Medicare: Employer Insurance Primary Payer for 11 Percent of Disabled Beneficiaries

Published by the Government Accountability Office on 1990-05-10.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

   w             United   States   General   Accountina   Office

GAO              Report to Congressional Committees

May 1990
                 Employer Insurance
                 Primary Payer for 11
                 Percent of Disabled

             United States

GAO          General Accounting  Office
             Washington, D.C. 20548

             Human Resources    Division


             May lo,1990

             The Honorable Lloyd Bentsen
             Chairman, Committee on Finance
             United States Senate

             The Honorable John D. Dingell
             Chairman, Committee on Energy and Commerce
             House of Representatives

             The Honorable Dan Rostenkowski
             Chairman, Committee on Ways and Means
             House of Representatives

             The Omnibus Budget Reconciliation Act of 1986 (OBRA) (P.L. 99-509)
             made Medicare the secondary payer for medical expenses of certain dis-
             abled beneficiaries covered by large group health plans (UHF%). Section
             93 19 of the act also directed that we determine (1) the number of benefi-
             ciaries for whom Medicare became the secondary payer because of their
             own or a family member’s employment, (2) the resulting annual Medi-
             care cost savings, and (3) the provision’s effect on the employment of
             disabled individuals and family members, and the cost and availability
             of employer health insurance.

             This report describes our work on the first requirement. A subsequent
             report will address the remaining requirements. Our field work on this
             subject, completed on August 28, 1989, was performed in accordance
             with generally accepted government auditing standards. Our methodol-
             ogy, discussed in appendix I, included use of mall questionnaires (see
             app. II and III) and data from the Health Care Financing Administration
             (HCFA) and the Social Security Administration (sA). HCFA and SA are
             agencies of the Department of Health and Human Services (HHS); HCFA
             administers the Medicare Program.

             Medicare is a federal health insurance program that covers most Ameri-
Background   cans aged 65 or older and some disabled persons under age 65. In 1988,
             the program covered on average 3.1 million disabled individuals under
             age 65.’

                  l   100,000 were not working due to disability but met one or more of
                      HCFA'S indicators3 of employee status and also were covered by an
                      employer-sponsored LGHP. HCFA supplied this estimate.

                      HHS provided written comments on a draft of this report and generally
Agency Comments       agreed with our conclusions (see app. IV).

                      HHS does not question that our figures should be given a high degree of
                      confidence but suggests that our estimate of beneficiaries eligible for
                      coverage through a spouse’s employment is extremely conservative. We
                      estimate, using scientific sampling methodology, that this group repre-
                      sents 214,000 beneficiaries or 8 percent of the Medicare disabled popu-
                      lation.HHS estimates from actuarial projections and contractor
                      experience that this group represents 16-17 percent of the disabled

                      We disagree that our estimate is as conservative as HHS suggests. Our
                      sample methodology does exclude nearly 11 percent of the disabled
                      Medicare population for various reasons (see p. 1 l), but this group is not
                      large enough to explain our differences with HHS’Sfigures. Our estimate
                      provides 96-percent confidence that between 7 and 8.6 percent of the
                      sampled disabled population had LGHP coverage through a working
                      spouse or other family member. Moreover, our figure is consistent with
                      an earlier GAOestimate based on data from the Bureau of the Census’s
                       1984 Current Population Survey (see p. 10).

                      HHSstates that our referring to disabled individuals having employee
                      status as having health coverage from a “previous” or “former”
                      employer (1) suggests that we believe HCFA'S definition is inconsistent
                      with the statute and (2) could be used against HCFA in the event of litiga-
                      tion. We agree that our wording could suggest that HCFA'S classifying dis-
                      abled individuals who are not actively working as current employees is
                      inconsistent with the statute and have made appropriate revisions. OBFL4
                      authorized HCFA to define “employee” in regulation. This report is not
                      intended to dispute or endorse the appropriateness of HCFA’S proposed
                      definition. Our wording was an attempt to communicate that these dis-
                      abled individuals, while classified as “employees,” are not actively

                      3HCFA,under proposedregulations,setsforth three situations in which an individual is consideredto
                      have employeestatus. In addition. the proposedregulationsListfive factors,the presenceof any one
                      of which could be the basisfor concludii that an individual is an employee.An exampleis a disabled
                      individual who participates in an employer’sbenefit plan in which only employeesmay participate or
                      has right8 to rehrn to duty if his/her condition improves.

