Burden of illness among commercially insured patients with Alzheimer’s disease

Alzheimer’s & Dementia 3 (2007) 204 –210

Burden of illness among commercially insured patients with Alzheimer’s disease Amie T. Joycea, Yang Zhaob, Lee Bowmanb, Jennifer A. Flynnb, Chureen T. Cartera, Daniel A. Ollendorfa,* a

PharMetrics Inc (a unit of IMS), Watertown, MA, USA b Eli Lilly & Co, Indianapolis, IN, USA

Abstract

Background: The purpose of this study was to examine the direct costs of care for patients newly diagnosed with Alzheimer’s disease (AD) by using retrospective healthcare claims data. Methods: Patients aged 65 years or older with ⱖ1 claims containing a listed diagnosis of AD between January 1999 and November 2003 were selected. The first observed AD claim was deemed the index date. Control patients with no evidence of AD or dementia also were identified. Patients had a 12-month pre-index period and minimum 30-day follow-up. AD patients were matched to control patients on the basis of age, gender, and follow-up duration. Annualized utilization and costs were calculated; generalized linear models (GLM) were undertaken, controlling for demographic and clinical characteristics. Analyses focused on differences in costs among AD patients and controls. Results: The average age of AD and control patients was 82 years; 38% were men (n ⫽ 2,475 AD; 4,950 controls). Average total costs for AD patients were more than five-fold higher compared with controls ($28,263 vs $5,880; P ⬍ .001), driven primarily by inpatient costs. Total costs were significantly higher (P ⬍ .001) for AD patients in GLM modeling, with diagnosis group (AD vs control) as the most important predictor. Adjusted annual costs per patient were also five-fold higher ($21,150 vs $4,053 for AD vs control, respectively) during the follow-up period. Conclusions: The annual economic burden of AD to third-party payers is more substantial than previously estimated ($3,805 to $8,200). © 2007 The Alzheimer’s Association. All rights reserved.

Keywords:

Alzheimer’s disease; Cost of illness; Managed health care insurance plans; Direct service costs; Medicare plus Choice; Prescription insurance

1. Introduction Alzheimer’s disease (AD) is a progressive, neurodegenerative disease and is the most common cause of dementia in the elderly, with a prevalence rate of 5.7% among those aged 65 years and older [1,2]. More than 4 million older Americans have AD; this number is expected to quadruple by the year 2050 as more individuals live into their 80s and 90s [3]. Age is the most important risk factor for

Dr Carter is now employed by Ortho Biotech, LLC, Raritan, NJ, USA. *Corresponding author. Tel.: 617-393-8322; Fax: 617-972-8587. E-mail address: [email protected]

developing the disease; the number of persons with AD doubles every 5 years beyond age 65. Other risk factors include genetics and the presence of the protein apolipoprotein E [4]. AD has a substantial impact on patients, caregivers, and the healthcare system. A patient with AD will expend on average $174,000 in lifetime costs of care [3]. Approximately 70% of people with AD are cared for at home by family or friends [5]; annual out-of-pocket expenditures for relatives of AD patients are estimated to be more than $12,500 [6]. The government also bears a significant portion of the cost of AD care. Most of AD-related direct healthcare costs, such as nursing home costs, are assumed by third-

1552-5260/07/$ – see front matter © 2007 The Alzheimer’s Association. All rights reserved. doi:10.1016/j.jalz.2007.04.373

