The prevalence of diabetes has grown at double enrollees with diabetes were also Recent studies have demonstrated a strong link between diabetes …
I S S U E P A P E R
k aios e ri s s i o n c mm
medicaid
and the
on
uninsured
October 2007
An Overview of Medicaid Enrollees with Diabetes in 2003
Mindy Cohen The Urban Institute
Executive Summary
Diabetes was the 6th leading cause of death in the US in 2004 and is among the top 10 most expensive medical conditions in the country The prevalence of diabetes has grown at double digit rates over the last 40 years and in some parts of the country is now considered an epidemic1 Although more than one in seven diabetics in America rely on the Medicaid program for their health coverage, little is known about who they are or what the program spends on their behalf This brief provides a first look at this issue using diagnoses codes reported by the states to the federal government through the Medicaid Statistical Information System MSIS It highlights spending and enrollment patterns of the nearly 2 million Medicaid enrollees with diabetes in 2003 using administrative data from the MSIS for Federal Fiscal Year 2003 Key Findings: Roughly 19 million Medicaid enrollees had diagnosed diabetes in FY 2003, about 6 of the Medicaid population The majority of these enrollees were elderly or
disabled Elderly enrollees had the highest prevalence rate at 22, while children had the lowest, 03 The 6 of Medicaid enrollees with diabetes accounted for 16 of total Medicaid spending On average, enrollees with diabetes spent 16,967 per capita Across all eligibility groups, enrollees with diabetes had higher spending than enrollees without diabetes Figure 1 Enrollees with diabetes used significantly more acute care services than enrollees without diabetes This was particularly apparent with inpatient care and prescription drugs, as would be expected for individuals with chronic disease Long term care use was not affected by diabetes status, with enrollees with and without diabetes incurring similar long term care spending
1330 G STREET NW, WASHINGTON, DC 20005 PHONE: 202-347-5270, FAX: 202-347-5274 W E B SI T E : W W W K F F O R G
Figure 1
Spending Per Medicaid Enrollee with and without Diabetes by Group, 2003
Enrollees with Diabetes Enrollees without Diabetes
18,843 15,342
18,982
14,613
6,481 3,367
6,366
2,047
Elderly
Disabled
Adults
Children
Source: Urban Institute estimates of MSIS 2003 data prepared for the Kaiser Commission on Medicaid and the Uninsured Totals
only include full-benefit enrollees
Just under one third of enrollees with diabetes were also diagnosed with mental illness in FY 2003 Recent studies have demonstrated a strong link between diabetes and mental illness nationally, which is mirrored in the Medicaid population Per capita spending for this group was two times higher than for diabetic enrollees without mental illness
Policy Implications Medicaid enrollees with diabetes are a high cost population with significant health complications and high levels of health care use Effective approaches to improve care management for Medicaid enrollees with diabetes hold great potential for improving health while reducing costs; however, more research is required to determine which approaches might be most effective given the characteristics and service use of the Medicaid population with diabetes Such research should be informed by claims level analyses of Medicaid enrollees with diabetes to help better coordinate their care Unfortunately, these analyses are not supported by current national Medicaid data sources However, state efforts to address Medicaid enrollees with diabetes and other chronic conditions using state level claims
data and targeted programs are growing more prevalent each year and hold great promise for the development of better approaches to caring for this population
00
Introduction
Diabetes was the 6th leading cause of death in the US in 2004 Almost 15 million people had diagnosed diabetes in 2005, up from 85 million in 1995, and another 6 million people are thought to live with undiagnosed diabetes today2 Roughly 13 million new cases are diagnosed every year in people over 20 and prevalence has increased dramatically over the last 40 years, often growing at double digit rates3 This increased prevalence has been the primary reason for increased spending on diabetes over the last few decades, and has made diabetes one of the top 10 most expensive medical conditions in the US4 Medicaid covered about 15 of all individuals with diagnosed diabetes in the country in FY 20035 These beneficiaries account for a substantial portion of Medicaid program costs even though they are a relatively small percentage of this population Much of these costs are due to the complications caused by diabetes, which are often the result of inadequate primary care and are therefore preventable6 While the
