Racial and Ethnic Disparitiesin Diabetes Care and Impactof Vendor-Based DiseaseManagement ProgramsDiabetes Care 2016;39:743–749 | DOI: 10.2337/dc15-1323

OBJECTIVE

We examined the existence of disparities in receipt of appropriate diabetes careamong California’s fee-for-service Medicaid beneficiaries and the effectivenessof a telephonic-based disease management program delivered by a disease man-agement vendor on the reduction of racial/ethnic disparities in diabetes care.

RESEARCH DESIGN AND METHODS

We conducted an intervention-control cohort study to test the effectiveness of a3-year-long disease management program delivered to Medicaid fee-for-servicebeneficiaries aged 22 to 75 with a diagnosis of diabetes in Los Angeles and Alamedacounties. The outcome measures were the receipt of at least one hemoglobinA1c (HbA1c) test, LDL cholesterol test, and retinal examination each year. We usedgeneralized estimating equations models with logit link to analyze the claims datafor a cohort of beneficiaries in two intervention counties (n = 2,933) and eight controlcounties (n = 2,988) from September 2005 through August 2010.

RESULTS

Racial/ethnic disparities existed in the receipt of all three types of testing in theintervention counties before the program. African Americans (0.66; 95% CI 0.62–0.70) and Latinos (0.77; 95% CI 0.74–0.80) had lower rates of receipt for HbA1c

testing than whites (0.83; 95% CI 0.81–0.85) in the intervention counties. After theintervention, the disparity among African Americans and Latinos compared withwhites persisted in the intervention counties. For Asian Americans and PacificIslanders, the disparity in testing rates decreased. We did not find similar dispar-ities in the control counties.

CONCLUSIONS

This disease management program was not effective in reducing racial/ethnic dis-parities in diabetes care in the most racially/ethnically diverse counties in California.

Diabetes disproportionately affects racial/ethnic minority populations. Comparedwith white adults, the risk of having a diabetes diagnosis is 77% higher amongAfrican Americans, 66% higher among Latinos/Hispanics, and 18% higher amongAsian Americans (1). Despite the high prevalence of the condition, minorities ex-perience lower quality of care and greater barriers to self-management comparedwith white patients (2,3). Racial/ethnic minorities are less likely to receive recom-mended services for diabetes, such as annual hemoglobin A1c (HbA1c) testing,

1UCLA Center for Health Policy Research, LosAngeles, CA2Division of General Internal Medicine andHealth Services Research, Department of Medi-cine, University of California, Los Angeles, LosAngeles, CA3Truven Health Analytics, Washington, DC4St. Jude Medical, Los Angeles, CA5Department of Health Policy and Manage-ment, UCLA Fielding School of Public Health,Los Angeles, CA6Department of Health Services Administration,University of Maryland School of Public Health,College Park, MD

Corresponding author: Ying-YingMeng, yymeng@ucla.edu.

Received 18 June 2015 and accepted 18 February2016.

This article contains Supplementary Data onlineat http://care.diabetesjournals.org/lookup/suppl/doi:10.2337/dc15-1323/-/DC1.

© 2016 by the American Diabetes Association.Readersmayuse this article as longas thework isproperly cited, the use is educational and not forprofit, and the work is not altered.

Ying-Ying Meng,1 Allison Diamant,1,2

Jenna Jones,3 Wenjiao Lin,4 Xiao Chen,1

Shang-Hua Wu,1 Nadereh Pourat,1,5

Dylan Roby,1,6 and Gerald F. Kominski1,5

Diabetes Care Volume 39, May 2016 743

CLIN

CARE/ED

UCATIO

N/N

UTR

ITION/PSYC

HOSO

CIAL

annual LDL cholesterol (LDL-C) testing,and an annual retinal examination (4).Racial/ethnic disparities in health carecan be found everywhere in the U.S.health care delivery system, even in pub-lic insurance programs, including Medic-aid (5). Medicaid is the largest providerof health insurance for low-income andminority populations, with;60%of ben-eficiaries being racial/ethnic minorities(6). With the implementation of the Pa-tient Protection and Affordable Care Act(ACA), states were allowed to expandMedicaid to previously uninsured indi-viduals. After full implementation ofthe ACA, Medicaid now covers morethan 18% of all American adults (7). Fed-eral and state government agencies havedeveloped and implemented diseasemanagement programs (8) for MedicareandMedicaid beneficiaries to reduce thedouble-digit increases in health carecosts and to address system-wide healthcare quality issues (9).Two main types of disease manage-