                      page 3                          GAO/‘HlU%99-79Medicare Costa for Disabled Shift to Employer
Page ti   GAO/HlUN@79   Medlm   Costa for Disabled Shift to Employer

GAO       General Accounting Office
HCFA      Health Care Financing Administration
HHS       Department of Health and Human Services
LGHP      large group health plan
OBRA      Omnibus Budget Reconciliation Act of 1986
SSA       Social Security Administration

Paee7                  GAO/HRDW79 Medicare Costafor Ohbled Sldft to Employer
  Appendix I
  Objective,   Skope, and Metlmdolc$y

. Not older than age 63-3/4 as of September 30,1986.’

  Our final universe was composed of 2,600,941 beneficiaries. From this
  universe, we selected a simple random sample of 5,776 beneficiaries, to
  whom we sent a questionnaire during April 1987. In it, we asked about
  marital status, employment, and health insurance coverage, and employ-
  ment of any spouse-all as of December 31,1986, the day before the
  law became effective. The response rate was 88 percent.

  Through the survey, we learned that some disabled beneficiaries might
  be covered by an employment-based LGHPthrough a family member
  other than a spouse. These beneficiaries are disabled adult dependents
  who are (1) primarily sons or daughters of deceased, retired, or disabled
  workers; (2) at least 18 years of age; and (3) disabled before age 22. To
  the 1,046 beneficiaries in this subgroup, we sent a second questionnaire
  to obtain more specific information about health coverage through an
  employed family member. The response rate for the subgroup sample
  was 83 percent. (See app. II and III for the questionnaires.)

  The overall response rate for the two questionnaires was 87 percent,
  and after removing unusable questionnaires, 82 percent. In making our
  statistical estimates, we assumed that the nonresponding and in-
  adequately responding beneficiaries were not materially different from
  those who responded.

   For beneficiaries who indicated coverage under a spouse’s or other fam-
   ily member’s employer-sponsored health insurance plan, we determined
   whether the plan was a large or small group plan. Using various
   resource materials and telephone contacts, we classified employers as
   having fewer than 100 or 100 or more employees as of
   December 31,1986. References we consulted included Dun’s Marketing
   Services Million Dollar Directory, Standard and Poor’s Register of Cor-
   porations, and the American Hospital Association Guide. We contacted
   companies with fewer than 100 employees to confirm their size and that
   they did not belong to a multiple employer health plan or a labor union
   plan that covered employees of at least one company that employed 100
   or more employees. Table I.1 shows our sample results.

   ‘Our design called for dropping beneficiariesolder than 633/4 years as of September30,19&K This
   WBSbecause they would be in “aged Medicare"rather than “disabled Medicare” at the beginning of
   1983when we planned to resurvey beneficiarieswith employedspouseaas part of our study of the
   cost savingsassociatedwith OBRA.Disabledpersons66 and over generally a~ coveredunder Medl-
   careon the basisof ageand may beaffected by its secondarypayer provisions that pertain to aged
   employedindividuals and spouses.Wedo not believethat excluding this group materially affectsour

   Page 9                               GAO/‘EEDm79 Medlm     Costa for LMsabledShift to Employer
Table 1.2 Disabled Medicare
Beneficiaries With LGHP Coverage            Btatua                                                               1999           1998
Throu h a Working Family Member (1986       Disabled Medicare      beneficlarles on the rolls as of July 1
and 19%8)
                                              and living wlthin   the Unlted States, less those who          2,852,856      2,995,174
                                              Have end-staae      renal disease                               I1 18.4631     (139.3751
                                              Have risk-based     health mamtenance organization
                                                 coverage a                                                    (64,189)       (67,391)
                                              Are in extended     trial work’                                  WVJW           (30,~)
                                            Subtotal of disabled beneficiaries                               2.640.204     2.799.409
                                              Percentage of beneficiaries with a family member
                                                who has LGHP coverage                                             7.75           7 75
                                            Total disabled beneficiaries with LGHP coverage
                                              through a family member                                         204,907        213,993
                                            aOur estimate.

