Racial and Ethnic Disparities in Prescription Coverage and ......plans, or M+C. (Our models excluded...

This study compared drug coverage andprescription drug use by race and Hispanicethnicity for Medicare beneficiaries withthree chronic conditions: diabetes, hyper-tension, or heart disease. We found thatamong beneficiaries without any drug cov-erage black persons and Hispanics used 10to 40 percent fewer medications, on aver-age, than white persons with the same ill-ness, and spent up to 60 percent less in totaldrug costs. Having drug coverage some-what lessened these dif ferences although theef fect was consistent with only M+C pre-scription benefits. Substantially lower med-ication use remained for dually eligibleblack beneficiaries and Hispanics withemployer-sponsored drug benefits.

INTRODUCTION

This study examines the access thatblack and Hispanic Medicare beneficiarieshave to prescribed drugs for chronic con-ditions. We know little about how race andethnicity influences medication use despitesubstantial research showing that, for mosthealth care services, minority beneficiariesuse fewer services compared with whitepersons (Gornick, 1999, 2000, 2003;Gornick et al., 1996; Gornick, Eggers, andRiley, 2001; Murray, 2000). One exceptionto this pattern—and a possible indicator of

medication underuse—is a higher thanaverage need for procedures used to treatthe complications of chronic illnesses. Forinstance, elderly black beneficiaries arethree to four times more likely than whitebeneficiaries to undergo amputations oflower limbs or implantations of shunts forrenal dialysis due to uncontrolled diabetes(Gornick, 1999, 2000; Gornick et al., 1996).Such differences have been generallyinterpreted as evidence of widespreadinsensitivity in the acute care setting(Mayberry, Mili, and Ofili, 2000). An alter-native explanation is that minority benefi-ciaries may be facing persistent problemsin getting necessary medications thateventually lead to the most debilitatingeffects of unmanaged chronic illness.

Black and Hispanic Medicare beneficia-ries may be particularly susceptible tomedication underuse for economic rea-sons because outpatient prescription drugswill not be included under the traditionalMedicare benefit until 2006. Until then,beneficiaries must negotiate some form ofdrug benefits or else pay out-of-pocket fortheir medication expenses. Options forgaining drug coverage include earningcomprehensive retiree health benefits,buying personal insurance, enrolling intoM+C plans with a drug coverage option, orqualifying for public assistance (e.g.,Medicaid or State pharmaceutical assis-tance programs). Some of these avenuesmay be less accessible to minority popula-tions, while other types place considerabledemands on personal income and savings.For example, private drug plans often require

HEALTH CARE FINANCING REVIEW/Winter 2003-2004/Volume 25, Number 2 63

Becky Briesacher and Rhona Limcangco are with the Universityof Maryland, School of Pharmacy. Darrell Gaskin is with theJohns Hopkins Bloomberg School of Public Health. Theresearch in this article was funded by the Henry J. KaiserFamily Foundation. The views expressed in this article are thoseof the authors and do not necessarily reflect the views of theHenry J. Kaiser Family Foundation, the University of Maryland,the Johns Hopkins Bloomberg School of Public Health, or theCenters for Medicare & Medicaid Services (CMS).

Racial and Ethnic Disparities in Prescription Coverage andMedication Use

Becky Briesacher, Ph.D., Rhona Limcangco, M.Pharm., and Darrell Gaskin, Ph.D.

substantial cost sharing in the form ofcopayments for each prescription fill,deductibles, and monthly premiums.Access to employer-sponsored drug cover-age depends on consistent employmentopportunities in industries offering retireebenefits. Enrollment into M+C plans withdrug coverage has becoming increasinglylimited for residents of Southern Stateswhere many minorities live. Similarly, onlythree States offer substantial drug assis-tance programs for Medicare beneficiaries,and they are all in the Northeast (NewJersey, New York, and Pennsylvania). Forchronically-ill black and Hispanic benefi-ciaries with regular medical needs andscarce personal resources, inadequatedrug coverage may translate to disparitiesin medication use.

Few studies have compared racial dis-parities in medication use by drug insur-ance status despite the intuitive relation-ship between affording drug therapies andmanaging disease (Espino et al., 1998;Fillenbaum et al., 1993; Fillenbaum et al.,1996; Nelson, Norris, and Mangione, 2002;Svetkey et al., 1996; White-Means, 2000).We used the wide variation in prescriptiondrug coverage among Medicare beneficia-ries to study prescription spending and useby race and Hispanic ethnicity for threegroups with persistent medication needs—those with diabetes, hypertension, or heartdisease. Cardiovascular disease and dia-betes are two of three chronic conditions(the third is HIV/AIDS) identified as tar-gets for Federal initiatives to eliminateracial/ethnic disparities in health (U.S.Department of Health and HumanServices, 1999). We also examined benefi-ciaries with hypertension as a conditioncommonly identified as sensitive to raceand ethnicity, in terms of disease preva-lence, treatment selection, and health careuse (Sung et al., 1997). All three conditionsare commonly treated with drug therapy to

minimize the debilitating effects of pro-gressive disease. Our main objective wasto distinguish whether drug coveragelessens or eliminates racial and ethnic dif-ferences in the use of medications forchronic conditions, and whether certaintypes of coverage are more effective atimproving access.

METHODS

Data

We used data from the 1999 MCBS Costand Use File to study prescription drugcoverage, expenditures, and use across dif-ferent race and ethnic groups. The MCBSis a longitudinal panel survey of a repre-sentative national sample of the Medicarepopulation conducted under the auspicesof CMS. Beginning in fall 1991, more than12,000 Medicare beneficiaries have beeninterviewed three times a year using com-puter-assisted personal interviewing. Eachrespondent is followed for up to 4 years.MCBS interviewers collect extensive infor-mation on individuals’ use and expendi-tures for health services including sourceof payment, as well as information onhealth insurance, access to care, healthand functional status, socioeconomic sta-tus, and demographic characteristics.Prescription drug utilization data in theMCBS are based on self-reports of eachprescription filled and refilled during theyear. To assure accurate recall, respon-dents are asked to keep bill records andprescription containers to show interview-ers during the three yearly interviews.