literature has shown mixed results on the effectiveness of disease management programs, it has been proven that diabetes complications can be reduced with comprehensive monitoring and primary care7 Concerned about rising health care costs in the Medicaid program, many states have shifted their focus to disease management programs with a specific emphasis on diabetes in FY 2006, 12 states implemented Medicaid disease management programs, and 26 more states adopted programs in FY 20078 Although Medicaid covered a substantial percentage of individuals with diabetes, little is known about this group This brief highlights the spending and enrollment patterns of Medicaid enrollees with diabetes in FY 2003 Previous research details spending amounts, prevalence rates, and health care utilization rates for all diabetics in the country, but this brief is the first to present spending and enrollment patterns specific to diabetics with Medicaid coverage We begin this brief with general information on diabetes and provide national estimates of prevalence and costs for diabetes in the US We then present our findings on diabetes in Medicaid with enrollment and spending by eligibility group and
managed care status We then provide information on total spending and spending for long term and acute care services, and compare service use of enrollees with diabetes to those without diabetes We also analyze enrollees with diabetes who also suffer from mental illness, as they account for a sizable percentage of diabetics and have very high costs
Background on Diabetes
General Overview Diabetes is a disorder in which the body does not produce enough insulin, a hormone that converts sugar into energy Untreated diabetes causes a build up of sugar in the bloodstream, and can lead to serious health problems Individuals can have either Type 1 diabetes or Type 2 diabetes Type 1, previously called insulin dependent, accounts for about 5 10 of all
diagnosed diabetes cases and usually affects children and young adults; methods of prevention are unknown Individuals with Type 1 must use insulin to control their diabetes Type 2 diabetes, previously called adult onset or non insulin dependent, accounts for all remaining diagnosed cases of diabetes, and is typically associated with factors such as older age, obesity, family history, physical inactivity, and race/ethnicity9 Type 2 diabetes
can be regulated through diet, exercise, and oral medications Diabetes causes a wide range of complications, and if poorly controlled will often result in cardiovascular disease, kidney failure, blindness, and a number of other co morbidities Diabetes is the number one cause of kidney failure, non traumatic leg amputations, and new cases of blindness among adults10 However, comprehensive primary care, which includes blood sugar monitoring, proper nutrition and physical activity, greatly reduces the risk of these complications Educating individuals with diabetes on how to control their disease is also a critical part of their care National Prevalence Diabetes affects all ages and ethnicities, although not equally Elderly people are the most likely group to have diabetes and to experience long term complications; 103 million individuals over 60 years old had diabetes in 2005, which was 209 of people in this age group Children have historically been the least likely group to develop diabetes, although with the incidence of childhood obesity on the rise diabetes is being diagnosed more frequently in children and is the most common chronic disease in kids and adolescents11 Most of the
individuals with diabetes both diagnosed and undiagnosed are non Hispanic whites, however they are the least likely group to develop diabetes only 87 of this group are diabetic, compared to 133 of non Hispanic blacks, 95 of Hispanics, and 151 of American Indians/Alaska natives Similar numbers of men and women had diabetes, 109 million 105 and 97 million 88, respectively National Costs As a result of the high risk of complications, diabetes is a very costly disease to treat Total costs associated with diabetes was 132 billion 2002; 92 million for direct medical costs, and 40 billion in indirect costs due to lost work time, disability, and premature mortality12 Average per person spending for someone with diabetes was 13,243 in 2002, compared to 2,560 for non diabetics13 People with diabetes are more than twice as likely to develop cardiovascular disease, the most common diabetes complication, which requires expensive hospital stays and often multiple hospital visits As mentioned previously, comprehensive primary care is critical to preventing costly complications One study showed that national inpatient hospital costs for diabetes with complications was about 38 billion in 2001, a