                                            Associated with the statistical estimate that 7.75 percent of the disabled
                                            Medicare population has LGHPcoverage through a working spouse or
                                            other family member is a standard error of 0.39 percentage points.
                                            Thus, although we use the point estimate of 7.76 percent for the sake of
                                            simplicity, strictly speaking we can estimate only (with 95-percent con-
                                            fidence) that the percentage is between 6.97 percent and 8.52 percent.
                                            Likewise, we use the point estimates of 204,607 and 213,663 for 1986
                                            and 1988. But strictly speaking, we can estimate only (with 9.5~percent
                                            confidence) that the figure for 1986 is between 183,943 and 225,071;
                                            and for 1988, between 192,178 and 236,148.

                                            The percentage and absolute estimates of the average number of dis-
                                            abled Medicare beneficiaries with LGHPcoverage through a working
                                            spouse or other family member are conservative. They exclude nearly
                                            11 percent of the disabled Medicare population, that is, beneficiaries

                                        l live outside the United States (roughly 3.57 percent of the disabled
                                          Medicare population in 1986);
                                        9 have end-stage renal disease or are end-stage renal disease/disabled
                                          (roughly 4.15 percent of the disabled Medicare population in 1986);
                                        l are in a risk-based health maintenance organization (roughly 2.25 per-
                                          cent of the disabled Medicare population in 1986); or
                                        l are in the final 3 years of extended Medicare eligibility resulting from
                                          the completion of trial work with substantial gainful activity (roughly
                                          0.97 percent of the disabled Medicare population).

                                            Page 11                             GAO/KUBBMB Medicare Cab for LMsabledSldft to Employer
                            Objective,   Scope, and Methodology

                            The end result of these calculations was a rough estimate of 26,413 dis-
                            abled Medicare beneficiaries with employer coverage in a LHGP through
                            their own active employment. We rounded it to 26,000 to reflect the
                            approximation inherent in it.

                            The third group, disabled Medicare beneficiaries who come under OBRA
Determining                 through their nonworking “employee status” relationship with their
Beneficiaries With          employer at the time they became disabled, was delineated by HCFA in
LGHP Coverage               April 1988, about 1 year after we mailed our initial sample question-
                            naire. Because our data collection began before this group was defined,
Through Their HCFA-         we chose not to study this group through an independent sampling pro-
Defined Employee            cess. Instead, we relied on HCFA'S rough approximation of the group’s
                            size. A HCFA analyst estimated that between 50,000 and 150,000 benefi-
Status After Becoming       ciaries are unable to work but still are “employees” and have LGHP insur-
Disabled                    ance through the employer for whom they actively worked at the time
                            they became disabled. We used the midpoint of his estimated range
                            (lOO,OOO),which he agreed would be his “best guess.” This analyst and
                            other HCFA officials noted that HCFA has little information available for
                            estimating this group’s size. In fact, HCTFAofficials advised us that the
                            50,000-150,000 estimate was a preliminary figure and might be too low.
                             They were unable, however, to provide a more precise estimate.

                             Our overall estimate of 340,000 disabled beneficiaries affected by OBRA
Overall Estimate a           thus represents the sum of three very different types of estimates:
Cumulation of
Disparate               l    A statistically valid projection of 214,000 beneficiaries insured through
                             a family member’s employment, based on our sample of 5,776 disabled
Subestimates                 Medicare beneficiaries;
                        l    An estimate of 26,000 beneficiaries insured through their own employ-
                             ment, based on adjustments to and extrapolations from SA estimates
                             from a l-percent sample of Medicare beneficiaries (i.e., roughly
                             30,000 disabled beneficiaries); and
                        l    A HCFA estimate of 100,000 beneficiaries who are insured by their
                             employer at the time they became disabled.

                             Because of the difference in the quality of the underlying estimates, we
                             cannot assign an overall confidence interval to our estimate of
                             340,076 disabled beneficiaries affected by OBRA.

              IQUESTIONS, PLEASE ASK A CLOSE     :

2.   Uhat       is             the      full           name                of        your         spouse?

     (PLEASE                   PRINT           CLEARLY)

                                                         I                                                        I
            First               Name                                      Middle             Name                                 Last        Name

3.   Uhat       is             the      birth            date                   of     your        spouse?

     (PLEASE                   PRINT           CLEARLY)

                                           I                     *I
                              Month              Day                        Year

4.   Uhat   was                  the      working                       situation                   of     your              spouse      on
     December                    31,      1986?                       [CHECK                ONE)

                     Working                   full!                                        PLEASE   TURN '0
       [I                              Or                        --                         PAGE 3 AN0 ANSUER                         ----------
                     part              time                  I                              QUESTION    5

                       Not           uorklng,                         :
       [ 1             but           looking                          I
                       for           work                                                                       .-
                       ilomemaker                                         --------_
       [ 3              Unable             to          work:
       [I               Retired                                                             THIS         ENDS         ftlE      3UE3:13NS.
                                                                                            PUT     THE      ]kJ;$'::VU::::                        /
                                                                      I                     THE AOORESSEJ   E'c,f-,zt q/u
       [ 1              Other
                                                                                            HAIL IT AS SOON iS sclSS:3L:.
       [I              Oon't          know                                                       THANK YOU!