Sample

Our sample consisted of non-institution-alized Medicare beneficiaries age 65 orover who identified their race and Hispanicethnicity in the survey. American Indians,

64 HEALTH CARE FINANCING REVIEW/Winter 2003-2004/Volume 25, Number 2

Asians, or Pacific Islanders were excludedbecause their sample sizes were insuffi-cient to provide reliable estimates. In addi-tion, we excluded persons who did not pro-vide specific racial/ethnicity information.We created three mutually-exclusive racialand ethnicity groups: non-Hispanic white,non-Hispanic black, and Hispanic. Thesecategories conform to recommendationsby minority health task groups to identifyHispanic Americans independently fromrace (Zambrana and Carter-Pokras, 2001).We used self-reports of diseases to identifyindividuals with diabetes, hypertension, orheart disease. Prior study has shown thataccuracy of self-reported health varies bymedical conditions, but is generally notsensitive to race or ethnicity (Bergmann etal., 1998). In comparing our three diseasegroups, we found a large overlap amonghypertensive beneficiaries who also reportedheart disease and/or diabetes. Thus, ouranalysis of persons with hypertensionexcludes those who also had heart diseaseor diabetes. Our unweighted samples sizesare: 4,355 beneficiaries with heart disease(n=3,760 white persons, n=334 black per-sons, n=261 Hispanics), 1,568 with dia-betes (n=1,196 white persons, n=218 blackpersons, n=154 Hispanics), and 2,157 withhypertension (n=1,762 white persons,n=235 black persons, n=160 Hispanics).

Statistical Analysis

Our analysis examined five measures ofprescription use and expenditures: (1)annual number of prescriptions filled, (2)total prescription cost, (3) average unitprice (ratio of total prescription cost overnumber of prescription filled), (4) out-of-pocket costs, and (5) use of medicationsfrom broad therapeutic classes commonlyrecommended in the management of thestudy condition. These include oral hypo-

glycemic drugs and hormones such asinsulin for diabetes, and cardiac agents,cardiovascular drugs and diuretics forhypertension and heart disease. Our mainstudy variables were indicator variables ofrace and Hispanic ethnicity. We furtherclassified each group by source of pre-scription drug coverage. For covariates weincluded socioeconomic traits (age, sex,and income relative to the FPL) and sever-al measures of health status: self-ratedhealth status, number of comorbidities,and limitations in daily living activities orinstrumental activities of daily living.

We calculated descriptive statistics ofselected personal characteristics, prescrip-tion use, and expenditures for each diseasegroup stratified by race and ethnicity andsource of drug coverage. In our bivariateanalyses, we tested for statistically signifi-cant differences in prescription use ofblack persons relative to white persons andHispanics relative to white persons. Forthe multivariate models, we limited oursample to people with any drug use andone of four types of drug coverage: no drugcoverage, Medicaid, employer-sponsoredplans, or M+C. (Our models excluded per-sons with Medigap drug plans or otherpublic drug assistance programs due tosmall sample sizes.) We then estimatedfour identically specified linear and logisticregressions:

y = α + β1x1 + β2x2 + e

where (y) in the linear form takes thelogged form of the number of medications,total medication costs, and out-of-pocketcosts, (α) is the constant, (x1) is a set ofdummy indicators for black persons andHispanic ethnicity, (x2) is a set of covariatespreviously described, and e is an error term.For the logistic regression, (y) is a binaryvariable indicating whether beneficiaries


received recommended drug treatmentsand (x1) and (x2) are the same as in the lin-ear regressions.

Each set of models was run separatelyfor each disease group and stratified bymajor type of drug coverage for a total of48 models (three disease groups by fourdrug coverage types by four outcome mea-sures). This approach was taken to isolateany racial or ethnic disparities in medica-tion use among people with the same dis-ease and same type of drug coverage.Preliminary analyses showed that the aver-age effect of insurance status varied con-siderably by race/ethnicity and type of ill-ness, which prohibited simple pooledregressions. In each model we examinedthe relative differences in prescriptionexpenditures and utilization of black per-sons and Hispanics compared with whitepersons. Diagnostic tests included an over-all Chow test for statistically different coef-ficients and variance for the three racial/ethnic groups (F=6.11, p <0.001). Allregressions were statistically significant,and the R-squared ranged from 8 to 19 per-cent). (Regression outputs are availablefrom the authors upon request.) All analy-ses used survey estimator modules inSTATA® Software Version 7 (Stata, CollegeStation, TX) to estimate standard errors inthe population weighted analyses.

RESULTS

Sample Description

In 1999, the size of the Medicare popula-tion age 65 or over and living in the com-munity numbered approximately 31 mil-lion people (26.6 million white persons, 2.3million black persons, and about 2.0 mil-lion Hispanics) (Table 1). Among them,nearly one-half (43.6 percent) reportedhaving some form of heart disease, anoth-er 20 percent had uncomplicated hyperten-

sion, and 16 percent had diabetes. Theprevalence of heart disease and hyperten-sion was roughly similar across the threeracial/ethnic groups except for diabeteswhich is far more common in minority ben-eficiaries: one in four black and Hispanicbeneficiaries reported having diabetescompared with only 15 percent of whitebeneficiares. The rest of this table showshow beneficiaries with chronic illnesses(especially those with heart disease anddiabetes) had socio-economic and healthdisadvantages compared with the generalMedicare population. Beneficiaries withheart disease, hypertension, or diabeteswere generally older than the average ben-eficiary, and they more often had incomesbelow the FPL. About one-third of benefi-ciaries with heart disease and diabetesdescribed their health as fair or poor com-pared with about one-fifth of the totalgroup. Lastly, the vast majority (97 to 82percent) of the three disease groups hadother chronic illnesses in addition to theirstudy condition.