good portion of which was the result of inadequate primary care It is estimated that up to 25 billion of this spending could have been saved with appropriate preventive care, including 386 million in savings to Medicaid alone14
00
Data Sources and Methods
The data used in this analysis come from the FY 2003 Medicaid Statistical Information System MSIS Summary File maintained by the Centers for Medicare and Medicaid Services CMS The MSIS contains demographic, eligibility, and expenditure information for all Medicaid enrollees We designate all enrollees using the Basis of Eligibility in MSIS as elderly age 65 and older, disabled under age 65, or other adults or other children not classified as disabled Spending is aggregated into over 30 types of services, including inpatient care, drugs, personal care, and others Due to the aggregated nature of spending reported in the MSIS, we present total spending by service for individuals included in our study, but are unable to separate out spending for health care services specifically related to diabetes Expenditures reported in MSIS do not include disproportionate share payments to providers, payments to Medicare, or administrative
payments 15 Individuals in MSIS are classified by the level of benefits they receive enrollees are either entitled to the full scope of Medicaid benefits termed full benefit or only a subset of services, such as pregnancy related or substance abuse services termed restricted benefit As these restricted benefit enrollees often display notably different patterns of spending from full benefit enrollees, we excluded them from our analysis and only included full benefit enrollees in all data calculations Individuals had to be classified as full benefit enrollees for every month of FY 2003 or we considered them restricted benefit enrollees This excludes roughly 225 million people from our analysis, or 41 of total Medicaid enrollment in FY 2003 MSIS also contains information on monthly eligibility status In order to provide more accurate calculations of average spending per enrollee amounts for a full year, we annualized spending amounts for any individuals who were only enrolled for part of the year based on the number of months they were enrolled In this analysis, we include all enrollees in both fee for service Medicaid and Medicaid managed care plans, unless otherwise stated in the
tables This analysis focuses on individuals who are flagged as having diabetes during FY 2003 in Medicaid 16 This flag, created by CMS, is determined by the presence of certain ICD 9 codes related to diabetes on the individuals health care claims during the year17 Underreporting of diagnosis codes is a limitation of our study, which occurs primarily for two major reasons First, MSIS only captures individuals who received treatment for their diabetes, and will therefore exclude individuals with undiagnosed diabetes or diagnosed diabetes but no treatment in FY 2003 Second, although an individual may have received treatment for their diabetes during the timeline of our study, that information may not be captured due to claim reporting errors or lack of reporting requirements Some states had high missing rates of claims or diagnosis codes for certain services, predominantly long term care claims Other states have a maximum number of diagnosis codes allowed to be reported per fee for service claim Claims submitted by managed care plans typically exclude diagnosis information, which occurs in most states
Findings
Prevalence Roughly 19 million enrollees had diagnosed diabetes in 2003,
or 6 of the Medicaid population Table 1 This is higher than the national prevalence of diagnosed diabetics at 146 million or 49 of the entire US population18 The vast majority of Medicaid enrollees with diabetes were elderly or disabled, with non disabled adults and children accounting for about 16 The highest prevalence rate was among the elderly, 22, and non disabled children the lowest, 03
Table 1 Enrollment and Spending for Diabetics in Medicaid by Eligibility Group, 2003 Enrollment Spending
Eligibility Group Elderly Disabled Adults Children Total
Number of Enrollees 769,073 798,345 238,445 52,029 1,857,892
Percent of Eligibility Group with Diabetes 22 13 5 03 6