ment programs, provider-based andthird party (vendor)–based, have beenadopted by private and public insurers(10). Almost all third party–based dis-ease management programs rely on apatient-focused telephonic interventionas a strategy for reaching large popula-tions with one or more chronic illnessesthat may be at risk for exacerbation(11,12). The Centers for Medicare &Medicaid Services (CMS) have tested avariety of disease management pro-grams, including provider-based, third-party–based, and hybrid models, forMedicare populations with chronicconditions, dating back to 1999 (10).However, there is little evidence to con-clusively state that disease managementprograms have decreased hospitaliza-tions and emergency department visits,improved prescription drug adherence,lowered costs (13), or alleviated racialand ethnic disparities among the partici-pants (14). Also, telephonic disease man-agement programs have been criticizedrecently for their inability to produce suf-ficient savings and improvement in healthoutcomes (15,16).The California Department of Health

Care Services (CDHCS) conducted a CMS-approved pilot disease management pro-gram for Medi-Cal (California Medicaidprogram) fee-for-service adult beneficia-ries with a diagnosis of diabetes and otherselected chronic conditions from 1

September 2007 to 31 August 2010 (17).A vendor (McKesson Health Solutions)was selected through a competitive pro-curement process to provide telephonicdisease management services to the tar-geted populations in Alameda and LosAngeles Counties (17). The disease man-agement program was designed to reg-ularly contact high-risk, actively engagedbeneficiaries to assess their health status,encourage receipt of appropriate screen-ings and care from their providers, and pro-vide coaching or follow-up to encourageadherence to their personalized diseasemanagement plans (please see details ofthe disease management program in theSupplementary Data). However, the over-all rate of active engagement was ;10%perprogramyear.Most of thosewhowereever contacted received only one call,with a median of three calls per personover their full duration of eligibility.

Our recent literature search suggestslimited information is available regard-ing disparities in diabetes care withinthe Medicaid fee-for-service populationand whether disease management pro-grams are effective in reducing or elim-inating such differences among thispopulation. This information is extremelyimportant because the chronically illMedicaid population comprises one-thirdof the nation’s Medicaid population andaccounts for an estimated 80% of totalMedicaid expenditures (18). To addressthe current gaps in the literature, thisstudy examined two important questionson the quality of diabetes care amongMedi-Cal fee-for-service populations inthe intervention counties:

1. Were there any racial/ethnic differ-ences in diabetes care among thesebeneficiaries?

2. Did a telephonic-based and patient-focused disease management pro-gram, delivered by a private vendor,improve diabetes care and decreaseracial/ethnic disparities in the re-ceipt of comprehensive diabetescare?

RESEARCH DESIGN AND METHODS

Study Population and Data SourceMedi-Cal claims data from the CDHCSbetween 1 September 2005 and 31 Au-gust 2010 were used. The study popula-tion includes fee-for-service Medi-Calbeneficiaries, ages 22–75, diagnosedwith diabetes residing in two intervention

counties (Alameda and Los Angeles) andeight control counties (Fresno, Riverside,Sacramento, San Bernardino, San Diego,San Francisco, San Joaquin, and SantaClara). The two intervention countiescomprise 30% of Californians and havemore racially/ethnically diverse popula-tions (71%) than the overall Californianpopulation (60%). The selection of thecontrol counties was based on a reason-able match of intervention counties usingcluster analysis on paid claims (costs), dis-ease rates, demographic characteristics(age, sex, ethnicity, and language), andservice use (number of hospitalizations,emergency department visits, and doc-tor visits) at the county level for the pur-pose of expenditure comparison. For thisstudy, we used the Healthcare Effective-ness Data and Information Set (HEDIS)definition to define the beneficiarieswith diabetes (type 1 and type 2), whichrequires that they have had at least twoor more outpatient visits with a docu-mented diabetes diagnosis, one or moreacute inpatient or emergency depart-ment visits with a documented diabetesdiagnosis, ormedication dispensedwith adocumented diabetes diagnosis duringthe measurement year and the prioryear. The beneficiarieswere also requiredto be continuously enrolled, defined ashaving no more than a 1-month gap inMedicaid coverage, in the years beforemeasurement and in the measurementyears.