Prescription Drug Coverage

Table 2 describes the different sourcesof drug coverage for beneficiaries with dia-betes, hypertension, or heart disease.Although most white, black, and HispanicMedicare beneficiaries maintained someform of prescription drug coverage in1999, the type of coverage differed greatly.In general, minorities relied far more heav-ily than the white individuals on public pro-grams for assistance with prescriptiondrug costs. Between 25-29 percent ofHispanic and black beneficiaries receiveddrug benefits from Medicaid—througheither traditional Medicaid or the QualifiedMedicare Beneficiary Plus (QMB+) pro-grams—compared with only 5 percent ofwhite persons. Other public drug coveragewas more comparable across the three


groups through sources such as theDepartment of Veterans Affairs and State-funded pharmacy assistance programs,which ranged from 12 to 17 percent.

Conversely, minority beneficiaries wereless likely to have private sources of drugcoverage. For instance, in 1999 only 12 per-cent of Hispanics and 25 percent of black


Table 1

Select Characteristics of Medicare Beneficiaries, by Chronic Conditions: 1999

Characteristic All Diabetes Heart Disease1 Hypertension2

Total (in Millions) 30.656 4.916 13.485 7.052

Percent by Race/EthnicityWhite 85.6 78.0 87.1 82.7Black 7.6 12.0 7.0 9.5Hispanic 6.7 10.0 5.9 7.8Age65-69 Years 25.0 22.8 20.3 22.870-74 Years 26.5 30.6 24.5 26.375-79 Years 22.8 24.0 25.2 24.480 Years or Over 25.6 22.6 30.0 26.5SexFemale 57.8 53.9 54.8 63.5Income as Percent of FPL≤100 20.4 24.9 21.1 19.6101-200 31.5 34.1 33.4 31.3>200 48.1 41.0 45.5 49.1Self-Reported Health StatusPoor 6.7 10.3 8.7 3.0Fair 16.9 23.9 22.0 13.0Good/Excellent 78.2 65.8 69.3 84.0Burden of Chronic Conditions3

1 16.4 3.3 4.3 18.12 23.8 12.2 16.3 42.03 or More 51.2 84.5 79.4 39.91 Includes individuals reporting they have been told they have angina/coronary heart disease, myocardial infarction, atherosclerosis, or other heartdisease.2 Includes individuals reporting they have been told that they have hypertension. It excludes hypertensive individuals with diabetes and/or heart disease.3 Chronic conditions include self-reported conditions of hypertension, stroke, heart disease (angina/coronary heart disease, myocardial infarction, atherosclerosis, other heart disease), diabetes, arthritis (osteoarthritis, rheumatoid arthritis), Alzheimer’s disease, osteoporosis, mental disorder, lungdisorder (emphysema, chronic obstructive pulmonary disease, asthma), and cancer.

NOTE: Excludes beneficiaries with end stage renal disease entitlement.

SOURCE: Centers for Medicare & Medicaid Services: Data from the Medicare Current Beneficiary Survey Cost and Use File, 1999.

Table 2

Prescription Drug Coverage of Medicare Beneficiaries with Chronic Conditions1, byRace/Ethnicity: 1999

BeneficiariesCoverage2 White Black Hispanic

PercentM+C 17.7 **21.1 **25.2Medicaid 5.0 **24.9 **29.3Employer Sponsored 38.2 **25.9 **12.6Medigap 15.6 ** 7.3 ** 6.7Other3 12.5 **17.0 **15.6No Coverage 23.7 **21.3 **23.1

** p<0.05 black non-Hispanics compared with white non-Hispanics and Hispanics compared with white non-Hispanics.1 Chronic conditions: diabetes, hypertension, or heart disease.2 Categories are not mutually exclusive.3 Includes other public (such as State-funded prescription assistance program and Department of Veterans Affairs) and unknown source.

NOTE: n=22.6 million.


beneficiaries obtained drug coverage fromemployer-based insurance, compared withover 38 percent of white beneficiaries. Whitepersons are also far more likely to haveMedigap drug coverage, at rates two andthree times higher than black persons andHispanics, respectively. Only M+C coveragefavors minority populations: Hispanics andblack persons more often had drug coveragefrom Medicare HMOs than white persons(25.2 and 21.1 percent versus 17.7 percent).

Prescription Use and Expenditures byDiseases

In Table 3, we see how drug coverageinfluences the medication use and spend-ing of diabetic beneficiaries by race and

ethnicity. Without any drug coverage,white persons used one-third more med-ications, on average, than black personsand Hispanics, and spent 20 to 40 percentmore, respectively. Hispanic beneficiariestended to fill more expensive medicationsalthough white persons had the bestaccess to diabetic agents: nearly 70 percenttook insulin or oral diabetic agents duringthe year compared with only about 50 per-cent of black persons (p<0.05) or Hispanics(p<0.05). Having drug coverage somewhatlessened these differences although theeffect was consistent with only managedcare benefits. For minority beneficiarieswith Medicaid drug benefits, medicationuse remained much lower than for whitebeneficiaries, although average prescription


Table 3

Prescription Use and Expenditures for Medicare Beneficiaries with Diabetic Conditions1, by DrugCoverage and Race/Ethnicity: 1999


