esearch perc Bran and e Figure 4 d ntage use edu cat also decreas ion on dece lined d econ to o in 2m pe the ic s rcen ecuri first t tfor year y atnd em hose in hyplo in pertension tyhee e b theird nefi qua and t issrti ues le . will provide as monthly income for life. Your answers to the next two questions will allow us to estimate that Washington, D.C ? For a parti .-based market cipant who has -research firm. Th not yet reache e 2014 RCS d normal data ret collection w irement aag s funded b e, his or yher grants f pension rom appro benef ximatel it state y twm o d ent ozen also providers43(1 and 1), Novem defined-contr ber 20 ibution 05:1130 -pl–9. an sponsors have given significant thought to the best method to illustrate the a EBRI’s membership includes a cross-section of pension funds; businesses; trade associations; Don’t Know Much More Than Expected Replacement, Consumer-Directed Health Dependent Variables Plan With a Health Savings Account, by dyslipi and 1 d perc emia, ent wher for thos eas e for in depression, the hi Impact of CDHP ghest ge quar netile. ric fills were on Prescription Drug Utilization lower after one year. Dyslipidemia generic fills were higher by organizations. Th monthly would show amou e full report, R a nt proba for b you le CS acco Fact Sheet to runt eact balanc to. s, an d other e and resource an estima s are online at ted lifetime www.ebri.o income rg/surv stream eys/RCS/20 based on 14 the participant's Introduction Charlson, M.E. et al. “A New Method of Classifying Prognostic Comorbidity in Longitudinal Studies: Development and impact of changing these assumptions on projected monthly income through a series of sensitivity analyses. labor unions; health care providers and insurers; government organizations; and service firms. Somewhat More Than Expected About What Expected Figure 2 shows how the illustrated monthly income deviated from the amount expected, as a function of household Paul The impact Fronstin, of the Ph.D., HSA pla Empl n was oyee estimate Benefit d on Research both the Ins abtsolute itute, a and nd re M. lative Christopher use of generic Roebuck, and bra Ph.D., nd-name 0.14 at the end of the four-year period, and brand fills were lower by 0.20. Finally, generic prescriptions for Roebuck, M.C. et al. “Medicati current account balanceon Adh projected erence to wh Lea at ds to Lo would wbe er H avai ealth C lable aat re Use an the participant d Costs Despite Incre ’s normal retir ae sed Dru ment age g Spe based nding.” 5 The body Vali of datio research n.” Journa on ho l of Chron w consum ic Diseas er-directed e 40 ( h5e), 1987:3 alth plans 73– (CDH 83. Ps) affect the use and costs of health care CDHP Effect After CDHP Effect After CDHP Effect After CDHP Effect After Figure 2 http:// ? First, www.askebsa.dol.gov/lia/home how much money do you currently have saved in your employer-sponsored plan in total? Somewhat Less Than Expected Much Less Than Expected income. Does Hearing While 13 the percent Monthl ofy those Amount with Figure household Affect incom Conf es of idence less than in the $60,0 Abilit 00 th you to ght Sav the e Eno value ugh was to much Live less RxEconomi prescription cs dr ugs, and analyses were conducted on several levels. Overall use of prescription drugs, maintenance vs. depression stayed significantly lower (-0.14 fills) after four years. on assumed Health Affairs future 30(1): Jan contribu uar tions y 201 and 1: 91– One Year (2007) inve 99. stment returTwo Years (2008) ns. The projected Three Years (2009) account balance woul Four Years (2010) d then be The results in the 2014 Retirement Confidence Survey (RCS) suggest that less than 1 in 10 (8 percent) of the defined services—while still relativelyBaseline (2006) Health Care Services Utilization Means small—is EBRI’s work advances knowledge and unders growing (Bundorf, 2012). In recent work, tanding of emplo EBRI has examin yee benefits and their ed these 6 than Comfor expta ecte bly d , in 10Retireme percent ont? f those with household incomes between $60,000 and $99,999 had a similar assessment. Source: Employee Benefit Research Institute and Greenwald & Associates, 2014 Retirement Confidence Survey. nonmainte All data were Generic Fills nan weighted b ce medication y age, ss, ex, and and condit education to ion-spec reflect the ac ific therapies tual proporti wereons i