The disease management program ranfor 3 years, from 1 September 2007 to 31August 2010. Among beneficiaries withdiabetes, those who were aged 45–65,female, and white were more likely tobe actively engaged in the program (19).The sample included a cohort of 5,921beneficiaries with diabetes (2,933 in twointervention counties and 2,988 fromeight control counties) who were contin-uously enrolled from September 2005 toAugust 2010.

Outcome MeasuresThis analysis focused on three outcomesdefined by HEDIS regarding receipt of 1)at least one HbA1c blood test within thepast year, 2) at least one LDL-C screeningduring the past year, and 3) one or moreretinal examinations during the measure-ment year or prior year. HEDIS measure-ment specifications change slightly overtime; thus, the testing rates were ad-justed to reflect HEDIS specifications

744 Racial and Ethnic Disparities in Diabetes Care Diabetes Care Volume 39, May 2016

relevant to each measurement year ofthe study (20).

Other VariablesThe main independent variable of interestwas race/ethnicity, includingwhites (refer-ence), Latinos, African Americans, AsianAmericans and Pacific Islanders (AAPIs),and others.We also included the groupingof intervention counties and control coun-ties. This analysis used 1 September 2005through 31 August 2007 as preinterven-tion years and September 2007 to August2010 as postintervention years. We alsoincluded three-way interaction (interven-tion vs. race/ethnicity vs. program years)and two-way interactions in the models.The presence of a comorbidity was de-

termined by whether the individual had adiagnosis of one or more of the followingconditions: chronic obstructive pulmo-nary disease, asthma, congestive heartfailure, or coronary artery disease. A log-transformed disease severity score wasdefined by the Chronic Illness and Disabil-ity Payment System (21), using the ICD-9-CM codes and National Drug Codes fromthe claim data. Each individual in ourstudy population was assigned a diseaseseverity score at baseline and each inter-vention year to reflect any changes in hisor her disease severity.

Statistical AnalysesUnadjusted rates of the outcome mea-sures (receipt of annual HbA1c testing,annual LDL-C testing, and an annual ret-inal examination according to HEDISmeasures) and distributions of popula-tion by characteristics (race/ethnicity,sex, English-speaking capacity, comor-bidity rates) in the intervention and con-trol counties were tested using x2 tests.Generalized estimating equations wereused to examine the trends of receiptsof the examinations over time by inter-vention and control counties and byracial/ethnic groups controlling forage, sex, language, comorbidity, and dis-ease severity score with logit links tomodel the binary outcomes adjustingfor the nature of panel data. Specifically,we tested the significance of three-wayinteraction and two-way interactions.Post hoc tests were conducted to inves-tigate the disparity in receipt of annualHbA1c testing, annual LDL-C testing, andan annual retinal examination for pre-and postintervention years. Difference-in-difference analysis on probabilityscale was conducted separately for

intervention counties and control coun-ties to explore whether the variations indiabetes care between whites and otherraces/ethnicities were significant in thepreintervention (2005–2006) and post-intervention (2009–2010) periods. Theanalyses were conducted using Stata13 software.

RESULTS

Descriptive characteristics for the totalsample population in the baseline year,by intervention and control groups, areprovided in Table 1. During the baselineyear, the unadjusted testing rates variedbetween the intervention and controlcounties. HbA1c testing rates duringthe baseline year were 79% in the inter-vention counties and 71% in the controlcounties. Almost 80% of beneficiaries inthe intervention counties had receivedLDL-C testing during the baseline yearcompared with only 72% of individualsin the control counties (P , 0.001). An-nual retinal screening rates were 92%vs. 89% in the intervention and controlcounties, respectively (P , 0.001).