Total Spending 14,491,773,363 15,154,266,909 1,545,371,938 331,225,848 31,522,638,058
Share of Spending within Eligibility Group 25 17 9 1 16
Spending per Enrollee 18,843 18,982 6,481 6,366 16,967
Source: Urban Institute estimates based on MSIS 2003 data Data for Maryland are excluded
These trends are similar to national prevalence rates discussed above and are expected, as diabetes worsens with age and primarily affects older individuals Table 2 shows the subset of this population enrolled in Medicaid managed
care Total prevalence was slightly lower in this group 4 than all enrollees with diabetes 6, although this is expected as managed care claims often exclude diagnosis codes and will therefore underestimate the total number of managed care enrollees who have diabetes Of the roughly 835,000 managed care enrollees with diabetes, more than 40 were disabled and just under one third were elderly Elderly enrollees had the highest in group prevalence In addition, dual eligibles had much higher prevalence rates than non duals, 19 to 3, which is expected as duals are typically in much
Table 2 Distribution of Spending Comparisons Among Diabetics in Managed Care Plans Enrollees with Diabetes in Managed Care Percent of Eligibility Group Total Spending Per Enrollment with Diabetes Enrollee 250,267 18 15,090 365,891 11 16,393 181,835 5 5,636 37,413 0 5,242 835,406 4 13,162
Eligibility Group Elderly Disabled Non-Disabled Adults Non-Disabled Children Total
Source: Urban Institute estimates based on MSIS 2003 data Data for Maryland are excluded
00
poorer health than non duals data not shown19 Total Spending and Spending Per Enrollee The 6 of Medicaid enrollees with diabetes spent about 315
billion in FFY 2003 16 of total spending on Medicaid Table 1 and Figure 2 On average, all diabetics in Medicaid spent 16,967 per capita Spending per enrollee amounts were very similar for elderly and disabled Medicaid diabetics, 18,843 and 18,982, while non disabled adults and children spent about one third of that Elderly and disabled enrollees in managed care had notably lower average spending per enrollee than all elderly and disabled enrollees with diabetes 15,090 and 16,393 for those with managed care Table 2
Figure 2
Diabetic Enrollees as a Percentage of Medicaid Enrollment and Spending within Group, 2003
Enrollment
25 22 17 13 9 5 03 1 Elderly Disabled Adults Children Total 6
Spending
16
Source: Urban Institute estimates of MSIS 2003 data prepared for the Kaiser Commission on Medicaid and the Uninsured Totals only include full-benefit enrollees
Service Use Service use, as measured by spending per enrollee amounts, differed dramatically between diabetics and non diabetics Table 3, which evaluates spending per enrollee by service for the fee for service population only, shows that enrollees with diabetes spent much more on acute services than enrollees without diabetes,
across all eligibility groups20 Elderly diabetics spent almost three times more on inpatient services than elderly non diabetics, 1,620 compared to 566, and a greater percentage of all spending went to acute care services in general These differences are quite striking considering most elderly Medicaid enrollees are dual eligibles, and will therefore have much of their acute care services covered by Medicare as well Elderly enrollees with diabetes also spent notably more on prescription drugs, 3,136 to 1,969
Table 3 Distribution of Spending Comparisons Among Diabetics and Non-Diabetics in Medicaid by Eligibility Group - Fee-For-Service ONLY
Enrollees with Diabetes Share of Total Spending Per Spending Per Enrollee Enrollee Elderly Enrollment Long Term Care Inpatient Outpatient/Physician/Clinic Drugs Other Acute Total Disabled Enrollment Long Term Care Inpatient Outpatient/Physician/Clinic Drugs Other Acute Total Non-Disabled Adults Enrollment Long Term Care Inpatient Outpatient/Physician/Clinic Drugs Other Acute Total Non-Disabled Children Enrollment Long Term Care Inpatient Outpatient/Physician/Clinic Drugs Other Acute Total
Enrollees without Diabetes Share of Total Spending
Per Spending Per Enrollee Enrollee
518,806 13,628 1,620 927 3,136 1,376 20,687 66 8 4 15 7 100
1,656,206 14,296 78 566 3 495 3 1,969 11 928 5 18,254 100
432,454 6,897 5,023 3,141 4,181 1,986 21,227 32 24 15 20 9 100
2,317,571 9,853 2,111 1,919 2,331 1,827 18,040
55 12 11 13 10 100
56,610 301 3,994 2,604 1,641 679 9,218 3 43 28 18 7 100 79 1,535 1,432 540 340 3,926
974,291 2 39 36 14 9 100
14,616 1,883 2,602 1,773 1,264 1,742 9,264 20 28 19 14 19 100 225 496 620 232 618 2,191
3,708,953 10 23 28 11 28 100