examined n the adult popul for hypertensio ation. n, dyslipidemia Charlt ? Next, on Mar ho y w E. et al., “Effects of Health Sa much money in total do you vings and Acco your unt– employer Eligible Plans currently on Utiliz contribut ation a e nd E to yo xur pend em iture ployer s.” -sponsored American Journ plan a l importance to the nation’s economy among policymakers, the news media, and the public. It contributio converted n participants to an estimated said the lifetime monthly income amount stream was much of paylm ess ents, than assuming expected, that while the about participant a quarter woul(2 d 7 ret per ire cent) at relationships using panel data from a large employer that implemented a full-replacement CDHP with a health savings ? A full-replacement HSA plan was associated with a 4.7 percentage-point rise in the generic-drug dispensing rate (GDR) Study Group Control Group The Only we 3 percen ight All prescriptions ed t perce of thn e tage respo ofn dents respondents indicated that ththou at -0.43 thght ey ** w th ee reill mu ustrated ch l0.04 ess value confident was muc in th heir 0.17 low ab eility r than to e save xpect eno 0.19 ed uwas gh to o nly live (abnormal cholesterol), diabetes, does this b asthma/chronic y conducting obstructiv and p e publishing policy re ulmonary disease sear (CO ch, analysis, PD), and depr and es special reports on sion. For each of annually? of Manag ” ed Care 17(1), January 2011:79–86. 7 th th What we do normal retirement age. This projected account balance and the related lifetime income payment would be had been overestimating the amount of retirement income that they would likely receive from their defined account Discussion The 25 after pe (HSA) o rcen ne ye tile w (Fronstin ar, aan s $831, a d s and ettled nd Roe the 75 3.4 b uck, percentage percen 2013; tile w points Fro as $4 nstin, high ,799. Ser e pulv a(N=13,093) fter e d fo a, ur ayears. nd Roebuck, (N=13,093) The GDR Jun for e 20 main % Bias 13;tenance medic Fronp-value stin, Sepulve ationd s a, and Endnotes Variable Maintenance prescriptions -0.29 * -0.02 0.13 0.13 employee benefits issues; holding educational briefings for EBRI members, congressional and 4 comfortably percent for in t hose retirem witeh nt ho as usa ehold result income of thiss information, of $100,000 but or more. 17 per Com cent binin replie g t d he that “m tuch hey le were ss” an then d “somewhat somewhat les les ss ” these categories, four dependent variables were examined: contributio expressed n plInpatient hospital days an in(at current least dollars as calculat on the ed usi pensio ng th n e bene EBSA’s fit state safe m 0.16 harb ent. or assump 0.15 tions and the 0.60 modified 0.63 methodology 1 Roebuck, Nonmaintenance prescriptions December 2013). Among other salient -0.14 results, ** it was fou0.07 nd that adoption o0.04 f the HSA plan reduce 0.06d both the experienced a similar effect, while federal agen for noncy m s aintena taff, and th nce c e news onditions media; the and sponsoring public opinion survey GDR rose by 4.1 percentage points s on emplo after one yee 8 The Specifically, premise this of CDH analysis Ps is estimated that members a prope wiln l sity be more score cost-conscious probit model on wh membership en makingin thte hie r study health-c group are-consumption (i.e., the employer Chen, Song, Regina A. Levin, and James A, Gartner. “Medication Adherence and Enrollment in a Consumer-Driven Health categori confident. A Another total of es, 2 perc 2Ho 2 42 3wever, ent responde respondents percent more ofd th th tprovided h ose at the an 3 with yi did n n 5 a hous res ot know defi p ehold nite ondents . re incom sponse (62 es perce for of less bot nt) hthan in of dic th $ ated ese 60,0 variabl t0 h0 at in th dicat e e s new in e d ad tinfo h dition at rmati the to val on the udi e age d was not at less im wpact hic thh an Emergency department visits 0.13 0.14 -0.50 0.67 Hypertension prescriptions -0.04 -0.01 0.06 0.07 mentioned above). The fact that mor benefit issues. e than halEBRI’s Ed f (58 percuc enation and Re t) thought that searcthe h F und illustrated (EBRI-ERF) performs monthly income the char was itin able line , number year, of 1) but pr Th ee wa scriptions ns umbe just 1.7 r of fille percentage 30 d -day and ov adjusted erall points pharmacy higher generic a ffills. ter cost fo s ur