Total (in Millions) 3.84 0.6 0.47

No CoverageAverage Annual Drug Use 31.2 25.2 25.2Average Unit Price $37.50 **$28.40 $31.70 Average Out-of-Pocket Cost $1,095.40 **$670.5 $889.53 Percent Out-of-Pocket to Total Cost 97.8 93.7 100.0Percent Prevalence of Any Diabetic Drug Use2 69.9 **45.0 **52.3Medicaid Average Annual Drug Use 51.6 39.6 39.6Average Unit Price $32.00 $35.10 $33.50 Average Out-of-Pocket Cost $334.80 *$182.50 $181.90 Percent Out-of-Pocket to Total Cost 27.9 19.2 18.2Percent Prevalence of Any Diabetic Drug Use2 71.20 **49.5 **73.2Employer Sponsored Average Annual Drug Use 34.8 *28.8 *26.4Average Unit Price $56.40 $53.30 $71.83 Average Out-of-Pocket Cost $463.40 $411.80 $363.93 Percent Out-of-Pocket to Total Cost 29.4 *35.3 26.23Percent Prevalence of Any Diabetic Drug Use2 68.8 **54.9 51.83M+CAverage Annual Drug Use 32.4 30.0 33.6Average Unit Price $31.70 $28.40 *$38.50Average Out-of-Pocket Cost $464.40 $456.50 $353.30 Percent Out-of-Pocket to Total Cost 53.0 51.5 42.2Percent Prevalence of Any Diabetic Drug Use2 73.6 **55.5 **60.5

** p<0.05 black non-Hispanics compared with white non-Hispanics and Hispanics compared with white non-Hispanics.

*p<0.10 black non-Hispanics compared with white non-Hispanics and Hispanics compared with white non-Hispanics.1 Includes individuals reporting they have been told that they have diabetes.2 Includes hormone such as insulin and oral hypoglycemic drugs.

NOTES: n=4.9 million. Unweighted n=1,196 white; n=218 black; n=154 Hispanic.


prices dropped to relatively similar levels.Out-of-pocket expenses also declined con-siderably although dually eligible whitebeneficiaries paid about twice as much fortheir medications compared with Hispanics(p>0.05) or black persons (p<0.05). Accessto diabetic agents became more similarbetween white and Hispanic persons withMedicaid, but not for black persons.Employer-sponsored drug benefitsshowed little ability to reduce racial/ethnicdisparities in drug use except in out-of-pocket costs. With M+C coverage, we see aleveling out of prescription use and spend-ing across the three groups. Average druguse is comparable between white personsand Hispanics and only 10 percent lowerfor black persons (p>0.05). Average unitdrug prices also look alike, although

Hispanics still used slightly more expen-sive medications. Out-of-pocket costs andgenerosity of coverage are nearly identicalfor white and black persons with M+C cov-erage, although Hispanics have slightlyless generous coverage (p>0.05). Access todiabetic agents is still problematic, though,as white persons are about 30 to 12 percentmore likely to use these medications thanblack persons or Hispanics, respectively(p<0.05).

In Table 4, we see similar patterns ofhigher spending and medication use forwhite beneficiaries with heart disease com-pared with Hispanic and black personswith the same condition, except for thosewith M+C drug coverage. Without any pre-scription plan, white persons filled four toseven more prescriptions during the year


Table 4

Prescription Use and Expenditures for Medicare Beneficiaries with Heart Disease1, byRace/Ethnicity: 1999


Total (in Millions) 11.75 0.95 0.80No CoverageAverage Annual Drug Use 27.6 24.0 **20.4Average Unit Price $36.70 $31.10 ** $46.8Average Out-of-Pocket Cost $944.60 **$677.7 $752.60 Percent Out-of-Pocket to Total Cost 93.6 93.1 100.0Percent Prevalence of Any Heart Disease Drug Use2 76.7 ** 79.4 ** 67.2 MedicaidAverage Annual Drug Use 44.4 39.6 *38.4Average Unit Price $36.20 $31.80 $34.40 Average Out-of-Pocket Cost $324.10 ** $168.6 **$200.2Percent Out-of-Pocket to Total Cost 27.8 **19.9 **15.5 Percent Prevalence of Any Heart Disease Drug Use2 85.1 **87.8 **89.1 Employer Sponsored Average Annual Drug Use 31.2 27.6 27.6Average Unit Price $55.30 *$46.5 ** $41.2Average Out-of-Pocket Cost $422.00 $347.50 $307.60 Percent Out-of-Pocket to Total Cost 31.6 34.1 32.9Percent Prevalence of Any Heart Disease Drug Use2 81.7 ** 84.2 **85.5M+CAverage Annual Drug Use 29.3 30.0 29.3Average Unit Price $33.40 $32.20 $32.40 Average Out-of-Pocket Cost $386.20 $409.10 $308.70 Percent Out-of-Pocket to Total Cost 50.3 48.3 44.6Percent Prevalence of Any Heart Disease Drug Use2 77.6 ** 88.6 ** 78.2

**p<0.05 black non-Hispanics compared with white non-Hispanics and Hispanics compared with white non-Hispanics.

*p<0.10 black non-Hispanics compared with white non-Hispanics and Hispanics compared with white non-Hispanics.1 Includes individuals reporting they have been told that they have hypertension. It excludes hypertensive individuals with diabetes and/or heart disease.2 Includes cardiac drugs, cardiovascular drugs, and diuretics.

NOTES: n=13.5 million. Unweighted n=3760 white; n=334 black; n=261 Hispanic.


and spent 20 percent more compared withblack persons (p>0.05) and 28 percentmore compared with Hispanic persons(p<0.05). Uninsured Hispanic beneficiariestended to fill more expensive medications,but their access to common heart thera-pies was the lowest: only 67 percent tookcardiac agents, cardiovascular medica-tions, or diuretics during the year com-pared with 77 to 79 percent of white andblack persons. Having Medicaid drug cov-erage improved overall access to heart dis-ease drugs for all groups, but the averagenumber of medications filled was stillmuch higher for white persons comparedwith the other beneficiaries. Employer-sponsored coverage consistently favoredwhite persons except in overall access toheart drugs, which was relatively similar

(ranging from 81 to 85 percent). With M+Ccoverage, disparity between racial and eth-nic groups appeared to nearly disappear,across measures of average use, averageunit price, out-of-pocket spending, andaccess to any heart drugs.