We found that the three-way interac-tion was not significant, indicating therewas no intervention effect on racial/ethnic disparity by program years. Thetwo-way interaction of race/ethnicity ver-sus program years was also not signifi-cant, indicating the disparity persistedafter the intervention. However, the

two-way interaction between interven-tion and race/ethnicitywas significant, in-dicating that the existence of disparitydiffered between the intervention andcontrol groups. The results for receipt ofannual HbA1c testing, annual LDL-C test-ing, and annual retinal examination forpre- and postintervention periods by in-tervention and control counties and byracial/ethnic groups, controlling for age,sex, language, comorbidity, and diseaseseverity score, are presented in Figs. 1–3.

Receipt of HbA1c TestingThe adjusted rate of receipt for HbA1ctesting in the intervention and controlcounties before and after the interven-tion by race/ethnicity is shown in Fig. 1.The results indicate that before the in-tervention, racial/ethnic minorities,such as African Americans (0.66; 95%CI 0.62–0.70) and Latinos (0.77; 95% CI0.74–0.80), had lower rates for HbA1ctesting than whites (0.83; 95% CI 0.81–0.85) in the intervention counties,whereas AAPIs and other racial/ethnicgroups in the intervention countieshad comparable rates of HbA1c testingwith whites. Persons in the control coun-ties did not have different rates of annualHbA1c testing among racial/ethnicgroups, except that AAPIs (0.78; 95% CI0.75–0.81) had higher rates than whites(0.71; 95% CI 0.68–0.74).

Overall, there were no changes in thereceipt of HbA1c testing from pre- to

Table 1—Descriptive summary of sample population by intervention and controlgroup at baseline (2005–2006)

Total (N = 5,912)

P valueIntervention (%) Control (%)

n = 2,933 n = 2,988

Testing ratesHbA1c 78.66 70.78 0.00LDL-C testing 80.33 71.92 0.00Retinal examinations 92.36 89.02 0.00

Race 0.00White 45.38 26.44Latino 18.99 20.25African American 15.85 13.89AAPI 12.41 23.93Others 7.36 15.5

Female sex (vs. male) 64.13 65.16 0.40

English, yes (vs. no) 25.3 37.62 0.00

Age-group (years) 19.33 29.52 0.00#50 43.03 42.651–59 37.64 27.88$60

Comorbidity, yes (vs. no) 57.82 48.39 0.00

care.diabetesjournals.org Meng and Associates 745

postprogram periods in the interventioncounties after adjusting for the baselinedifferences between control and inter-vention counties. The difference-in-difference results from the two-wayinteraction indicate that after the inter-vention, the probability of undergoingat least an annual HbA1c test in the inter-vention counties was still different for Af-rican Americans (0.69; 95% CI 0.65–0.73)and Latinos (0.75; 95% CI 0.71–0.79) thanfor whites (0.84; 95% CI 0.82–0.86),whereas the differences in testing ratesbetween AAPIs and others versus whitesalso remained nonsignificant through theend of the program year (Fig. 1). Further-more, differences in testing rates in thecontrol counties were not observed be-tweenwhites, African Americans, Latinos,and others, but higher rates were persis-tent among AAPIs.

Receipt of LDL-C TestingThe adjusted rate of annual LDL-C testingfor the pre- and postintervention years isshown in Fig. 2. Before the intervention,results of the regression model indicatedthat compared with whites (0.87; 95% CI0.85–0.89), Latinos (0.75; 95% CI 0.71–0.78), African Americans (0.65; 95% CI0.61–0.69), AAPIs (0.79; 95% CI 0.76–0.83), and others (0.78; 95% CI 0.72–0.84) had lower rates of having hadLDL-C testing in the intervention counties.However, in the control counties, therewere no differences in rates of LDL-C test-ing among racial/ethnic groups in theprior intervention years.After the intervention, the difference-

in-difference results indicate the disparities

in LDL-C testing rates remained betweenwhites, Latinos, African Americans, andothers in the intervention counties. How-ever, thedifferences in LDL-C testing ratesbetween whites and AAPIs were elimi-nated in the intervention counties. In ad-dition, we saw a similar pattern in thecontrol counties after the interventionyears, with no changes in LDL-C testingrates between Latinos, African Ameri-cans, and other versus whites (Fig. 2),but the testing rates among AAPIs werehigher than for whites.