over years. a four-year follow-up period. A closer examination decis implementing ions than the whHSA en they plan) wu ere sing not as cover regress ed ors by pre-period a CDHP. For (2006) examp values le, when of the sfollowin pending g is variables: below the ge (h nder, igh)age, deduct age ible, times ? Both lifetime income streams (i.e., the one based on the current account balance and the one based on the Plan.” American Journal of Managed Care 16(2) February 2010:e43 -e50. they expected to retire. At that point, the monthly income available at their stated retirement age was estimated they their e co xpe nficted. dence, Outpatient physician's office and clinic visits This while perc 13 ent percent age dec felt line somewhat d to 30 percen more t fconfid or those ent, 3.59 with and household 6 perc3.64 ent felt incomes much -1.30 bet more weec no 0.29 nfident $60,000 as and a result 9 Dyslipidemia prescriptions educational, and scientif -0.02 ic functions of the Instit 0.05 ute. EBRI-ERF 0.09 is a tax-ex ** empt organizatio 0.14 *** n Figure 2 with The conditional their expectations median wamay s $75help 0. explain why this information not only did not engender a major shift in retirement revealed that the reduction in prescription utilization also involved a decreased use of maintenance medications for gender, indicators for spouse and child, years of tenure, and number of members in the household. The Charlson individuals are presumably less likely to visit an emergency department for relatively mild ailments, such as the projected All prescriptions account balance) would be presented as estimated monthly payments based on the expected mortality supported by contributions and grants. $99, of based the 99 in on 9 forma an Diabetes prescriptions thd e o methodolo tnly ion. 12 perce gy nt described for those in with the A househol NPRM 0.00 usin d inc g omes the safe of 0.02 $1 harb 00, or 000 assump or mor tions e. 0.03 and the modificat 0.01 ions ? At the 2) Th end e of nutmbe he four-y r of 30 ear -day, follow-up adjusted periobran d, GDR d fills. was grea ter by 4.5 percentage points for hypertension, 15.4 per- Deviation of the Illustrated Monthly Income from the Amount Expected, confidence or anticipat bed retirement age, but for most also did not translate into an intention to increase current 10 chronic Chernew diseas , M.E., A.B. Rosen, a e, and a worsenin nd A.M. F g of ad endr herie ck. 2007. “Val nce. This report ue-Base looks d Insura at the nce Des effects io gn.” f th Health Affairs e HSA plan on 2 6 the : 204–2 absolut 07. e and Comorbidity Index Generic was incorporated into the model as a measure 6.86 of general health 7.04 status (Charlson -1.60 et 0.19 al., 1987; Deyo, common cold. Asthma/COPD In the co prescriptions ntext of prescription-drug 0.02 utili *zation, it is thou 0.03ght ** that the higher 0.02 patie * nt cost-sharing 0.01 of A total of 4 pe of the participant. rcent indicated that In additio they n would , if the both part retire later a icipant ha nd incr s a sp ease their cont ouse, the lifetime ributions. income streams would be based on mentioned above, and provided to the respondents as follows: centage points for dyslipidemia, and 7.8 as per a Function centage points of Household for asthma/COPD. Income No significant effects were detected for contribution levels Brand (only 17 percent of the respondents indicate 3.38 d they would 3.58 increase the -2.70 ir contributions 0.03 as a result of 3) The total number of 30-day adjusted fills. relative use Depression prescriptions of brand-name and generic drugs. It -0.13 was *** expected th-0.09 at patients ** were m-0.10 ore likely ** to choose -0.14 a gen ***eric Cherkin, and Ciol, 1992; Quan et al., 2005). Also included were baseline measures of the number of inpatient hospital days, Eleven More CDHPs than perc enco 10 4en ur 0% int ages 10 of t(he 4 e3 n rolle rp ee spo rce es ndents nt) to o choose fEBRI Issue B thos indicati e les mng s uch expensive trhat liefs essth are period or ey s generi o eme xpect icals providing exp what cs to ov less re er tire brand coat nfident ornames e rt evaluati before as a if trh they e ons of emplo e sult median view of hea th ay rin ose g ee benefit issues and e g of medications th65 e in thought format as ion the the joint lives of the participant and spouse. 