The M+C benefit observed previouslyfor minority beneficiaries is mixed in ourlast example, those with hypertensionuncomplicated by diabetes or heart dis-ease (Table 5). Here too, white beneficia-ries without drug coverage used moremedications, spent more on their drugtherapies, and had more access to hyper-tensive agents than the minority groups.Medicaid coverage still tends to favorwhite beneficiaries in terms of higher aver-age drug use and better overall access tomedications related to the disease, although


Table 5

Prescription Use and Expenditures for Medicare Beneficiaries with Hypertension1, byRace/Ethnicity:1999


Total (in Millions) 5.83 0.66 0.54No CoverageAverage Annual Drug Use 16.8 14.4 12.0Average Unit Price $39.30 $33.20 *$31.4Average Out-of-Pocket Cost $659.00 *$486.30 $526.10 Percent Out-of-Pocket to Total Cost 98.2 99.5 100.0Percent Prevalence of Any Hypertensive Drug Use2 75.1 **71.6 ** 58.8MedicaidAverage Annual Drug Use 31.2 **21.6 **21.6Average Unit Price $31.40 $34.20 **$39.6Average Out-of-Pocket Cost $190.10 $110.30 **$80.4Percent Out-of-Pocket to Total Cost 22.3 18.1 *13.0 Percent Prevalence of Any Hypertensive Drug Use2 78.8 **67.2 **73.8Employer Sponsored Average Annual Drug Use 21.6 14.4 25.2Average Unit Price $55.40 $47.10 $46.80 Average Out-of-Pocket Cost $326.80 $261.00 $457.40 Percent Out-of-Pocket to Total Cost 33.6 32.7 36.4Percent Prevalence of Any Hypertensive Drug Use2 76.6 ** 80.8 **75.3 M+CAverage Annual Drug Use 21.6 *16.8 19.2Average Unit Price $32.70 $30.70 $35.90 Average Out-of-Pocket Cost $294.00 $201.50 $222.30 Percent Out-of-Pocket to Total Cost 51.2 47.3 *42.6Percent Prevalence of Any Hypertensive Drug Use2 69.7 **73.7 **58.4

**p<0.05 black non-Hispanics compared with white non-Hispanics and Hispanics compared with white non-Hispanics.

*p<0.10 black non-Hispanics compared with white non-Hispanics and Hispanics compared with white non-Hispanics.1 Includes individuals reporting they have been told that they have hypertension.2 Includes hormones such as insulin and oral hypoglycemic drugs.

NOTES: n=7.0 million. Unweighted n=1762 white; n=235 black; n=160 Hispanic.


Hispanic and black persons had lower out-of-pocket expenses. Employer-sponsoredcoverage improved the number of filledprescription for both white and Hispanicpersons, but not for black persons. However,black persons with retiree drug benefitspaid the least out-of-pocket and had thebest access to hypertensive drugs. WithM+C drug coverage, the measures of med-ication use fluctuated across the groupsand show no discernable pattern. Whitepersons with hypertension had the highestaverage mediation use, Hispanics took themost expensive drugs, but black personsmanaged the best overall access to antihy-pertensives.

In Table 6, we tested whether the differ-ences in medication use, spending andaccess previously observed were due tounderlying population dissimilarities indemographics, economic status, or healthstatus. After adjustment for these factors,black and Hispanic persons still generallyfilled fewer medications and spent less onthem than white persons, although somedifferences narrowed with certain types ofdrug coverage and diseases. Not havingdrug coverage is particularly problematicfor black persons with diabetes who werefar less likely than white persons to fill anydiabetic agents (OR: 0.39). On average,black beneficiaries overall medication usewas 63 percent lower (p<0.10) and drugspending was 69 percent less (p<0.05). Incontrast, uninsured black beneficiarieswith heart disease or hypertension experi-enced about the same levels of drug useand spending as white beneficiaries, andaccess to condition-specific medicationsappeared slightly better, although not sta-tistically significant. Medicaid coverageexerted surprisingly negative impacts forchronically ill black beneficiaries. BlackMedicaid recipients with diabetes usedsubstantially fewer medications than whiterecipients and had far lower access to

insulins or oral hypoglycemics. Neither didMedicaid improve medication use for blackpersons with heart disease whose patternsare comparable to those without any druginsurance, while access substantively wors-ened for those with hypertension. HavingM+C or employer-sponsored drug cover-age closed many gaps in medication useand out-of-pocket spending between blackand white persons, across all three condi-tions. Neither insurance type could improvethe compromised access that black per-sons had to diabetic agents, although bothdid increase the use of heart medicationsand hypertensive agents.

The impact of drug insurance on themedication use of Hispanics is also sensitiveto condition type, although in ways that areunique from that of black individuals. Nothaving drug coverage decreases overallmedication use, spending, and access forHispanics relative to white persons,although particularly for those with hyper-tension or heart disease. Medicaid coveragedoes not appear to lessen the disparities,except for out-of-pocket drug costs: duallyeligible Hispanic beneficiaries pay two tothree times less than white beneficiaries.Under employer sponsored and M+C, gapsin medication use and expenditure haveclosed between Hispanic and white personswith hypertension. However, differencespersist in those with diabetes and heart dis-ease. Among the diabetics, Hispanics withemployer-sponsored coverage had lowermedication use compared with white per-sons. Similarly among the diabetics,Hispanics with M+C had lower access todrugs specific to diabetes compared withwhite persons. Among those with heart dis-ease, out-of-pocket costs for Hispanics withemployer sponsored and M+C were one-third lower than white persons. Total spend-ing was significantly lower in Hispanics withemployer-sponsored coverage comparedwith white persons.