Receipt of Annual Retinal ExaminationResults of the regression model (Fig. 3),adjusting for covariates of interest,demonstrated that African Americanshad lower rates (0.84; 95% CI 0.81–0.87) than whites (0.94; 95% CI 0.93–0.95) of annual retinal examinations inthe baseline year; however, all of theother racial/ethnic groups had compara-ble rates with whites in the interventioncounties. The beneficiaries in the con-trol counties had comparable rates ofreceiving an annual retinal examinationamong racial/ethnic groups, except forAAPIs, who had higher rates of testing(0.95; 95% CI 0.94–0.96).

After the intervention, the difference-in-difference estimate reflects a decreasein retinal examination rates, especiallyamong African Americans (0.78; 95% CI0.74–0.82), from pre- to postprogramyears in the intervention counties (Fig.3). As a result, the disparity in retinal ex-aminations remained between whites(0.90; 95% CI 0.88–0.92) and AfricanAmericans. Again, the beneficiaries in

the control counties had lower rates ofreceiving an annual retinal examinationthan the preintervention years, but com-parable patterns among racial/ethnicgroups remained.

CONCLUSIONS

Our findings indicate that racial/ethnicdisparities existed in the receipt of ap-propriate diabetes care in the interven-tion counties, most pronounced forAfrican Americans, but also for Latinosandother racial/ethnic groups, before theimplementation of a disease manage-ment pilot program among California’sfee-for-service Medicaid beneficiaries.Additionally, we found that the vendor-based disease management program inCalifornia did not improve diabetescare or reduce racial/ethnic disparitiesin care among these beneficiaries. Theresults show that the disparities in allthree diabetes care indicators remainedat the end of the intervention period forAfrican Americans and Latinos in the in-tervention counties compared withwhites. However, the testing rates forAAPIs were higher or the same in com-parison with whites in both interventionand control counties.

Our study is one of a very limited num-ber of studies that shows racial/ethnicdifferences regarding receipt of appro-priate care for Medicaid beneficiaries(22). The observed disparities in diabe-tes care among Medicaid fee-for-servicepopulations are consistent with the find-ings in the overall population (23,24)and in the Medicare population (25–27).

Figure 1—Predicted HbA1c testing rates by race/ethnicity, intervention, and preintervention and postintervention. Error bars represent the 95% CI.

746 Racial and Ethnic Disparities in Diabetes Care Diabetes Care Volume 39, May 2016

These findings are important becauseracial/ethnic minorities, specifically La-tinos and African Americans, are dispro-portionately more likely than whites tobe enrolled in Medicaid. These findingsalso have important clinical implicationsbecause racial/ethnic disparity in receiv-ing these critical clinical services could beassociated with minorities having muchhigher rates of diabetes-related complica-tions and death, including heart disease,blindness, end-stage kidney disease, pe-ripheral neuropathy, and nontraumaticamputation (28,29).Our findings suggest that the vendor-

based disease management programwasnot effective in reducing racial/ethnic

disparities in diabetes care for the Med-icaid fee-for-service population. Diseasemanagement programs exist in a varietyof settings, with a focus on both chronicdisease management and preven-tive care (13). The perceived benefitsof disease management programs aretheir emphases on patient involvementthrough education and self-activationwith the goal of improving receipt ofnecessary and appropriate care, whichshould result in improved health out-comes and cost savings (12). Althoughsome disease management programsthat include counseling, informationfeedback, education, and other patientsupport mechanisms are found to be

positively correlated with improvedhealth outcomes (9), most of the pro-grams have mixed findings with respectto improving quality of care, reducingdisparities in care, and controlling costs(14, 30, 31). The frequency of contacts ofdisease management programs may bean important factor. One study sug-gested that moderate or high frequencyof contact led to an improvement com-pared with low frequency of contact(32). Low frequency of contact by thedisease management vendor in thisstudy may explain why this particular pi-lot program was not effective in improv-ing the appropriate care and reducingdisparities in care. Although the disease

Figure 2—Predicted LDL-C testing rates by race/ethnicity, intervention, and preintervention and postintervention. Error bars represent the 95% CI.