11 diabetes GDR, but the measure for depression was lower by 8.4 percentage points after 2010. Approximately t Maintenance prescriptions he same percentage of respondents found the illustration useful regardless of their quartile of illustrated income amount. Congressional Budget Office (CBO). “Offsetting Effects of Prescription Drug Use on Medicare’s Spending for Medical Brand Fills hearing the monthly income estimate). trends, as well as critical analyses of employee benefit policies and proposals. EBRI Notes is a over emergency a brandep d-name artment druvisits, g undphysician’s er the terms offi of ce the visits, newly and im pres pos cription ed HSA drug plan. fills—the dependent variables described below. value mentione equival Our “If was ents yod u mu tan or he ch d y comparativ yo woul less ur d tha em inn ployer crea ely ex pect se effective w their ed, ere whi to contri alter contin le only butions. natives. ue6 contri perce In In thi co buti nt mparison, sof ng analysis, th at ose the e11 x evi same pecting perce denc percenta en to t is of mi re tho tir xe ge ed s aft e o that fwho e cr o mpensation this a g wee is re 65t more h had e case. a until con simi fiden you lar t or 4) The generic dispensing rate (GDR), which equals the number of generic fills divided by the total number of Generic 4.59 4.69 -1.00 0.40 All prescriptions 90% -0.95 *** -0.74 *** -0.63 *** -0.69 *** ? Accordin Services.” W g to the ashin ANPRM, gton, DC: C the monthly pension BO; 2012 [c period bene ical providing current informati fit ited 201 statem 3 March ents woul 23]. Availa d also cbl oon on a variety of employ e from: ntain an “understandablee b e expl enefit topics. anation” o f 12 ? GDR increases were due to individuals discontinuing use of brand-name drugs without substituting generic therapy. Finally, to help achieve covariate balance between the study and control groups, higher-order (squared and cubed) terms for See VanDerhei (2013) for an example of the impact of the annuity purchase price on the comparative value of a defined contribution plan retire, some retirement calculators estimate that your employer-sponsored plan could provide a monthly assessment. did not have Combinin a changeg in th co e nfidence “much less” indic aated nd “somew they wou hatl d less” increase categor the ies ir contri resulte bution d in 31 s. percent of those with expected Brand 2.69 2.89 -2.90 0.02 fills (both 30-day adjusted). The GDR is a metric routinely used by pharmacy benefit managers (PBMs) to Maintenance prescriptions -0.71 *** -0.59 *** -0.48 *** -0.58 *** EBRIef is a weekly roundup of EBRI research and insights, as well as updates on surveys, It is, of course, possible that these respondents’ current participation in employment-based plans has already publications http://www.cbo.gov/sites/default/files/cbofiles/attachments/43741-MedicalOffsets-11-29-12.pdf the assumptions behind the lifetime income stream illustrations. Pension benefit statements also would contain a PBMs all continuous After one year routinely measures cite under the the also GDRentered attaine full-replacement HSA plan, 0.43 fe d the via sptecification. heir managem Following ent ofwer guidance pres generic cription offered and drug 0.95 by bAu en fewer stin efits(2011), brand-name (e.g., CVS this analysis Caremark, prescriptions perf 2013). ormed were vs. a defined ben 80%efit plan. Nonmaintenance prescriptions retirement retireme Nonmaintenance prescriptions ag nt es inc of ome 65 or for ea lif rlier e ofi n [$ dicamount] ating that . tKee h -0.24 e val p *** in u e mi was nd that less -0.15 this thandoes *** expecte not daccount . Thi-0.15 s perce fo*** r any ntagre e tireme declined -0.11 nt savings to ***23 per- studies, litigation, legislation and regulation affecting employee benefit plans, while EBRI’s Prior Research assess plan design effectiveness (e.g., CVS Caremark, 2013). provided them the education and information necessary for an appreciation both of the projected total and the statement that projections and lifetime income stream illustrations are estimates, not guarantees. Generic 2.27 2.36 -2.10 0.09 When A a 1:1 nearest-n GDR filled, th th e at con h is an average. high eighbor match within 1/5th of a standard ge in er confi thand th ence e metric was ’s associated underlying, wit secula h thdevi e quart r at tren ion il d e of