A.T. Joyce et al / Alzheimer’s & Dementia 3 (2007) 204 –210

party payers and are expected to continue to escalate in the future. It has been reported by the Alzheimer’s Association that nursing home care ranges between $42,000 and $70,000 per patient annually [3]. By 2010, Medicare is anticipating healthcare expenditures of $49.3 billion for its beneficiaries with AD [7], which will represent a 54.5% increase from 2000. In comparison, Medicare spends $26 billion annually on cardiovascular disease–related hospitalizations [8], and 30% of all Medicare expenditures ($72.3 billion annually) are attributable to beneficiaries with diabetes [9]. It has been estimated that in 2010, Medicaid will spend close to $33 billion in residential care costs related to dementia. Currently, the U.S. spends at least $100 billion a year in direct and indirect costs for the treatment and care of patients with AD, which is considered the third most costly disease to treat [2,10]. As the prevalence of AD grows as a result of rapid aging of the U.S. population, the economic burden of AD will become increasingly important to thirdparty payers. Although there have been a number of studies examining the costs attributable to AD in the U.S. [11–17] and globally [18 –20], the U.S. studies have been largely limited to government beneficiaries or individual commercial insurers. One-year attributable costs among those with AD have been estimated to range between $3,805 to $8,200 in prior published studies [12–15]. These studies were primarily conducted during the mid-1990s, and the majority focused on patients with AD or related dementias [12–15]. Another study that focused on AD patients only did not include pharmacy data or nursing home data [16]. In contrast to the methodologies used above, our study focused on the attributable costs of AD among Medicare patients enrolled in managed care organizations (known as Medicare Risk, Medicare⫹Choice, or Medicare Advantage plans). Such an analysis might be viewed as timely and warranted as a result of the recent implementation of a prescription drug benefit for Medicare beneficiaries under the Medicare Modernization Act (MMA) and the expected increase in enrollment in managed care organizations by Medicare beneficiaries. Our study focused on patients with a diagnosis of AD; those with evidence of dementia and no definitive AD diagnosis were not included. We examined AD patients and demographically matched controls who were enrolled in a Medicare Risk plan between 1999 and 2003 within multiple managed care organizations to examine the attributable direct medical costs associated with AD. 2. Methods 2.1. Data source Medical and pharmaceutical claims were obtained from the PharMetrics Patient-Centric Database for the period January 1998 to December 2003. At the time of the study,

205

this retrospective database included fully adjudicated claims from 78 health plans across the U.S. Patients in the database are representative of the national commercially insured population on a variety of demographic measures including geographic region, age, gender, and health plan type. The data are also longitudinal, with an average member enrollment time of 2 years. Inpatient and outpatient diagnoses (International Classification of Diseases, 9th Revision, Clinical Modification [ICD-9-CM] format), procedures (Current Procedural Terminology, 4th edition [CPT-4], and HCFA Common Procedural Coding System [HCPCS] formats), and standard and mail order prescription records are included in the data set. Reimbursed payments (ie, payments by health plans for services rendered and medications dispensed, net of any patient copayment or coinsurance) are available for all services rendered, as well as dates of service for all claims. Additional data elements include demographic variables (eg, age, gender, geographic region), plan type (eg, health maintenance organization [HMO], preferred provider organization [PPO]), payer type (eg, commercial, self-pay), provider specialty, and plan enrollment start and stop dates. 2.2. Sample selection Patients aged 65 years and older with one or more claims containing a listed diagnosis of AD (ICD.9.CM 331.0) between January 1999 and November 2003 were first selected. Only patients enrolled in a Medicare Risk plan were included in the study (because full utilization and cost data were not available for patients with other third-party payer types because of coordination of benefits with other payers). The date of the first observed AD medical or facility claim was deemed the index date. Patients were required to have a 12-month pre-index period with no diagnosis for AD or dementia (ICD.9.CM 290.xx). Because of the high morbidity rate among patients with AD, follow-up was allowed to vary among patients. This allowed patients to be observed and the cost burden assessed for the duration of time patients were enrolled in their health plan (to prevent a biased sample of “survivors”). Follow-up varied beyond a minimum of 30 days and terminated as of the date of patient disenrollment or the end of available data, whichever came first. All medical and pharmacy claims spanning the period January 1998 to December 2003 were extracted for each eligible patient. Patients not continuously eligible for drug and health benefits during their entire pre-index and follow-up periods were excluded, as were those whose insurance coverage did not include retail pharmacy or whose health plan did not report days supplied or quantity dispensed. A cohort of patients with no evidence of AD or dementia (defined as a diagnosis of dementia, dementia-related diagnoses, and AD medications) in their claims history was identified as a control group. Patients in the control group