Figure 3—Predicted retinal examination rates by race/ethnicity, intervention, and preintervention and postintervention. Error bars represent the95% CI.

care.diabetesjournals.org Meng and Associates 747

management vendor attempted to de-liver proactive telephonic interventionsto all of the actively engagedmembers, asignificant proportion of the “engaged”members received little or no interven-tion. Furthermore, the intensity of theintervention among the actively enrolledpopulation was low, with ;2.7 to 3.4monitoring calls per person during the3-year intervention.One of the issues that affected this

program implementation was missingor incorrect contact information for eli-gible beneficiaries because the vendordelivered the intervention almost exclu-sively through mailings and telephonecalls. Of the 54,051 individuals whowere ever eligible for the disease man-agement program during the 3-year pro-gram period, the vendor reportedthat;25% had incorrect or missing con-tact information for some periods oftheir eligibility, although correct contactinformation was eventually found formany of these individuals.Another possible explanation is lan-

guage barriers. For instance, no informa-tion is available documenting whetherthe interventions were conducted in lan-guages other than English, given that75% of the participants in this study pop-ulation reported not speaking English.Although telephonic translation wasavailable between English-speakingnurses and beneficiaries with limited En-glish proficiency, such as those speakingSpanish and Armenian, this option maynot have fully resolved language barriersto participation.Our findings highlight the complex

nature of disparities in health care, es-pecially for Medicaid populations (12).Medicaid beneficiaries could face signif-icant barriers due to their language, lit-eracy, culture, disability, mental illness,poverty, and abilities to find a primarycare doctor (12,33). Disease manage-ment programs that involve providers,are incorporated at all levels of care,and are tailored to the cultural needs ofvarious racial/ethnic groups might bemore effective (14,34). For instance, dis-ease management programs that includehealth literacy and education outreachare associated with enhanced self-efficacyand self-care behaviors (12,35).Several limitations related to the data

may prevent us from fully explaining theeffectiveness of the diseasemanagementprogram. First, limited information is

available about the content and intensityof the disease management intervention,especially specific efforts made to reduceracial/ethnic disparities in care. Thoughthe intervention information was avail-able at the individual level, only 10% ofthe eligible population was actively en-gaged, and most of them received onlyone call. It is difficult to determinewhether there was a possible “dose re-sponse” of the intervention as well asusing a categorical method to determinethe effect of receiving any calls versus nocall. Second, no information is availableon the clinical status of participants withrespect toHbA1c and LDL-C levels becauseno laboratory values are available fromthe claims data. Third, information de-scribing the study population was limitedto claims data only. We lacked informa-tion on physician practices for these pa-tients and the vendor’s effect on theproviders’ practices, although the vendorwas supposed to contact the providers ofthe eligible population to coordinate caredelivery. Fourth, the observed disparitiesin the intervention counties do not seemto be generalizable to the control coun-ties. Because no information was avail-able regarding the availability of otherdisease management programs and theadequacy of providers accepting Medic-aid patients in the intervention or controlcounties, it would be difficult to fully ex-plain the differences between interven-tion and control counties. So, the controlcounties mainly serve as a comparisongroup for contextual information, whichallows us to ascertain whether anychanges in diabetes care are attribut-able to disease management programinterventions and not to other seculartrends.