     Page 16                                                               GAO/EItD9@79 Medlcme Costa for Disabled Sblft to Employer
Questionnaire Sent to Disabled
Adult Dependents


                          GAO/HltD99-79 Medicare Costafor Disabled Sbift to Employer

*&~ents From the Department of Health and
Human Services

               DEPARTMENT      OF HEALTH & HUMAN            SERVICES                        Olfice Of InspeCtOrGeneral

                                                                                            w,I”mplo”   DC 20201


            Ms. Janet      L. Shikles
            Director,      Health     Financing       and Policy        Issues
            United    States     General
               Accounting      Office
            Washington,      D.C.       20548

            Dear   Ms.   Shikles:
            Enclosed     are    the Department's            comments on your draft         report,
            "Medicare:        Employer     Insurance         Primary    Payer for 11 Percent        of
            Disabled     Beneficiaries."             The comments represent         the tentative
            position     of the Department             and are subject       to reevaluation       when
            the final      version      of this      report     is received.

            The Department    appreciates              the opportunity           to   comment    on this
            draft report   before     its         publication.
                                                            Sincerely      yours,


                                                            Richard      P. Kusserow
                                                            Inspector      General


                   Page19                             GAO/RBD9079MedieareCostsforMsabledShirttoEm~loyer
                    c!olnments hoIn the DL?partment of Hultb

               Setian     9319 of the Cmihe    B.@et Rgnriliatim              &A of 1986 M            the
               ksP forthedisabledprwvisim.            AlthaqhthaseMte              auer&mt     inclrded
               healthaueragefor"f-~~"as~                                        towh.i&MBaieretid
               ~secarjary,~~-agreemant                         eliminatadvlis,jncl~inthe
               fimlpxwibimmlyaxerage              hasdm~~status.                                  *H.R.
               Pap. No. 1012, 99th "s.,       2d Sesp. 320-321 (1986).             In implemmtin3       the
               EZEir           lhe prcpxed     rule plblisbed      at 55 Fed. Ibeg. 8491 (March 8,
               1990) adepts this position.         lhecaodraftmportreferstothesepeE.ons
               ashaviqhselthcovarage           au6 "previcms 6mplcyars" or "former
               enplayers."     Useofsuchl~clmrlysqysLsthatuY,ke.lieves
               HCPA's*lsentatimis            incus*           with ttm staWtory     autbrizatim,
               ti   uxldbeused       againstHCE4      inthaeventof       litigatim   overHCF?.'s
                     *    tim.    Wehvepmvidedqecificlaquagesubstitutimsbel~to

Now on p. 2

Now on p. 3    Page 4, line     3, delete    "fcamer."
Now on p. 8    Page 7, ParagraFh “1,” "place of qlm"                      shcllld be dwqed        to "current
               erploymmt"in%oplaces.              Ihasam~~dbmedainthecaptim
Now on p 8     m page 8. Also mpga             8, lhl,       rrtkirplaceofemploynwtpriorto
Nowonp    13   Page17, inthecaptim,          Wuygha       PrehabilityFx@oyex?'shmldbe
Nowonp    13   Disbled.~l       Alsompage17,      inlh     3 ofthelastparqqh,
               "predisabilityenployer"       sharldte&aq&towen@oyeratthetiuettq
               became disable%"
Nowon o. 13    paSe18,lines    3 and4,"bef~temhqdisabkd"                        shcllldbechanqedto
               "at the time they beaaae disabled.n
VowonD    73

               lbreisalsoan-infcutmt8                1%dl.idlmaykeoorrected                  if   specific
               reportlanguageiSaddressed.      I.nfoohpte1mpage2,%verage65"
vow on p 1     slxuld be charqxl to "age 65 ard over."
Requests for copies of GAO reports     should be sent to:

U.S. General Accounting    Office
Post Office Box 6015
Gaithersburg,  Maryland    20877

Telephone   202-275-6241

The first five copies of each report   are free. Additional   YII@
$2.00 each.