206

A.T. Joyce et al / Alzheimer’s & Dementia 3 (2007) 204 –210

were matched (2:1) to the AD cohort on the basis of age on the index date (within 1 year), gender, and duration of follow-up (30-day intervals). Candidate controls were first selected on the basis of age (as calculated on the basis of patient birth year) and gender; the index date associated with the AD patient was then assigned to each candidate control patient. To ensure comparable periods of observation, patients were then matched on the basis of duration of follow-up (within 30 days), and all unsuccessfully matched patients were excluded. 2.3. Measures Patient demographics (ie, age, gender, plan type, physician specialty, all measured as of the index date, as well as pre-index health care costs), clinical characteristics (ie, ADrelated comorbidities and AD-related symptoms; see Appendix for codes), and baseline comorbidity profile (as defined by the Charlson Comorbidity Index score) [21] were measured during the pre-index period. Utilization and costs (ie, health plan payments to providers) of all inpatient, outpatient, and pharmacy services were examined for both the prediagnosis and follow-up periods. Measures of interest included the frequency of confinements (ie, inpatient stays in nursing home, hospital, psychiatric facility) and corresponding confinement days, the number of physician, emergency department, hospital outpatient, and home health visits, laboratory tests, outpatient AD-related and other pharmacy claims (retail and mail order), and all other ancillary services. Measures were examined on an overall basis as well as stratified by diagnosis group (AD vs controls). 2.4. Analyses Utilization and direct medical costs were calculated for a 12-month follow-up period. Because the duration of follow-up varied, all measures of utilization and cost were reported on an annualized basis. All costs were expressed in 2003 dollars and were adjusted as necessary by using the medical care component of the U.S. Consumer Price Index (unpublished data, U.S. Bureau of Labor Statistics, 2004). Univariate statistics were calculated with Wilcoxon rank-sum tests for each measure. In addition to unadjusted comparisons, generalized linear models (GLM) with a gamma distribution and log-link function [22] were used to examine the difference in total costs between the study groups. Candidate explanatory variables as described above were introduced into the model; variables were retained if significant (at P ⬍ .10) and if model results changed materially. Covariates in the total cost model included age (65 to 69 years, 70 to 74 years, 75 to 79 years, 80 to 84 years, 85⫹ years), presence of comorbidities (ie, diabetes, coronary artery disease, cerebrovascular disorders), pre-index AD-related abnormality of gait, pre-index use of antipsychotics, and pre-index 1-year health

care costs. An offset variable that weighted the number of days a patient was observed was included in the GLM models to account for differential follow-up (ie, rightcensoring). All patients were therefore weighted to yield expected measures at 12 months. All analyses were conducted with Statistical Analysis Software (SAS Institute Inc, Cary, NC), version 8.2. 3. Results 3.1. Patient population A total of 15,272 patients with a diagnosis of AD who were 65 years or older and enrolled in a Medicare Risk plan were identified. After applying enrollment criteria exclusions (ie, continuous enrollment 1 year before and 30 days after first AD diagnosis), 4,152 dementia patients remained in the study sample. Following the matching procedure, 2,475 AD patients and 4,950 controls were successfully matched. The average age in both groups was 82 years; 38% were men (Table 1). The average duration of follow-up was approximately 9 months (minimum, 31 days; maximum, 1,214 days). Plan type differed significantly (P ⬍ .001) between the two groups (90.5% vs 87.8% for HMO and 4.4% vs 7.3% for PPO for AD and controls, respectively). Not surprisingly, patients in the AD group were also more likely (P ⬍ .001) to be seen by a neurologist on their index date as compared with controls (15.3% vs 0.2%) and less likely to be seen by a cardiologist (3.2% vs 4.9%). Average 12-month pre-index costs were approximately $6,000 higher among AD patients ($11,513 vs $5,576; P ⬍ .001). Patients in the AD group had a significantly (all P ⬍ .001) higher prevalence of the majority of the cardiovascular comorbidities of interest compared with control group patients, including acute myocardial infarction (3.4% vs 1.8%), coronary artery disease (22.4% vs 19.0%), cerebrovascular diseases (21.1% vs 6.0%), and peripheral vascular disease (8.8% vs 5.9%), whereas the prevalence of hyperlipidemia was significantly lower in the AD group (20.5% vs 28.5%; P ⬍ .001). Patients with AD had a higher average Charlson score (1.7 vs 1.2; P ⬍ .001). As expected, patients in the AD group had a higher prevalence rate of symptoms related to AD and dementia (data not shown). Most notably, 15.0% of patients in the AD group had evidence of malaise/fatigue, as compared with 7.2% of controls (P ⬍ .001). More than three times as many AD patients had evidence of memory disturbance (21.7% vs 6.1%; P ⬍ .001) or abnormality of gait (7.9% vs 1.8%; P ⬍ .001). 3.2. Observed annualized utilization Resource utilization results are reported in text only. Patients in the AD group had significantly (all P ⬍ .001)