Source: Urban Institute estimates based on MSIS 2003 data Data for Maryland are excluded Long Term Care includes nursing facilities, ICF/MR, inpatient mental health facilities, home health, personal care, and home and communitybased services Other Acute includes dental, hospice, targeted case management, labx, midwife, nurse practitioner, private duty nursing, other practitioners, sterilization, abortion, rehab, therapy, transportation, and other services
Disabled enrollees with diabetes also spent significantly more on inpatient services 5,023 to 2,111 and drugs 4,181 to 2,331 than disabled enrollees without diabetes Disabled enrollees with diabetes were much less
reliant on long term care services than disabled non diabetics, but still had higher average total spending, despite the high expense typically associated with long term care services
00
Non disabled adults and children with diabetes followed similar trends to elderly and disabled diabetics Non disabled adults had very high average spending per enrollee amounts for inpatient and outpatient/physician/clinic care, somewhat close to levels for disabled enrollees Forty three percent of spending for non disabled adults with diabetes was for inpatient care, with average spending of 3,994 Non disabled children with diabetes were quite expensive relative to children without diabetes, with average spending more than 4 times that of non diabetic children 9,264 and 2,191 respectively Children with diabetes spent over 5 times more on inpatient care and prescription drugs than children without diabetes Diabetics with Mental Illness Recent studies have shown there is a strong link between diabetes and mental health individuals with Type 2 diabetes are twice as likely to develop Alzheimers and depression, and the rate of diabetes in individuals with mental illness is twice as high as the
general population21 While the links between these two diseases are still being researched, this is a critical population to study as they suffer from two vastly different conditions that both require constant monitoring and comprehensive health care Among Medicaid enrollees with diabetes, just under one third were also diagnosed with mental illness Table 4 Not surprisingly, these individuals were significantly more expensive than enrollees with diabetes but without mental illness, spending on average 26,710 to 12,708 respectively While individuals with both diagnoses only accounted for a third of total enrollment of diabetics, their total spending of 15 billion accounted for almost half of spending on all diabetics in Medicaid Elderly enrollees with diabetes and mental illness were the most expensive group of those with both diagnoses, with average spending of 33,275 Disabled individuals with diabetes and mental health diagnoses had lower spending per enrollee amounts, 25,983, but were still costly Non disabled children had noticeably high spending, with average spending at 13,789 These children, while technically classified as non disabled for the purposes of Medicaid
eligibility, most likely suffer from multiple comorbidities and require significant acute care
Table 4 Spending and Enrollment for Diabetic Enrollees with a Mental Illness Diagnosis by Eligibility Group, 2003
Eligibility Group Elderly Disabled Adults Children Total
Enrollment 186,915 313,454 53,025 11,672 565,066
Percent of Diabetic Enrollees with Diabetes and Mental Illness 24 39 22 22 30
Total Spending 6,219,505,322 8,144,600,603 568,140,509 160,940,758 15,093,187,192
Spending Per Enrollee 33,275 25,983 10,715 13,789 26,710
Spending Per Enrollee for Diabetics WITHOUT Mental Illness 14,210 14,456 5,270 4,219 12,708
Source: Urban Institute estimates based on MSIS 2003 data Data for Maryland are excluded
Conclusions and Discussion
Medicaid enrollees with diabetes accounted for 6 of enrollment and 16 of spending nationally Among all groups, diabetics in Medicaid are significantly more expensive than their non diabetic counterparts Medicaid diabetics in this analysis rely heavily on acute care, especially inpatient care, which is expected with individuals with chronic disease The vast majority of diabetics in Medicaid are elderly or disabled, already an expensive population