There is a 25% discount    on orders for 100 or more copies    11~
single address.

Orders must be prepaid bv cash or bv check or IIIUII~~~:~~~~~-... __
out to the Superintendent of ticuments.
-,,-       ,,.-   ,,                                 -,-

United States                         First-Class Mail  L
General Accounting    Office        Postage & Fees #‘I8
Washington,   D.C. 20548                      GAO
                                      Permit No. GlOO
Official   Business
Penalty    for Private   Use $300
Appendix V

Major Contributors to This Report

                          Jane L. Ross, Senior Assistant Director, (202) 275-6195
Human Resources           Terence J. Davis, Assistant Director
Division, Washington,     Peter J. Oswald, Assignment Manager
D.C.                      William A. Eckert, Design and Data Analysis Group

                          Frank G. Pasquier, Issue Area Manager
Seattle Regional Office   Walter R. Eichner, Evaluator-in-Charge
                          Lori D. Pang, Site Senior
                          Evan L. Stoll, Jr., Technical Assistance Group

                          Page 22                  GAO/HltD9&79 Medicare Coats for Disabled shut to Employer
                          u-u 0ffia31Xarr8          "I&&-.                 mlw
      Inmrawe     Primary ker     for 11 Fkrmnt of Disabled             Beneficia%

A significant     wea)oress of the rqnrt      lies inthslaclcof            data orttkepoor
qxalityofdata.         0f the 340,000 individuals        identified       as potentially
beneficiaries     ho are workiq       (26,000 ir&VidUls),           "XtiVe     irdividuals"
m lol-qer achJally working (100,OOO Mviduals),                   and wcare-e         ligible
family m           of anplayed i&iv&duals         (214,000 individuals).            Kh.ile the
estimate of 214,000 is the stmzqststatistically,theothertm
estimates ana less so, withtbeestimate              of Qctive       individuals"       being the
nmtunreliable,        based as it is on 'Wum~guesses~                  fmMedica.z
plrgmm staff.

Hmever, tile       the estimate of 214,0G.l is statistically            strong
 (representw      VP-        tely 8 pwcentofthe          disabledpqxlation),           ax
figuresarenm~mtheorderof16til7percent.                            'IWGWfigumsare
projections    buttressed    byacIxalcontzactor~ieme.Wedonot
beneficiaries     eligible    for wverage thmqh        their qcusds          enplcynmt    is,
if anythhq,      extremely oarservative.

W-Ale thedata    pxc4lem rerhKas smz&k              she reliability      of the report, we
believeCaDhasdonsthe~itcanwith~tisavailable.                                    Wealso
believe there am sufficient         %aveats"       in the report      so that the reader is
p.3tonnotice    cc8mahq       the gulityof          thedata.

Inpkrentatim          of section 6202 of the Cam&us Bxiget Remmiliation                   Act of
1989 myprovide          theHealth      Qxe Financ~~tion                       (HCFA) with a
b&terbasis         fordetennhhgthe           nm&rofdisabled          in3ividualswhohave
employer grcup health cmemge thrtxgb their spxse.                       This legislation
mr&tesaseriesofdatamat&esbetweentkeIntem3l                               Revenue Service,
the Social Searity          prhninistratim     snd HCFK Itwill          assist HCFA in
identifyirq      secom%rypayer-.                In particular,     this provision      will be
useful inidentifyirq           both aged arddisabledkeneficiarieswith                health
averagethmqhtheirq.cuse~sBlployergrmphealthplan.                                 RR15,within
mxe cmp1et.s infomtion              on thisgruq~ofdisabledbmeficiaries.

      page20                             GAO/tlIU29079MedicareCostr,foor~bledShiPttoEmployer
Appendix   ill
Qneatlonnalre  Sent to JHmbled
Adult Dependents

 Page 18                         GAO/HUB@@79   Medicare   Costs for Disabled   Sbif-t to Employer
         Appendix II
         Qnestiod       Sent to Dlaabled
         Medim     Ben.dIduies

                 ;QUESTIONS,  PLEASE ASK A CLOSE      I
                 :FRIENll OR RELATIVE TO ASSIST YOU.;

    5.    Uhat was       the    name of         your    spouse's         employer       on
          December       31,    1986?