A.T. Joyce et al / Alzheimer’s & Dementia 3 (2007) 204 –210

207

Table 1 Demographic characteristics of the study sample (AD vs control) AD Total patients Age (mean, SD) Age groups (n, %): 65–69 y 70–74 y 75–79 y 80–84 y 85⫹ years Gender (n, % male) Plan type (n, %): HMO Indemnity POS PPO Other Unknown Physician specialty (n, %): FP/GP Internal medicine Cardiologist Neurology Neurosurgery Psychiatry Other Unknown Pre-index AD comorbidities: (n, %) Diabetes Hyperlipidemia Hypertensive heart disease Acute myocardial infarction Coronary artery disease (CAD) Cerebrovascular diseases Peripheral vascular disorder (PVD) Charlson Comorbidity Index (mean, SD) Pre-index 12-month total healthcare costs (mean, SD)

Control Group

2,475 82.0

8.3

114 329 629 670 733 940

4.6 13.3 25.4 27.1 29.6 38.0

2,241 2 117 110 0 5 262 210 80 379 24 93 1,105 322 537 508 1,351 85 555 521 219 1.68 $11,513

4,950 82.0

P Value 8.3

1.000

228 658 1,258 1,340 1,466 1,880

4.6 13.3 25.4 27.1 29.6 38.0

1.000

90.5 0.1 4.7 4.4 0.0 0.2

4,345 1 216 362 0 26

87.8 0.0 4.4 7.3 0.0 0.5

⬍.001

10.6 8.5 3.2 15.3 1.0 3.8 44.6 13.0

439 590 245 11 43 6 2,084 1,532

8.9 11.9 4.9 0.2 0.9 0.1 42.1 30.9

⬍.001

21.7 20.5 54.6 3.4 22.4 21.1 8.8 1.92 $28,174

907 1,413 2,591 91 940 299 293 1.20 $5,576

18.3 28.5 52.3 1.8 19.0 6.0 5.9 1.66 $12,586

1.000

⬍.001 ⬍.001 0.068 ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001 ⬍.001

Abbreviations: SD, standard deviation; POS, point of service; FP/GP, family practice/general practice.

higher annualized pharmacy, outpatient, and inpatient utilization as compared with the control group. Patients in the AD group averaged 18 office visits, four emergency department visits, and 13 laboratory/diagnostic tests annually as compared with nine office visits, one emergency department visit, and seven laboratory/diagnostic tests for matched controls. On average, each patient in the AD group was hospitalized twice per year; in contrast, controls averaged 0.24 confinements annually (P ⬍ .001). Correspondingly, confinement days were significantly higher in the AD group (22.8 vs 1.6 days; P ⬍ .001). 3.3. Annualized costs Average total costs for AD patients were more than five-fold higher in the AD group as compared with matched controls ($28,263 vs $5,880; P ⬍ .001) (Table 2). Average total pharmacy costs were approximately $800 higher, outpatient costs were approximately $2,300

higher, and inpatient costs were approximately $19,000 higher in the AD group (pharmacy: $1,797 vs $1,024; outpatient: $4,877 vs $2,508; inpatient: $21,589 vs $2,348; all P ⬍ .001). 3.4. Multivariate adjusted annualized costs Total costs were significantly higher for AD patients in multivariate analyses, with diagnosis group (AD vs control) as the most important predictor of total costs (parameter estimate, 1.448; standard error, 0.039; P ⬍ .001) (Table 3). Adjusted annual costs per patient were also five-fold higher ($21,150 vs $4,053 for AD vs control, respectively) during the follow-up period; although absolute differences in pharmacy and outpatient costs were reduced relative to the observed data, the magnitude of these differences was similar (pharmacy: $1,357 vs $725 for AD vs control; outpatient: $3,597 vs $1,771 for AD vs control). Multivariate analyses of inpatient costs were not feasible, because not all

208

A.T. Joyce et al / Alzheimer’s & Dementia 3 (2007) 204 –210

Table 2 Annualized 1-year cost for AD vs control groups AD (n ⫽ 2,475) Medications & Services Pharmacy claims Acetylcholinesterase inhibitors Antidepressants Anxiolytics Antipsychotics Other pharmacy Services Total pharmacy Outpatient claims Outpatient management Outpatient emergency department Laboratory/diagnostic tests Outpatient ancillary Total outpatient Inpatient claims Confinement costs Grand total

Control Group (n ⫽ 4,950)