su of th gges mo e propen nthly ts that isity ncome plan score. de illustrated, sign After an reducing d ot 29 her percent to pro the grams of area those of in Hypertension prescriptions Blog supplements our regular pub -0.08 ** lications, offering commentary -0.03 -0.05 on questions received from -0.08 *** that you might have outside of your employer-sponsored plan.” cent for those expecting to retire after age 65. CVS Caremark. Insights Report 2013. Available at: http://cvscaremarkfyi.com/insights2013/ Studies of CD 70% HPs have examined a variety of prescription-drug-use measures including spending (Charlton et al., monthly income estimate, and thus a greater alignment of those projections with their expectations. Brand 0.69 0.69 -0.20 0.85 Dyslipidemia prescriptions news reporters, policy -0.10 ***makers, and others. EBRI’s -0.14 *** Fun -0.15 damen ***tals of Employee Ben -0.20 *** efit common support, the final study and control groups each contained 13,093 individuals. As reported in Figure 1, standardized the design bottom ed to qu drartile ive gener mention ic drug ed use they ar were e ha ving much their less int or esnode mewhat d effect. less In confident, this analysis, butthe only ado 23p perce tion of nt aof full- those in the Statistical Analysis As 201noted 1; Fronstin abovHypertension prescriptions e, and the Roe lifetime buck, income 2013; Na illustra ir et tions al., 20 contem 09; Par pl eated nte, Fel by dtman he A,N and PRM Ch depe en, nd 2008; heavily Par eon nte, th Fel e use dman, of ce artai nd n Xu, Programs offers a straightforward, basic explanation of employee benefit programs in the Diabetes prescriptions -0.02 -0.03 -0.01 -0.03 differences (% bias) in covariates across the two groups at baseline (in 2006) were lower than the commonly cited 10 per- Resp Does s replac econd ond ement Hea quarti ents ring 60 % C lwere e D an HP the d then with Mon 21 perc ask an thlH e ent yd SA A a of m ge se ount thos neral ries eof Figure ly inquestions l e the d to thir Cause incr d to e quar ase determine: an s tile in Adjus were GDR t in that me thi nt commenced s in category. Expectations Th wiis thin perc th for entag e first Fue ture ydecreas ear and eco d to ntin ued Deyo, R.A., D.C. Cherki bn, and M.A. Ciol. “Adapting a Clinical Comorbidity Index for Use with ICD-9-CM Administrative For each of the dependent variables, a difference-in-differences, ordinary least squares regression model was How Would Define Generic d Contrib private and ution Participants R public sectors. The 1.03 EBRI Datab eact to Lifetime Inc ook on Emplo 1.03 0.10 yee Benef o0.96 me its is a statistical Asthma/COPD prescriptions assumptions. Moreover, the ANPRM requires that -0.05 plan administrators -0.04use only “reasonable -0.03 ” assumptions, -0.01 taking into 2010); number of drug fills (Charlton et al., 2011; Chen, Levin and Gartner, 2010; Fronstin, Sepulveda, and Roebuck, 2 cent threshold of tolerance, which suggested the control group adequately resembled the study group prior to introduction C 7 for o pnt e at rcent rleast ibu tfor io fons ur those ? years in post-i the himpleme ghest quarti ntatio le. n. This was true for overall, maintenance, nonmaintenance, hypertension, Databases.” Journal of Clinical Epidemiology 45 (6), June 1992:613–9. estimated. Brand reference work on employee bene0.32 fit programs and0.35 work force-related issues. -1.60 0.19 Depression prescriptions 0.00 0.02 0.04 0.00 Illustrations? Evidence from the 2014 Retirement Confidence Survey, by Jack ? Whether 50% the monthly amount was more, less or about what they had expected. account June 2013; certai Grn eene pro fet essional al., 20 08; standards Nair etwhe al., n 20 developing 09); medicliafetime tion ad income herenc e illustrations. (Chen, Levi n However, and Gartn it er, provides 2010; lFro egal nstin, safe of the CDHP with HSA (Austin, 2011). Dyslipidemia prescriptions A dys total lipidof emia, 81 percent and asthma of th/C e OPD respondents drugs. Nindicated o changes that in GDR they w wou erel d det conti ected nue for to dia contri betes bute , wh at ile their the cu derrent pressrat ion es GDR after Total Fills 2 VanDerhei, Ph.D., EBRI harbors to the plan administrators under which certain assumptions are deemed reasonable. Sepulveda, and Roebuck, December 2013; Greene et al., 2008; Nair et al., 2009; Parente, Feldman, and Chen, 2008); Was it Useful to Generic Hear an Estimate of the Monthly Retire0.32 ment Income? 