Tab

le 6

Ad

just

ed P

resc

rip

tio

n U

se a

nd

Ou

t-o

f-P

ock

et S

pen

din

g,b

y D

rug

Cov

erag

e an

d R

ace/

Eth

nic

ity:

1999

Ben

efic

iarie

sB

lack

His

pani

cE

mpl

oyer

Em

ploy

er

Chr

onic

Con

ditio

n an

d D

rug

Use

No

Cov

erag

eM

edic

aid

Spo

nsor

edM

+C

No

Cov

erag

eM

edic

aid

Spo

nsor

edM

+C

Dia

bet

esP

erce

nt D

iffer

ence

in U

se*-

38

**-3

1-1

1-1

2-2

0**

-37

**-3

76

Per

cent

Diff

eren

ce in

Tot

al C

ost

** -

59-3

1-1

4-2

-33

**-4

1-1

925

Per

cent

Diff

eren

ce in

Out

-of-

Poc

ket

Cos

t**

-66

33-3

-21

-30

**-2

50-2

5-1

3P

roba

bilit

y of

Fill

ing

Any

Ant

i-Dia

betic

Dru

g1**

OR

:0.3

9**

OR

:0.3

6**

OR

:0.4

6**

OR

:0.4

0O

R:0

.57

OR

:0.9

2O

R:0

.41

**O

R:0

.36

Hea

rt D

isea

seP

erce

nt D

iffer

ence

in U

se-1

3-1

4-9

-7**

41-1

7-1

7-1

5P

erce

nt D

iffer

ence

in T

otal

Cos

t-2

0-2

2*-

25-8

*-40

-16

**-4

1-1

2P

erce

nt D

iffer

ence

in O

ut-o

f-P

ocke

t C

ost

*-29

19-1

9-1

5*-

37**

-277

**-3

0**

-32

Pro

babi

lity

of F

illin

g A

ny H

eart

Dis

ease

Dru

g2O

R:1

.41

OR

:1.4

9O

R:1

.03

**O

R:2

.92

OR

:0.6

1O

R:1

.39

OR

:1.0

3O

R:1

.02

Hyp

erte

nsi

on

Per

cent

Diff

eren

ce in

Use

-1**

-52

8-8

**-4

1*-

2714

-5P

erce

nt D

iffer

ence

in T

otal

Cos

t-1

1**

-51

-11

-10

**-5

9 5

78

Per

cent

Diff

eren

ce in

Out

-of-

Poc

ket

Rx

Cos

t-2

-59

-7-1

9**

-50

**-3

0314

-21

Pro

babi

lity

of F

illin

g A

ny H

yper

tens

ive

Dru

g2O

R:1

.20

OR

:0.6

6O

R:1

.72

OR

:2.3

3O

R:0

.88

OR

:0.8

1O

R:1

.39

OR

:0.9

1

*p<

0.10

bl

ack

non-

His

pani

cs c

ompa

red

with

whi

te n

on-H

ispa

nics

and

His

pani

cs c

ompa

red

with

whi

te n

on-H

ispa

nics

.

**p<

0.05

bla

ck n

on-H

ispa

nics

com

pare

d w

ith w

hite

non

-His

pani

cs a

nd H

ispa

nics

com

pare

d w

ith w

hite

non

-His

pani

cs.

1 In

clud

es h

orm

ones

suc

h as

insu

lin a

nd o

ral h

ypog

lyce

mic

dru

gs.

2 In

clud

es c

ardi

ac d

rugs

, ca

rdio

vasc

ular

dru

gs,

and

diur

etic

s.

NO

TE

S: n

=22

.6 m

illio

n.A

djus

ted

for

age,

sex

, in

com

e, s

elf-

repo

rted

hea

lth s

tatu

s, a

nd n

umbe

r of

com

orbi

ditie

s.

SO

UR

CE

:Cen

ters

for

Med

icar

e &

Med

icai

d S

ervi

ces:

Dat

a fr

om t

he M

edic

are

Cur

rent

Ben

efic

iary

Sur

vey

Cos

t an

d U

se F

ile,

1999

.

DISCUSSION

The Medicare Program has dramaticallyimproved the health of older and disabledpersons by removing many financial barri-ers, yet we know that some populationshave fared better than others by most mea-sures of medical care use and outcomes.Differences in the care of minority Medicarebeneficiaries have been puzzling to under-stand as a problem of economic barriergiven the near-universal enrollment intothe program and uniformity of the benefit(Mayberry, Milli, and Ofili, 2000). Recentevidence in younger adult populations hasconfirmed the view that health insurancealone does not eliminate racial/ethnic dis-parities and in fact may play a rather mod-est role in ameliorating the differences.Investigations into the primary drivers ofunequal use of medical services havefound that insurance influences access, butmuch about racial/ethnic differencesremains unexplained (Weinick, Zuvekas,and Cohen, 2000; Zuvekas, 1999; Zuvekas,and Taliaferro, 2003). Nevertheless, drugcoverage status is far from uniform in theMedicare population and we know type ofinsurance strongly influences medicationuse (Stuart, Shea, and Briesacher, 2000).That black and Hispanic Medicare benefi-ciaries use fewer or less expensive medica-tions than white beneficiaries may still be aproblem grounded in socioeconomic caus-es with far-reaching consequences. Notbeing able to afford necessary medicationsmay explain, at least in part, why black andHispanic persons more often than whitepersons experience some of the worsteffects of chronic illnesses. One study thathas linked drug coverage to racial/ethnicdifferences in use of other medical careservices comes from an analysis of patientswho gained Medicare coverage throughthe ESRD program (Daumit et al., 1999).Medicare beneficiaries with ESRD entitle-

ment are among the few to receiveMedicare reimbursement for criticaloutpatient medications—erythropoietin foranemia and immunosuppressants. Daumitet al. (1999) found that a three-fold differ-ences in the use of clinical procedures bypatient ethnicity nearly disappeared follow-ing the acquisition of the special ESRDMedicare coverage. The researchers attrib-uted the decline largely to the Medicarebenefit and concluded that equity in caremay be attainable for all Medicare benefi-ciaries if coverage is truly comprehensive,including for necessary prescription drugs.