Mean

SD

Mean

SD

$279 $125 $16 $177 $1,200 $1,797

$475 $283 $75 $585 $4,777 $4,944

$0 $26 $7 $10 $981 $1,024

$0 $126 $49 $260 $2,077 $2,134

$1,287 $508 $239 $2,842 $4,877

$2,890 $2,219 $547 $6,721 $8,807

$591 $135 $134 $1,648 $2,508

$4,162 $850 $339 $4,526 $6,793

$21,589 $28,263

$55,934 $59,415

$2,348 $5,880

$14,250 $16,907

Abbreviation: SD, standard deviation. NOTE. All comparisons between AD and Control group significant at level P ⬍ .001.

patients in each group were hospitalized during the study period (ie, a subset of patients had zero values for inpatient costs). 4. Discussion This study of patients enrolled in a Medicare Risk plan within multiple managed care organizations suggests that patients with AD have 1-year total costs that are nearly five-fold higher than patients without evidence of AD or dementia (adjusted costs, $21,150 vs $4,053 for AD and controls, respectively). This study indicates that the attributable costs of AD are more than $17,000 annually. The majority of this attributable cost is due to inpatient costs, although both pharmacy

costs and outpatient costs are also higher among AD patients. Total estimated direct medical costs for patients with AD are higher in our study ($21,150) than those reported in previous studies examining costs to managed care organizations ($7,284 to $13,487) [11,12,15]. However, our study expressed costs in 2003 dollars, whereas the remaining studies estimated costs between 1996 and 1999. In addition, two of the studies included an imputation of costs based on the Medicare Current Beneficiary Survey [11] or the Medicare fee schedule for provider costs [15], possibly resulting in an underestimation of costs. Finally, the average age reported in our study was 82 years as compared with 79 to 80 years in the studies by Hill et al [15] and Leon and Neumann [11]. This might

Table 3 GLM model comparing total cost during follow-up for AD vs control groups Wald 95% Confidence Limits Variable

Estimate

Standard Error

Lower Limit

Upper Limit

␹2

P Value

Intercept AD vs Control group Age 65–69 y vs ⱖ 85 y Age 70–74 y vs ⱖ 85 y Age 75–79 y vs ⱖ 85 y Age 80–84 y vs ⱖ 85 y Pre-index diagnosis of diabetes Pre-index diagnosis of CAD Pre-index diagnosis of cerebrovascular diseases Pre-index use of antipsychotics Pre-index diagnosis of abnormality of gait Log (total pre-index costs)

1.5624 1.4479 ⫺0.2903 ⫺0.1266 ⫺0.0195 0.0510 0.2158 0.2949 0.1703 0.2420 0.2384 0.1368

0.0561 0.0388 0.0865 0.0573 0.0471 0.0460 0.0451 0.0451 0.0574 0.0748 0.0926 0.0063

1.4524 1.3719 ⫺0.4598 ⫺0.2389 ⫺0.1117 ⫺0.0390 0.1275 0.2065 0.0577 0.0954 0.0569 0.1245

1.6723 1.5239 ⫺0.1208 ⫺0.0143 0.0728 0.1411 0.3041 0.3832 0.2828 0.3886 0.4199 0.1492

775.62 1392.91 11.27 4.89 0.17 1.23 22.94 42.75 8.79 10.47 6.62 470.81

⬍.001 ⬍.001 ⬍.001 .027 .679 .267 ⬍.001 ⬍.001 .003 .001 .010 ⬍.001

Abbreviation: CAD, coronary artery disease.