0.34 -1.30 0.30 Feldman, Roger, Stephen T. Parente, and Jon B. Christianson. “Consumer-Directed Health Plans: New Evidence on ? Whet All prescriptions her 40% hearing that monthly-amount figu-1.39 re wo *** uld cause th-0.69 em to *** adjust their -0.46 future * contributions -0.50 (and * by how 2 Contact EBRI Publications, (202) 659-0670; fax publication orders to (202) 775-6312. hearin actually g th decli e pr ned ojected over mont the fo hlllow-u y income p per amo iod. unt, while 17 percent replied that this information would lead them to Results The regression model included the study group flag, all covariates used in the propensity score model, as well as References mail order use (Pa Brand rente, Feldman, and Chen, 2008; Feldman, Par 0.45ente, and Christia 0.51 nson, -2.70 2007); an 0.03 d generic and Maintenance prescriptions Subscriptions to -1.00 EBRI Issue Bri *** efs are included -0.61 *** as part of -0.35 EBRI membership, or as part of -0.46 ** a The vast Spen majority ding an of d Utilizati respondents on.” Inquiry said the 44 (1), Spr retirement ing 2 inco 00m 7:26– e projectio 40. n was useful; more than 1 in 3 (36 percent) of much). increase the amount they were contributing. Of those responding that their illustrated value was much less or Orders/ Figure geographic 1 prese region nts and variabl year e m duemmies. ans for Four the study interaction and contr terms ol between groups at the baseline study group (2006 fl)ag . The and sampl the post-period e consisted year of Since the pr 30 esentation % of lifetime income illustrations on defined contribution statements contemplated by the ANPRM Diabetes prescriptions Goda, Gopi Shah, Colleen Flaherty Manchester and Aaron Sojourner, “What Will My Account Really Be Worth? ? In May Nonmaintenance prescriptions 2013, the U.S. Department $199 annual sub of Labor's -0.38 Employe scription *** to e Bene EBRI Not fits -0.08 Securit es andy E ABRI Issue dministration -0.11 Briefs (EBSA) . Change of published Address: -0.04 an adv EBRI, ance brand-name dispensing rates (Feldman, Parente, and Christianson, 2007; Greene et al., 2008; Haviland et al., 2011; the Of course, responden GDR ts can thought rise if tha its t numerator it was very (t uhe seful numbe to hea r of r a ge n ne estimate ric fills) of in tcreases he mont or hly its re denomi tirement nator inco (t m he e they numbe might r of total somewhat less than expected, 35 percent indicated they would increase their contributions. Only 10 percent of those indicators were also Generic added to the models (i.e., the difference-in-differences 0.16 estimators). 0.17 This -0.60 analysis pr 0.61 esents coefficient is 51 a perc relatively ent Hypertension prescriptions ma new les, 42 innovatio –43 per nc , ent little 1100 13th St. NW, Suite 878, Washington, policyhol empirical ders, evi-0.11 d 2enc 3 perc e ee xn ists t spouses, regar-0.05 dina g npotential d DC, 20005-4051, (202) 659-0670; fax number, 34–35 pe par rce ticipant 0.01 nt depe respo ndent nse. childr Go -0.01 da, en. The ? Whether hearing that monthly-amount figure would cause them to adjust their expected retirement age (and in Experimental Evidence On How Retirement Income Projections Affect Saving,” Research Dialogue, TIAA-CREF Fronstin, Paul, and M. Christopher Roebuck. “Health Care Spending after Adopting a Full-Replacement, High-Deductible Parente, notice Feldman, of proposed and rulemakin Chen, 2008 g (A ). NPRM ) focusing on lifet ime income illustrations. The 2014 Retirement Confidence Subscriptions 20% expect fills) decrease from th s, eir or plan, a com an bd inanother ation of the 49 percent two. In tho respo ught nse it to was the some HSA wplan, hat use it was ful. O fo nund ly 5 that percent highte hr ought GDRs it wer was e largely not Brand 0.13 0.14 -0.20 0.88 indicating the illustrated amount w(202) 775-6312 as what they ; e-mail: expected wo subs uld criptions@ebr increase their i.org contri Membe butions. rship Information: Likewise, only Inquiries 10 percent Dyslipidemia prescriptions -0.12 ** -0.09 * -0.06 -0.06 estimates from these specifications with robust standard errors. All analyses were conducted using Stata/MP 12.0. More average age was 30-31 years, and workers had been with their respective