Our study also detected statistically sig-nificant and sometimes large differences inthe drug use and spending patterns ofchronically-ill Medicare beneficiaries byrace and ethnicity. These findings showedwide variation that persisted even amongindividuals with the same disease andsame type of prescription coverage. Aswith studies of other medical services, wefound that minority beneficiaries tend toget less of chronic medications comparedwith the majority of beneficiaries who arewhite. Drug coverage from M+C plans wasthe most successful in eliminating the dif-ferences although some remained, particu-larly use of any diabetic agents by blackbeneficiaries. What might explain theimproved equity in drug use associatedwith M+C drug coverage? Speculationsinclude the mandatory assignment of a pri-mary care doctor and disease managementprograms, although the research is mixed.Hargraves, Cunningham, and Hughes(2001) did not find that managed care poli-cies such as gatekeeper requirementsreduced racial/ethnic disparities in havinga usual source of care or visiting a physi-cian in the last year. Haas et al. (2002)found some improvements in preventivecare services for Hispanic persons in man-aged care plans relative to FFS enrollees,but not for black persons. Lastly, Schneider,


Zaslavsky, and Epstein (2002) noted moreequitable use of β-blockers by race withM+C enrollment, but only in plans with bet-ter overall quality standards.

This study has several limitations. First,we used broad therapeutic classes in ana-lyzing medication use. A more refined ther-apeutic classification may provide moreinformation on the underlying pattern dif-ferences observed here. Similarly, analyz-ing access to new therapies could provide abetter marker for studying issues on dis-parities because they are generally per-ceived to be superior compared with oldertherapies.

Second, our models explained less thanone-third of the variation in medication useand access, which means other critical fac-tors influence this behavior that were notconsidered here. Notably, we did not con-trol for differences in geography. We knowthat Medicare minorities more often live inurban areas (white persons 75 percent,black persons 81 percent, Hispanics 86percent) and in the south (white persons33 percent, black persons 53 percent,Hispanics 56 percent), while Hispanicsreside disproportionately in the west(white persons 19 percent, black persons 6percent, Hispanics 25 percent). As a relat-ed limitation, we categorized our sampleinto three racial/ethnic groups and thisclassification may not accurately capturevariation in culture, biology, or values(LaVeist, 1994). Lastly, we do not know ifwhite Medicare beneficiaries are overus-ing medications, especially expensivebranded products, and it may be thatbehavior which contributes to the largeracial/ethnic differences.

Notwithstanding these limitations, wehave shown that access to prescriptiondrugs is compromised for black and

Hispanic Medicare beneficiaries with con-sistent need for medicines. Having someform of drug coverage may help in easingout-of-pocket burden, especially for peoplewith few personal resources, but it alonewill not eliminate racial/ethnic disparities.Even Medicaid, the most generous of cov-erage, could not erase the different med-ication use patterns experienced by minori-ties relative to white persons. We can onlyspeculate about the other potential causesof disparity. A policy implication of our find-ings is that the future Medicare expansioninto universal drug coverage should pro-grammatically address closing the gaps inmedication use by race and ethnicity.Surely one starting point is to more closelystudy the features of M+C drug coveragethat appeared to create more equitableaccess to drug therapy. Having a usualsource of care (either physician or clinic)may be a promising first step as access toprimary care has been linked to less dis-parity in other medical services (Williams,Flocke, and Strange, 2001). Identifying themechanisms of managed care that promotemore equitable access is especially impor-tant as black and Hispanic persons enrollinto these types of plans at higher ratesthan do white persons (Cunningham andKohn, 2000) Lastly, our finding thatHispanic beneficiaries tended to use moreexpensive medications is puzzling andrequires more study, perhaps whetherpreferences for branded medications arerelated to ethnicity.

ACKNOWLEDGMENTS

The authors wish to acknowledgeMarsha Lillie Blanton and Tricia Neumanfor their assistance in developing theresearch objectives for this study.