          (PLEASE      PRINT     CLEARLY)

    6.    On December          31,     how long        had     your     spouse      worked     for    that

          (PLEASE       PRINT        CLEARLY)

                                           Years                        Months

    7.    At that       time,    did     your      spouse's           employer     have a health
          insurance        program       that      provided           medical     coverage to u?

           (CHECK      ONE)

         L: 1    Yes                        I: 1 NO                                [ ] DON'T         KNOW

                                      THANK YOU!

         Page 16                                GAO/lilUb9W9          Medium     coata for DiMbled    shin   to Employu
Questionnaire Sent to Disabled
Medicare Beneficiaries

                                          U. 5. GENERAL ACCOUNTINC OFFICE
                                            STUDY OF NEOICARE’S   SECONDARY
                                     PAYER PROCRAN FOR DISAEILITY    BENEFICIARIES

                                                                                          <               ---            LABEL

              Please       read      each    of the  questions     carefully.         Answer     only    those
              questions          that     apply   to you or your       spouse     (husband     or wife).       We
              would     like       you to complete        this  questionnaire          as soon     as you
              receive        it.       IF YOU NEED HELP ANSWERING             THESE QUESTIONS,        PLEASE
              ASK A CLOSE FRIEND                OR RELATIVE    TO ASSIST       YOU.

              Please   give          us a telephone       number              where    you          can   be         reached      in
              case   we have          questions     about    your             answers:

                                I                     1
                                    Area   Code                       Phone    Number

              1.         What was your            marital        status       as   of      December            3:,      1986?
                         (CHECK ONE)

                    1 1                                                                                                   [ 1
              loivorced                Widjwed                [ ’
                                                            Single!                     IMaLriled                   Separated!
               --------------               ------------                                 --------              -----------

                                                                                                PLEASE TURN TO
                                                                                                PAGE 2 AND ANSWER
                                                                                                OUESTION 2

                         THIS   ENOS T;IE OUESTIONS
                         PLEASE   PUT THE QUESTIONNAIRE
                         IN THE AOORESSEO ENVELOPE      AND
                         NAIL   IT AS SOON AS POSSIBLE.


                           Page 14                               GAO/HIUMO-79           Medlm         Coats for Disabled         SMft to Employer
                     Appendix     I
                     Objective,       Scope, and Methodology

                     To estimate the average number of disabled Medicare beneficiaries in
Determining          1988 with employer coverage in a LGHPthrough their own active
Beneficiaries With   employment, we used Social Security statistics and discussions with SA
LGHP Coverage        officials. First, we estimated the number of those on trial work with sub-
                     stantial gainful activity (generally those earning at least $300 per
Through Their Own    month) or in the 39-month period of extended Medicare eligibility while
Current Employment   performing substantial gainful activity. We assumed that disabled Medi-
(Trial Work)         care beneficiaries working at lower levels of income generally would not
                     have health coverage from their employer and thus would not be subject
                     to the Medicare secondary payer provision.

                     To make our estimate, we obtained figures as of December 1987 and
                     December 1988 from SA for the number of persons in the final 36
                     months of substantial gainful activity after completion of trial work. SA
                     cautions that these figures were based on a l-percent random sample.
                     They included an unknown (but presumed small) number of other bene-
                     ficiaries whose Social Security cash benefits had been suspended (for
                     example, prisoners or beneficiaries who had refused vocational rehabili-
                     tation). Averaging the two yearend figures, we estimated the mid-1988
                     figure (31,960) and used it to represent the average number of benefi-
                     ciaries in 1988. We reduced this figure to 30,000 beneficiary years to
                     provide some recognition of the non-trial-work beneficiaries. Then we
                     increased it by one-third to incorporate an estimate of those in trial
                     work with substantial gainful activity or the first 3 months of substan-
                     tial gainful activity after completion of trial work. These adjustments
                     provided an estimate of 40,000 beneficiaries in either trial work with
                     substantial gainful activity or the 39 months of extended Medicare eligi-
                     bility while performing substantial gainful activity after completing
                     trial work.