employer for a mean of 13 years. Manchester which di an rectio d Sojourn n). er (2013) used administrative data, prior to and following the provision of income and Institute, Issue No. 109, August 2013. Health Plan With a Health S a avings Account: A Five-Year Study.” EBRI Issue Brief no. 388 (Employee Benefit Survey (RCS) included a series of questions concerning mont hly income illustrations similar in many respects to those too reache usefu d thro l anu d Asthma/COPD g10 h re percent ductions th prescriptions in ought utilizatio it was n. n Speci ot usefu fically, l at th all e . num ber of brand fills declined under the new plan for regarding EBRI membership and/or contributions to EBRI-ERF should be directed to EBRI Diabetes prescriptions -0.02 -0.01 0.02 -0.01 of those indicating the illustrated amount was much more or somewhat more than what they expected would increase information about difference-in-difference models can be found in Fronstin and Roebuck (2013). 10% balanc Concerni e projec ng health-servic tions, to meas es uure tilization, the eff pat ect ieon nts the consumed level of contributions. 0.15 to 0.16 in Th paey tien fo t und hosp th itat al days, participa 0.13 nt sto who 0.14 wer emer- e sent The findings on generibc and brand-name dispensing rates are mixed. Feldman, Parente, and Christianson (2007) Research Institute, July 2013). provided by Generic the EBSA’s online President Dallas Lifetime Income Salisbur Calculator. y at th e a 0.12 bove address, (202) 659-0670; 0.13 -1.70 e-mail: salisbur 0.16 y@ebri.org Asthma/COPD prescriptions -0.03 -0.02 0.00 0.00 overall, ? How mai he nt arin enance, g this nonm estimate aintena of retir nce, ement hypertension, income af afecte nd dy dsl their ipide m conf ia—e iden ffe ce cts inth the at ip r e arb sil iste ity d to ove save r four enough years. to Elive ven their contributions. VanDerhei, Jack. “Reality Checks: A Comparative Analysis of Future Benefits from Private-Sector, Voluntary-Enrollment gency department visits, 3.59 to 3.64 outpatient physician office/clinic visits, and just over 10 prescriptions during retireme found that nt CD income HP enrol projectio lees tn es nde wer d e to not use o nly a himore gher pro likely portion to cha of nge bra th nd-nam eir contri e dru bution gs than levels oth , e but r co th horts, ey alswhil o increa e Parente, sed Brand 0.38 0.39 -0.70 0.59 A total of Depression prescriptions 90 percent of those whose illustrated va -0.12 lues **were lower -0.07 than expected fou-0.07 nd the estimates somewhat -0.14 ** or 0% the fill comfortably rates for ge in nerics retirem in e snt. ome cases were lower as a result of the HSA plan with two exceptions: 1) there were ? The vast majority of respondents said the retirement income projectio 3 n was useful; more than 1 in 3 (36 percent) of the 401(k) Plans vs. Stylized, Final-Average-Pay Defined Benefit and Cash Balance Plans” EBRI Issue Brief, no. 387 Fronstin, Paul, Editorial Board: Martín-J. Sepúl Dallas L. Salisbur < veda, an $60,00 y, publisher 0 d M. Christop ; Stephen Blakely her R , editor oebuck. “Cons . Any views expr umes er-Directe sed in this p d He ublicat alth Pla $1ion and th 00,000n or s Re Mo ose o redfu th ce Lo e author ng-T s sho erm uld Depression prescriptions $60,000–$99,999 t their he ye a Source: EBRI estimates based on administrative claims data. a nnua r prior l c oto ntributio the introductio ns by $8n 5 of more the th HSA an the plan control (Figur e grou 2). p. Tw o-thirds While their of tstudy hese fills benefit were ed generics, from using and ad 7ministr 1 percent ative Feldman, and Chen (2008) reported some evidence of less generic drug use. On the other hand, results by Greene et very useful, and nearly as many (86 percent) of those whose values were equal to what they expected also found the Of more those not be agener ascribed to the officers, indic ic ating prescript they ions wofi uld trust lled increase ees, for dyslipi mem c bers, or od ntributio emia other sponsors of the E medicat ns, 69 ion pes rcent in the tho mploye thir ught d e Benefit Research an they d fourth woul d ye in ars, creas Institute, the EBRI Educ and e the 2) tdollar here wer amou ation and e mo nt re by Generic 0.59 0.62 -1.30 0.31 re CDHP=consumer-directed health plan. sponde (Emplo nts yetho e Be ught nefit Rese that it w arch as