A.T. Joyce et al / Alzheimer’s & Dementia 3 (2007) 204 –210

have also influenced differences in the total cost estimates. Prior studies have reported 1-year attributable costs among those with AD ranging from $3,805 to $8,200 [12–15]. In a study of a large managed care organization, Gutterman et al [12] found that total costs were approximately 1.5 times higher for patients with AD or related dementia as compared with controls [12]. This study found that 75% of the increased costs were due to inpatient expenses. In their claims-based study of Georgia Medicaid recipients, Martin et al [14] also determined that the majority of these differences (85%) were due to inpatient costs. In our study, we found that 86% of the cost difference was contributed by inpatient costs. AD patients had an average of 20 additional hospital days and eight times as many emergency department visits. These results are more pronounced than those reported in another retrospective analysis of a managed Medicare population, which reported a two-fold difference in emergency department visits and 7 additional hospital days in comparison to matched controls [13]. Patients in our study were compared with the average elderly patient (ie, matched on the basis of age, gender, and duration of follow-up), whereas the previously mentioned two studies [13,14] were also matched on overall disease severity, which might have contributed to lower attributable costs of AD than observed in this study. However, in our study, even after adjustment for demographic and clinical characteristics in GLM modeling, patients in the AD group had more than five-fold higher total costs as compared with control patients ($21,150 vs $4,053). Our study is subject to some important limitations. First, our sample was restricted to include only patients 65 years of age or older who were enrolled in a Medicare Risk plan. These patients who are enrolled in a managed Medicare benefit might differ in certain respects from the overall U.S. elderly population (eg, demographics, severity). Therefore, our cost estimates might not be truly representative of all elderly persons with and without evidence of AD. However, we do believe that our findings will generalize well to the increasing percentage of Medicare beneficiaries who are currently enrolling in managed care organizations through the implementation of the Medicare drug benefit. A comparison with other previously published analyses with managed Medicare data [12,15,16] would add to an updated scholarly perspective on this issue, because most managed care organizations have historically had little insight into the costs of AD because of their previous lack of participation in managed Medicare arrangements. In addition, coverage of nursing home admissions by private insurers is limited. Because our analysis is limited to patients enrolled in a managed Medicare plan, costs associated with nursing home admissions are likely understated in this study.

209

We cannot rule out the possibility that factors influencing the differences between the diagnosis groups were not controlled for. As in most observational studies, we could only control for observed differences detectable in administrative claims data but were not able to control for unobserved differences. Nevertheless, observed differences in cost between AD patients and those without evidence of AD or dementia persisted when analyses controlled for betweengroup differences.

5. Conclusions Despite these limitations, our study has important implications. Our findings suggest that the attributable costs of AD to third-party insurers might be greater than previously estimated. Because the care of AD is likely to grow as a public health concern, postponing the emergence of symptoms of AD will bring direct benefits to patients living with AD, their families, and the healthcare system. With the introduction of a drug benefit as part of the MMA, access to treatment options for AD patients that slow (and eventually halt) disease progression is expected to improve. Treatment options that improve cognitive and daily functioning might lower overall healthcare costs and minimize the burden of AD.

Acknowledgments This study was funded through an unrestricted research grant from Eli Lilly & Co, Indianapolis, IN. This study was designed and conducted primarily by employees of PharMetrics, Inc (Ms Joyce, Mr Ollendorf, and at the time of the study, Dr Carter). Drs Zhao and Bowman as well as Ms Flynn are employees of Eli Lilly and provided assistance with study conception and design, review of findings, and manuscript review and comment.

References [1] Bird TD. Alzheimer’s disease and other primary dementias. In: Braunwald E, Fauci A, Hauser S, Jameson J, Kapser D, et al. Harrison’s principles of internal medicine. 15th ed. New York: McGraw Hill; 2001, p. 2391– 8. [2] Rice DP, Fillit HM, Max W, Knopman DS, Lloyd JR, Duttagupta S. Prevalence, costs, and treatment of Alzheimer’s disease and related dementia: a managed care perspective. Am J Manag Care 2001;7: 809 –18. [3] Statistics about Alzheimer’s disease. Available at: http://www.alz. org/AboutAD/Statistics.htm. Alzheimer’s Association, 2004. Accessed September 23, 2004. [4] Causes of Alzheimer’s disease. Available at: http://www.alzheimers. org/causes.htm. Alzheimer’s Disease Education and Referral Center, 2004. Accessed September 23, 2004. [5] Losing a million minds: confronting the tragedy of Alzheimer’s disease and other dementias. Washington, DC: US Congress Office of Technology Assessment, U.S. Government Printing Office; 1987.