                      We assumed the percentage of trial workers without health insurance
                      was the same as that for the U.S. population as a whole. Therefore, we
                      adjusted this number downward to reflect the percentage of civilian
                      nonagricultural workers (24.1 percent) without employer-sponsored
                      health insurance. (This percentage came from an issue brief published
                      by the Employee Benefit Research Institute, based on March 1986
                      Bureau of the Census data.) We also assumed that the percentage of cov-
                      ered trial workers in small rather than large group health plans was the
                      same (13 percent) as that obtained from our sample of family members
                      who provide employer-sponsored health insurance for disabled Medi-
                      care beneficiaries. Therefore, we further adjusted downward the num-
                      ber of trial workers to reflect this factor.

                      Page 12                                  GAO/HlUMlW9   Medicare   Costa for J&tabled   Shift to Employer
                                        Appendix I
                                        Objective, Scope,and Methodology

Table 1.1: Employment and Health
Insurance Status of Family Members of                                                                          Percent of
Disabled Beneficiaries (1986)                                                                                      usable      Error range’
                                        status                                                    Total        responses          (percent)
                                        Sample populatron of drsabled Medicare
                                          beneficianes                                            5,776                N/A              NIA
                                        Usable resoonses                                          4,738              100.00             N/A
                                        Beneficianes wrthout a workrng famrly
                                          member,                                                 3.699               78.07             1 20
                                          Beneficrary is single, drvorced, or wrdowed,
                                            and not a disabled adult dependent                    1,758               37 10             1 40
                                          Spouse IS lookmg for work                                  84                177              0 39
                                          Family member IS a homemaker, unable to
                                            work, retrred. or other                               1,154               24 36             1 24
                                          Beneficrary IS a disabled adult dependent
                                             without a workina famrlv member                        703               1484              1 03
                                        Beneficranes wcth a working family member:                1,039               21 93              1 20
                                          Beneficrary has a working famrly member
                                            but no employer-related health Insurance                618               13.04             0.98
                                          Beneficrarv has an emolover-swnsored
                                            health Insurance elan: ’                                421                8 89             0 83
                                             Large group health plan                                367                 7 75            0 78
                                             Small qrouo health elan                                 54                 1.14            0 32
                                        TVs are 95.percent confident that the disabled Me&care population does not drffet from our sample by
                                        more than the number of percentage pomts shown

                                        Using our sample results, we estimated for the entire disabled Medicare
                                        population the number of Medicare beneficiaries with LGHP coverage
                                        obtained through a working family member (see table 1.2). For the 1988
                                        estimates, we assumed that the percentage of JAXP coverage2 would not
                                        have changed dramatically in the l-1/2 years between the “as of” date
                                        of the sample-December 3 1,1986-and       July 1,1988.

                                        2Another 1.14percenthad similar coverage under smafl grouphealth plansnot affected by OBRA.
                                        Thesetwo groupsconstituted an &imxted 9 percentof disabledbeneficiarieswho had coverage
                                        through a WOrkklgfamily member-consistent with an estimatewe provided in an earlier report, The
                                        Chgreas Should 0msider AmendingThe MedicareSecondaryPayer ProvisionsTo Include Disability
                                        Beneficiaries(GAO--1021,         basedon data from the Bureauof the Census’s1984Cwent Pop
                                        ulation Survey.

                                        Page 10                           GAO/llRMl@79 Medicare Costs for LXsabledShift to Employer
Appendix I

Objective, Scope,and Methodology

                         Amendments to Section 1862(b) of the Social Security Act contained in
                         the Omnibus Budget Reconciliation Act of 1986 made Medicare the sec-
                         ondary payer for hospital and other medical expenses of disabled Medi-
                         care beneficiaries covered under a LGHP through their own or another
                         family member’s current employment. These amendments were effec-
                         tive January 1, 1987. OBRA also directed the Comptroller General to
                         study and report to Congress the effect of these amendments. The
                         report was to include an estimate of the number of disabled Medicare
                         beneficiaries for whom Medicare had become secondary payer.

                         In responding to this objective, we considered three groups of disabled
                         Medicare beneficiaries for whom employer-provided health insurance
                         has become the primary payer, namely those with LGHP coverage

                         1. a spouse’s current employment or a parent’s or other family member’s
                         current employment,

                         2. their own current employment (trial work) or its aftermath, and

                         3. their continued employee status, as defined by HCFA (see p. 3), after
                         becoming disabled.

                         For each group of beneficiaries, we used a different methodology.