REFERENCES

Bergmann, M.M., Byers, T., Freedman, D. S., andMokdad, A.: Validity of Self-Reported DiagnosesLeading to Hospitalization: A Comparison of Self-Reports with Hospital Records in a ProspectiveStudy of American Adults. American Journal ofEpidemiology 147(10):969-977, May 15, 1998.Cunningham, P.J., and Kohn, L.: Health PlanSwitching: Choice or Circumstance? Health Affairs19(3):158-164, May/June 2000.Daumit, G.L., Hermann, J.A., Coresh, J., and Powe,N.R.: Use of Cardiovascular Procedures AmongBlack Persons and White Persons: A 7-YearNationwide Study in Patients with Renal Disease.Annals of Internal Medicine 130(3):173-182,February 2, 1999.Espino, D.V., Lichtenstein, M.J., Hazuda, H.P., et al.:Correlates of Prescription and Over-the-CounterMedication Usage Among Older Mexican Americans:The Hispanic EPESE Study. Established Populationfor the Epidemiologic Study of the Elderly. Journalof the American Geriatric Society 46(10):1228-1234,October 1998.Fillenbaum, G.G., Hanlon, J.T., Corder, et al.:Prescription and Non-Prescription Drug UseAmong Black and White Community-ResidingElderly. American Journal of Public Health83(11):1577-1582, November 1993.Fillenbaum, G.G., Horner, R.D., Hanlon, J. T., et al.:Factors Predicting Change in Prescription and Non-Prescription Drug Use in a Community-ResidingBlack and White Elderly Population. Journal ofClinical Epidemiology 49(5):587-593, May 1996.Gornick, M.E.: The Association of Race/Socioeconomic Status and Use of MedicareServices. A Little-Known Failure in Access to Care.Annals of New York Academy of Sciences 896:497-500, 1999.Gornick, M.E.: Disparities in Medicare Services:Potential Causes, Plausible Explanations, andRecommendations. Health Care Financing Review21(4):23-43, Summer 2000.Gornick, M.E.: A Decade of Research onDisparities in Medicare Utilization: Lessons for theHealth and Health Care of Vulnerable Men.American Journal of Public Health 93(5):753-759,May 2003.Gornick, M.E., Eggers, P.W., Reilly, T.W., et al.:Effects of Race and Income on Mortality and Use ofServices Among Medicare Beneficiaries. NewEngland Journal of Medicine 335(11):791-799,September 12, 1996.

Gornick, M.E., Eggers, P.W., and Riley, G. F.:Understanding Disparities in the Use of MedicareServices. Yale Journal of Health Policy Law Ethics1:133-158, Spring 2001.Haas, J. S., Phillips, K.A., Sonneborn, D., et al.:Effect of Managed Care Insurance on the Use ofPreventive Care for Specific Ethnic Groups in theUnited States. Medical Care 40(9):743-751,September 2002.Hargraves, J.L., Cunningham, P.J., and Hughes,R.G.: Racial and Ethnic Differences in Access toMedical Care in Managed Care Plans. HealthServices Research 36(5):853-868, October 2001.LaVeist, T.A.: Beyond Dummy Variables andSample Selection: What Health ServicesResearchers Ought to Know About Race as aVariable. Health Services Research 29(1):1-16, April1994.Mayberry, R.M., Mili, F., and Ofili, E.: Racial andEthnic Differences in Access to Medical Care.Medical Care Research and Review 57 (Suppl 1):108-145, 2000.Murray, L.A.: Racial and Ethnic Differences AmongMedicare Beneficiaries. Health Care FinancingReview 21(4):117-127, Summer 2000.Nelson, K., Norris, K., and Mangione, C. M.:Disparities in the Diagnosis and PharmacologicTreatment of High Serum Cholesterol by Race andEthnicity: Data from the Third National Health andNutrition Examination Survey. Archives of InternalMedicine 162(8):929-935, April 22, 2002.Schneider, E.C., Zaslavsky, A.M., and Epstein,A.M.: Racial Disparities in the Quality of Care forEnrollees in Medicare Managed Care. Journal of theAmerican Medical Association 287(10):1288-1294,March 13, 2002.StataCorp LP: STATA® Software Version 7. CollegeStation, TX. Internet address: http://www.stata.com(Accessed 2004.)Stuart, B., Shea, D., and Briesacher, B.: PrescriptionDrug Costs for Medicare Beneficiaries: Coverage andHealth Status Matter. Commonwealth Fund IssueBrief Number 365, January 2000.Sung, J.F., Harris-Hooker, S., Alema-Mensah, E.,and Mayberry, R.: Is There a Difference inHypertensive Claim Rates Among MedicaidRecipients? Ethnicity and Disease 7(1):19-26, Winter1997.Svetkey, L.P., George, L.K., Tyroler, H. A., et al.:Effects of Gender and Ethnic Group on BloodPressure Control in the Elderly. American Journalof Hypertension 9(6):529-535, June 1996.


U.S. Department of Health and Human Services:The Initiative to Eliminate Racial and EthnicDisparities in Health. U.S. Department of Healthand Human Services. February 21, 1999.Weinick, R.M., Zuvekas, S.H., and Cohen, J.W.:Racial and Ethnic Differences in Access To and Useof Health Care Services, 1977 to 1996. Medical CareResearch and Review 57 (Suppl 1):36-54, 2000.White-Means, S.I.: Racial Patterns in DisabledElderly Persons’ Use of Medical Services. Journalof Gerontology, Series B, Psychological Sciences andSocial Sciences 55(2):S76-89, March 2000.Williams, R.L., Flocke, S.A., and Stange, K.C.: Raceand Preventive Services Delivery Among BlackPatients and White Patients Seen in Primary Care.Medical Care 39(11):1260-1307, November 2001.

Zambrana, R.E. and Carter-Pokras, O.: Health DataIssues for Hispanics: Implications for Public HealthResearch. Journal of the Health Care for the Poorand Underserved 12(1):20-34, February 2001.Zuvekas, S.H.: Health Insurance, Health Reform,and Outpatient Mental Health Treatment: WhoBenefits? Inquiry 36(2):127-146, Summer 1999.Zuvekas, S.H., and Taliaferro, G.S.: Pathways toAccess: Health Insurance, the Health Care DeliverySystem, and Racial/Ethnic Disparities, 1996-1999.Health Affairs 22(2):139-153, March/April 2003.

Reprint Requests: Becky Briesacher, Ph.D., University ofMaryland, School of Pharmacy, 515 W. Lombard Street, Room161, Baltimore, MD 21201. E-mail: [email protected]


Racial and Ethnic Disparities in Prescription Coverage and ......plans, or M+C. (Our models excluded...

Documents

Transcript of Racial and Ethnic Disparities in Prescription Coverage and ......plans, or M+C. (Our models excluded...