210

A.T. Joyce et al / Alzheimer’s & Dementia 3 (2007) 204 –210

[6] Rice DP, Fox PJ, Max W, Webber PA, Lindeman DA, Hauck WW, et al. The economic burden of Alzheimer’s disease care. Health Affairs 1993;12:164 –76. [7] Medicare and Medicaid costs for people with Alzheimer’s disease. Washington, D.C: Alzheimer’s Association; 2001. [8] Heart disease burden: chronic disease notes and reports. Atlanta: Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health; 2004. [9] Diabetes Self Management Training Act of 2005. Available at: http:// www.theorator.com/bills109/s626.html. Data referenced in bill presented to the US 109th Congress, 1st session, S. 626, March 15, 2005. Accessed: June 16, 2005. [10] 2001-2002 Alzheimer’s disease progress report. Bethesda, MD. National Institutes of Health. Publication number 03-5333, 2003. [11] Leon J, Neumann PJ. The cost of Alzheimer’s disease in managed care: a cross-sectional study. Am J Manag Care 1999;5:867–77. [12] Gutterman EM, Markowitz JS, Lewis B, Fillit H. Cost of Alzheimer’s disease and related dementia in managed Medicare. J Am Geriatr Soc 1999;47:1065–71. [13] Fillit H, Hill JW, Futterman R. Health care utilization and costs of Alzheimer’s disease: the role of comorbid conditions, disease stage, and pharmacotherapy. Fam Med 2002;34:528 –35. [14] Martin BC, Ricci JF, Kotzan JA, Lang K, Menzin J. The net cost of Alzheimer disease and related dementia: a population-based study of Georgia Medicaid recipients. Alzheimer Dis Assoc Disord 2000;14: 151–9. [15] Hill JW, Futterman R, Duttagupta S, Mastey V, Lloyd JR, Fillit H. Alzheimer’s disease and related dementias increase costs of comorbidities in managed Medicare. Neurology 2002;58:62–70. [16] Richards KM, Shepherd MD, Crimson ML, Snyder EH, Jermain DM. Medical services utilization and charge comparisons between elderly patients with and without Alzheimer’s disease in a managed care organization. Clin Ther 2000;22:775–91. [17] Menzin J, Lang K, Friedman M, Neumann P, Cummings JL. The economic cost of Alzheimer’s disease and related dementias to the California Medicaid Program (“Medi-Cal”) in 1995. Am J Geriatr Psychiatry 1999;7:300 – 8. [18] Rigaud AS, Fagnani F, Bayle C, Latour F, Traykov L, Forette F. Patients with Alzheimer’s disease living at home in France: costs and consequences of the disease. J Geriatr Psychiatry Neurol 2003;16: 140 –5. [19] Beeri MS, Werner P, Adar Z, Davidson M, Noy S. Economic cost of Alzheimer disease in Israel. Alzheimer Dis Assoc Disord 2002;16: 73– 80. [20] Atance Martinez JC, Yusta Izquierdo A, Grupeli Gardel BE. Costs study in Alzheimer’s disease. Rev Clin Esp 2004;204:64 –9. [21] Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 1987;40:373– 83. [22] McCullagh P, Nelder JA. Generalized linear models. London: Chapman and Hall; 1989.

Appendix Codes for selected AD comorbidities ICD-9-CM Description

ICD-9-CM Code

Diabetes Hyperlipidemia Hypertensive heart disease Acute myocardial infarction Coronary artery disease Cerebrovascular diseases Peripheral vascular disorder (PVD)

250.xx 272.xx 401.x to 405.x 410.x 411.x, 414.x, 429.x 434.x to 438.x 440.x to 441.x, 446, 459.x

Codes for selected AD and dementia symptoms Diagnosis Code Description

ICD-9-CM Code

Pseudobulbar palsy Hemiparesis, hemiplegia, unspecified Other specified paralytic syndrome, BabinskiNageotte Paralysis, unspecified Facial nerve disorders Psychophysical visual disturbances, visual: agnosia, disorientation syndrome, hallucinations Homonymous/heteronymous bilateral field defects Nystagmus and other irregular eye movements Difficulty in walking Unspecified disorder of muscle, ligament, and fascia Other malaise and fatigue, asthenia NOS; lethargy; postviral (asthenic) syndrome; tiredness Other general symptoms, memory disturbance, loss or lack Abnormal involuntary movements, abnormal head movements; fasciculation; spasms NOS; tremor NOS Abnormality of gait, gait: ataxic, paralytic, spastic, staggering Lack of coordination, ataxia NOS; muscular incoordination Aphasia Other speech disturbance, dysarthria; dysphasia; slurred speech Other symbolic dysfunction, acalculia; agnosia; agraphia NOS; apraxia Urinary incontinence

335.23 342.8x, 342.9x 344.89 344.9 351.x 368.16

368.46, 368.47 379.5x 719.7 728.9 780.79

780.9 781

781.2 781.3 784.3 784.5 784.69 788.3, 788.31, 788.33, 788.34, 788.39