Although this and other studies havedemonstrated a range of efficacy fordisease management programs, the ef-fectiveness of vendor-based diseasemanagement programs in reducingracial/ethnic disparities in diabetescare for Medicaid fee-for-service popula-tion remains questionable. Public and pri-vate efforts to improve self-managementskills and care-seeking behaviors of pa-tients with diabetes should carefullyexamine whether vendor-based andpatient-focused disease managementprograms can improve quality of care.As a result, the implementation of dis-ease management programs shouldinclude a prospective evaluation such

as this one, which can provide ongoingfeedback to the program designersabout the success or failure of the keycomponents of the program during theimplementation period. Because Califor-nia, especially the two interventioncounties, has one of the most diverseMedicaid populations in the country,we believe our results are applicable toMedicaid populations in other parts ofthe country with diverse populations.

Funding. This evaluation was funded by theCDHCS (contract number 06-55552).Duality of Interest. No potential conflicts ofinterest relevant to this article were reported.AuthorContributions. Y.-Y.M. and J.J. concep-tualized the research question and study designand wrote the manuscript. A.D. contributed towriting themanuscript and reviewed and editedthemanuscript. J.J.,W.L., andS.-H.W.conducteddata analyses under the directionsof Y.-Y.M. andX.C. N.P., D.R., and G.F.K. reviewed and editedthemanuscript. X.C. is the guarantor of thisworkand, as such, had full access to all the data in thestudy and takes responsibility for the integrity ofthe data and the accuracy of the data analysis.

References1. Centers for Disease Control and Prevention.2011 National Diabetes Fact Sheet. Atlanta,Centers for Disease Control and Prevention, Na-tional Center for Chronic Disease Preventionand Health Promotion Division of DiabetesTranslation, 20112. Boyle JP, Thompson TJ, Gregg EW, Barker LE,Williamson DF. Projection of the year 2050 bur-den of diabetes in the US adult population: dy-namic modeling of incidence, mortality, andprediabetes prevalence. Popul Health Metr2010;8:293. Mah CA, Soumerai SB, Adams AS, Ross-Degnan D. Racial differences in impact of cover-age on diabetes self-monitoring in a healthmaintenance organization. Med Care 2006;44:392–3974. National Committee for Quality Assurance.What is the current state of quality of care in di-abetes? [article online], 2011. Available fromhttp://www.ncqa.org/PublicationsProducts/OtherProducts/QualityProfiles/FocusonDiabetes/WhatistheCurrentStateofQualityofCare.aspx.Accessed 18 June 20155. Institute of Medicine. Unequal Treatment:Confronting Racial and Ethnic Disparities inHealth Care. Washington, DC, The NationalAcademies Press, 20036. Centers for Medicare & Medicaid Services(CMS). 2012 CMS Statistics [article online], 2012.Available from https://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/CMS-Statistics-Reference-Booklet/Downloads/CMS_Stats_2012.pdf. Accessed 18June 20157. CohenRA,MartinezME.Health Insurance Cov-erage: Early Release of Estimates From the Na-tional Health Interview Survey, January–March2015. Atlanta, Centers for Disease Control and

748 Racial and Ethnic Disparities in Diabetes Care Diabetes Care Volume 39, May 2016

Prevention, National Center for Health Statis-tics, 20158. Chrysochou C, Kalra PA. Current manage-ment of atherosclerotic renovascular disease–what have we learned from ASTRAL? NephronClin Pract 2010;115:c73–c819. Institute of Medicine. Crossing the QualityChasm: A New Health System for the 21st Cen-tury. Washington, DC, The National AcademiesPress, 200110. Bott DM, Kapp MC, Johnson LB, Magno LM.Disease management for chronically ill beneficia-ries in traditionalMedicare. Health Aff (Millwood)2009;28:86–9811. Mays GP, Au M, Claxton G. Convergence anddissonance: evolution inprivate-sector approachesto disease management and care coordination.Health Aff (Millwood) 2007;26:1683–169112. Roby DH, Kominski GF, Pourat N. Assessingthe barriers to engaging challenging popula-tions in disease management programs: theMedicaid experience. Dis Manag Health Out-comes 2008;16:421–42813. Mattke S, SeidM,MaS. Evidence for theeffectof disease management: is $1 billion a year a goodinvestment? Am J Manag Care 2007;13:670–67614. Kominski GF, Morisky DE, Afifi AA,Kotlerman JB. The effect of disease manage-ment on utilization of services by race/ethnicity:evidence from the Florida Medicaid program.Am J Manag Care 2008;14:168–17215. Motheral BR. Telephone-based diseasemanagement: why it does not save money.Am J Manag Care 2011;17:e10–e1616. Vollmer WM, Kirshner M, Peters D, et al.Use and impact of an automated telephone out-reach system for asthma in a managed care set-ting. Am J Manag Care 2006;12:725–733