                         To estimate the number of beneficiaries in the first group, we obtained
Determining              from HCFA a file that included all disabled persons receiving Medicare
Beneficiaries With       coverage as of September 30, 1986. We adjusted this file to obtain only
LGHP Coverage            beneficiaries with the following characteristics:
Through a Family     l Alive as of September 30, 1986;
Member’s Current     l On Medicare due to disability as of September 30, 1986;
Employment           . Not disabled due to end-stage renal disease (a related secondary payer
                       provision applies to disabled Medicare beneficiaries whose entitlement is
                       solely based on end-stage renal disease);
                     . Living within the United States;
                     l Not in a health maintenance organization that provides beneficiaries
                       with all Medicare-covered services for a fixed amount per person; and

                         pages                    GAO,‘HRBi3O-79MdcareCustsforMsabled   ShifttnEmployer

Appendix I
Objective, Scope, and
Appendix II                                                                                           14
Questionnaire Sent to
Disabled Medicare
Appendix III                                                                                          17
Questionnaire Sent to
Disabled Adult
Appendix IV                                                                                           19
Comments From the
Department of Health
and Human Services
Appendix V                                                                                             22
Major Contributors to
This Report
Tables                  Table I. 1: Employment and Health Insurance Status of                          10
                            Family Members of Disabled Beneficiaries (1986)
                        Table I.2 Disabled Medicare Beneficiaries With LGHP                            11
                            Coverage Through a Working Family Member (1986
                             and 1988)

                        Page 6                 GAO/ABD69.79 Medicare Coatsfor Disabled Shift to JZm~lo~er
 working. In most cases, HHS’s proposed revisions clearly achieve this. In
 several places, however, we used language different from HHS’s to clar-
 ify this distinction.

 HHS made additional comments of a more technical nature. These were
 considered and revisions were made to address them.

 We are sending copies to the Secretary of Health and Human Services,
 the Administrator of the Health Care Financing Administration the
 Commissioner of Social Security, the Director of the Office of Manage-
 ment and Budget, and other interested parties. Copies also will be made
 available to others upon request. Please telephone me at (202) 275-545 1
 if you or your staff have any questions about this report. Other major
 contributors are listed in appendix V.

‘Janet L. Shikles
  Director, Health Financing and
    Policy Issues

 Page 4                   GAO/‘llRDW79 lbiedlcare Co&a for Disabled Shift to Employer

                             Since 198 1, the Congress has enacted a series of amendments to the
                             Social Security Act that make Medicare the secondary payer if benefi-
                             ciaries are covered by certain other health plans. Prior to these amend-
                             ments, Medicare paid first (as primary payer) and the other plans paid
                             at least part of what Medicare did not pay (as secondary payer).

                             OBRA’S secondary payer provisions applied to three groups of disabled
                             beneficiaries who have “employee” status and LGHP coverage:

                             Individuals who have coverage through a working spouse or other fam-
                             ily member (the largest group),
                             Individuals who are employed in a trial work program,2 and
                             Individuals who are unable to work but whose link to an employer con-
                             stitutes employee status. Factors that HCFA lists in proposed regulations
                             as indicative of employee status include having a legally enforceable
                             right to return to work in the event disability ceases and participating it
                             an employer’s benefit plan in which only employees may participate.
                             HCFA published proposed rules defining employee status for disabled
                             beneficiaries on March 8, 1990.

                             We estimate that OBRA made Medicare the secondary payer for 340,000
Disabled Medicare            disabled beneficiaries during 1988, or approximately 11 percent of the
Beneficiaries Affected       3.1 million disabled Medicare beneficiaries:
                         l 214,000 had health insurance coverage through a family member’s LGHI
                           We based this estimate on a random sample of 5,776 beneficiaries,
                           drawn from HCFA files, to whom we sent a questionnaire regarding their
                           families’ employment and health insurance.
                         . 26,000 were in trial work programs, or had extended eligibility after
                           completion of trial work, and had health insurance coverage through
                           their employer’s LGHFJ.We estimated this group from HCFA and Social
                           Security Administration data.

                             ‘Under this program,disabledpersonscan attemptto m-enterthe work force by working up to
                             9 months,not neceSSari  ly corwcutively, and cont.inueto receivemonthly disability benefits and Me
                             carecoverageeven though they earned“substantial income”(defmedBSat least$300 per month).
                             disabledpersonwho continuesto work beyondthe 9-monthtrial work period can retain Medicare
                             benefits for up to an additional 39 months.