very Institute, useful June 201 to hear an 3). es timate of the monthly retirement income they might expect Use of Outpatient Physician Visits and Prescription Drugs.” Health Affairs 32(6), June 2013: 1126–1134. were ? How maintena useful nce it was medicatio to hear ns a fo nr estimate chronic disea of ths e e mont manageme hly retirement nt. income they might expect from their plan. data al. (20 with 08) a su six ggested -month that follo th we -up intro to ductio record n changes of a CDHP in di cod ntribu not greatly tion elec intions, fluence it also the use relie of d generic on the u dr se ugs. of hypoth Only Havi etical, land Research Fund, or their staffs. Nothing herein is to be construed as an attempt to aid or hinder the adoption of any pending legislation, regulation, estimates b somewhat or very useful. Don’t A som Know ewhat lower proportion Mu (7 ch 9 Mo perce re Than n t) Exp of ectt ed hose with illustrated values higher gene 10 perc ric COPD=chronic obstructive pulmonary disease. ent, fills fo while r Asthma/COP Brand another 13 D percent in the secon said d they and woul third d y in ecrease ars. 0.21 it by about 250.22 percent. Thr -0.30 ee percent 0.80indicated they from their plan, and another 49 percent thought it was somewhat useful. or interpretative rule, or as legal, accounting, actuarial, or other such professional advice. 11 additional contribution amounts and retirement ages to illustrate future retirement income. Moreover, the authors did et al. (2011) found a lower brand-name dispensing rate. than expectedSource: EBRI estimates based on administrative claims data. found the estimates somew Somewhah t Mo at re or Than ve E ry xpe us cted eful. About What Expected All prescription drug measures are adjusted using 30-days' supply equivalents. Fronstin, Paul, would increase Martín-J. Sepúl it by 50 percent, veda, an while d M. Christop another 8 perce her R nt oe th buck. “Medic ought they awould tion Utiliz doatio uble n a it. nd Adh Sevene re perce nce i nn t a Hea did not lth know a The impact of the HSA plan on the GDR is shown in Figure 3. Overall, the new plan design was associated with a COPD=chronic obstructive pulmonary disease. Presented are coefficient estimates of the CDHP effect in each of four follow-up years (compared to baseline) from ordinary least squares models, ? A total of 17 percent of the respondents indicated that this information would lead them to increase the amount they Somewhat Less Than Expected Much Less Than Expected EBRI No Savings Account- tes is registered in the U. Eligib S.le Patent and T Plan.” America rademarn k Office. Journa IS l of Mana SN: 1085 ?g 445 ed Car 2 1085 e Dece ?4452/9 mber 20 0 $ .50+. 13. 50 4.7 perc which included indicators for male, policyholder, three geographic regions, five age groups, four years, CDHP, and four CDHP ti entage-point rise in GDR after the first year, and—before inclining slightly in the second year—settled mes year to a were contributing. However, of those responding that their illustrated value was much less or somewhat less than Source: Employee Benefit Research Institute and Greenwald & Associates, 2014 Retirement Confidence Survey. interaction terms (presented difference-in-differences estimates), as well as Charlson Comorbidity Index, years of tenure, and household size. Statistical significance based upon robust standard errors denoted as follows: *** p<0.01; ** p<0.05; * p<0.10. ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org ebri.org expected, No No No No No No No No No No Nottttttttttte e e e e e e e e e es s s s s s s s s s s • March 2014 • • March 2014 • • March 2014 • • March 2014 • • March 2014 • • March 2014 • • March 2014 • • March 2014 • • March 2014 • • March 2014 • • March 2014 • 35 percent indicated Vol. 35, No. 3 Vol. 35, No. 3 Vol. 35, No. 3 Vol. 35, No. 3 Vol. 35, No. 3 Vol. 35, No. 3 Vol. 35, No. 3 Vol. 35, No. 3 Vol. 35, No. 3 Vol. 35, No. 3 Vol. 35, No. 3 they would increase their contributions. 13 10 14 11 15 9 7 5 3 6 2 © 2014, Employee Benefit Research Institute ?Education and Research Fund. All rights reserved. A monthl y newsletter from the EBRI Education and Research Fund © 2014 Employee Benefit Research Institute ebri.org Notes • March 2014 • Vol. 35, No. 3 ebri.org Notes • March 2014 • Vol. 35, No. 3 ebri.org Notes • March 2014 • Vol. 35, No. 3 12 4 8