17. National Health Law Program. ChronicDisease Management in the Medi-Cal Pro-gram. Health Consumer Alliance, 2008. Avail-able from http://www.healthconsumer.org/cs058ChronicDiseaseManagement.pdf. Accessed2 March 201618. Kaiser Commission on Medicaid and theUninsured. Medicaid and Managed Care: KeyData, Trends, and Issues. Washington, DC, 201219. Kominski GF, Pourat N, Roby DH, et al. Dis-ease Management Pilot Program in California:Evaluation Report. Los Angeles, UCLA Center forHealth Policy Research, 201220. National Committee for Quality Assurance.HEDIS & Performance Measurement 2005, 2006,2007, 2008, 2009, 2010,Washington, DC,NationalCommittee for Quality Assurance21. Kronick R, Gilmer T, Dreyfus T, Lee L. Improv-ing health-based payment for Medicaid beneficia-ries: CDPS. Health Care Financ Rev 2000;21:29–6422. DuBard CA, Yow A, Bostrom S, Attiah E,Griffith B, Lawrence W. Racial/ethnic differ-ences in quality of care for North Carolina Med-icaid recipients. N C Med J 2009;70:96–10123. Agency for Healthcare Research and Qual-ity. National Healthcare Disparities Report,2011. Rockville, MD, Agency for Healthcare Re-search and Quality, 201224. Mainous AG 3rd, Diaz VA, Koopman RJ,Everett CJ. Quality of care for Hispanic adultswith diabetes. Fam Med 2007;39:351–35625. Trivedi AN, Zaslavsky AM, Schneider EC,Ayanian JZ. Relationship between quality ofcare and racial disparities in Medicare healthplans. JAMA 2006;296:1998–200426. Chou AF, Wong L, Weisman CS, et al. Gen-der disparities in cardiovascular disease careamong commercial and Medicare managed

care plans. Womens Health Issues 2007;17:139–14927. Schneider EC, Zaslavsky AM, Epstein AM.Racial disparities in the quality of care for en-rollees in Medicare managed care. JAMA 2002;287:1288–129428. Carter JS, Pugh JA, Monterrosa A. Non-insulin-dependent diabetes mellitus in minori-ties in the United States. Ann Intern Med 1996;125:221–23229. Agency for Healthcare Research and Qual-ity. Diabetes Disparities Among Racial and Eth-nic Minorities. Rockville, MD, Agency forHealthcare Research and Quality, 200130. Chou AF, Brown AF, Jensen RE, Shih S,Pawlson G, Scholle SH. Gender and racial dispar-ities in the management of diabetes mellitusamongMedicare patients. Womens Health Issues2007;17:150–16131. Sequist TD, Adams A, Zhang F, Ross-DegnanD, Ayanian JZ. Effect of quality improvement onracial disparities in diabetes care. Arch InternMed 2006;166:675–68132. Pimouguet C, Le Goff M, Thiebaut R,Dartigues JF, Helmer C. Effectiveness of disease-management programs for improving diabetescare: a meta-analysis. CMAJ 2011;183:E115–E12733. White C, Fisher C, Mendelson D, SchulmanKA. State Medicaid Disease Management: Les-sons Learned from Florida. Durham, NC, TheHealth Strategies Consultancy LLC and DukeUniversity, 200534. Chin MH, Goldmann D. Meaningful dispar-ities reduction through research and translationprograms. JAMA 2011;305:404–40535. Cavanaugh KL. Health literacy in diabetescare: explanation, evidence and equipment. Di-abetes Manag (London) 2011;1:191–199

care.diabetesjournals.org Meng and Associates 749