that typically has multiple comorbidities, and they spend on average about 3 times more than adults or children with diabetes That diabetics are much more likely to be high cost enrollees than non diabetics has led state and federal policy makers to refocus on disease management programs While the literature has shown mixed results of the cost effectiveness of disease management programs, both the federal government and individual states increasingly view the monitoring and treatment of chronic disease as an important issue from both a quality and cost perspective While this brief provides valuable information on the Medicaid population with diabetes, it raises many questions It is unclear how many enrollees with diabetes are receiving an adequate level of primary care and for those that do receive it, how their utilization levels and health outcomes compare to other diabetics in Medicaid As states shift their focus to disease management and emphasize patient involvement in monitoring their conditions, it is important to better understand this population Further research using claims data to analyze health care utilization for diabetes related complications over a multi year period
could provide the additional detail necessary to shed light on the most appropriate way to care for this population
10
00
Notes
1
Kleinfeld NR Diabetes and Its Awful Toll Quietly Emerge as a Crisis New York Times, January 9, 2006 CDC National Diabetes Fact Sheet, United States 2005 Ford Earl, Ali Mokdad, Wayne Giles, Deborah Galuska, Mary Serdula Geographic Variations in the Prevalence of Obesity, Diabetes, and Obesity Related Behaviors Obesity Research 13 1 January 2005 Thorpe Kenneth, Curtis Florence, Peter Joski Which Medical Conditions Account for The Rise in Health Care Spending? Health Affairs, Web Exclusive, August 25, 2004 Urban Institute estimates of MEPS 2003 data Medicaid covers diabetes testing supplies and medications for diabetics, however they are optional services and this coverage can vary from state to state
2 3
4
5 6
7
Congressional Budget Office, An Analysis of the Literature on Disease Management Programs, October 2004 8 Smith Vernon, Kathleen Gifford, Eileen Ellis, Amy Wiles, Robin Rudowitz, Molly OMalley, Caryn Marks Low Medicaid Spending Growth Amid Rebounding State Revenues: Results from a 50 State Medicaid Budget Survey State Fiscal Years 2006 and
2007 Kaiser Commission on Medicaid and the Uninsured, October 2006
9
CDC National Diabetes Fact Sheet, United States 2005 CDC National Diabetes Fact Sheet, United States 2005 CDC Fact Sheet: SEARCH for Diabetes in Youth Hogan Paul, Tim Dall, Plamen Nikolov Economic Costs of Diabetes in the US in 2002 Report from the American Diabetes Association Diabetes Care 26 3, March 2003 American Diabetes Association Direct and Indirect Costs of Diabetes in the United States HCUP highlights Economic and Health Costs of Diabetes, January 2005, Pub 05 0034 We adjust aggregate spending amounts by state listed in the MSIS to mirror those listed on the FFY 2003 CMS Form 64, an aggregate audited report used to determine federal matching payments The totals on the CMS 64 are considered a more accurate representation of actual federal and state spending We excluded Maryland due to general data reporting errors that caused an overstatement of total expenditures in FY 2003 by roughly 500 million and misallocation of dollars between service categories The ICD 9 codes used to determine the presence of diabetes include 250xx, 3572, 3620x, 36641, 6480x CDC National Diabetes Fact Sheet, United States 2005
Dual eligibles are Medicaid enrollees who are jointly covered by Medicare We excluded enrollees with managed care coverage hmo, php or pccm in these tables because spending on specific services such as inpatient or home health care is often reported under the
10 11 12
13 14 15
16
17
18 19 20
11
managed care categories and not the actual service, causing an underestimation of dollars actually spent on a specific service
21
Kleinfeld NR In Diabetes, One More Burden For The Mentally Ill New York Times June 12, 2006; Grady Denise Link Between Diabetes and Alzheimers Deepens New York Times July 17, 2006
1
00
1330 G STREET NW, WASHINGTON, DC 20005 PHONE: 202 347-5270, FAX: 202 347-5274 WEBSITE: WWWKFFORG/KCMU
Additional copies of this report 0000 are available on the K a i s e r F a m i l y F o u n d a t i o n w e b s i e t w w w k f f o o n t h e Kaiser Family Foundations swebsitet at awwwkfforg r g
Additional copies of this report 7700 are available
The Kaiser Commission on Medicaid and the Uninsured provides information and analysis on health care coverage and access for the lowincome population, with a special focus on Medicaids role and coverage of the uninsured
Begun in 1991 and based in the Kaiser Family Foundations Washington, DC office, the Commission is the largest operating program of the Foundation The Commissions work is conducted by Foundation staff under the guidance of a bi-partisan group of national leaders and experts in health care and public policy
Source:medicare.gov
