Original articleThe accuracy of medicare claims data in identifying Alzheimer's disease
Introduction
Medicare claims records allow for the study of cost and outcomes in broadly representative samples of persons, but it is not known how reliable ICD-9-CM diagnosis codes contained in claims data are for identifying persons with Alzheimer's disease (AD). Several publications using Medicare claims records to identify prevalent cases of the disease appear to have seriously underestimated the prevalence of AD when compared with the estimated prevalence in the elderly population as a whole 1, 2, 3, 4, 5. Such undercounting may bias the results of work based on a Medicare claims-based estimate of prevalent cases of AD. This raises the issue of how reliable Medicare claims data are in identifying AD, whether the apparent undercount of prevalent AD is systematic, and if so, what types of persons with AD are most likely to be missed. If the undercount of AD cases is systematic, it can be modeled and perhaps adjusted for in studies that rely on Medicare claims to identify prevalent AD.
The present study attempts to determine the accuracy of Medicare claims data in identifying AD in a population of persons 65 years of age and over, all of whom have a clinical diagnosis of AD. Our goal is to provide practical guidance to those who wish to use Medicare claims to identify cases of AD. To do so, we examined 5 years of Medicare claims data (1991–1995) for 417 patients with AD who participated in the Consortium to Establish a Registry for Alzheimer's Disease (CERAD).
Section snippets
Subjects
Subjects were drawn from the AD patients entered into CERAD. CERAD, funded by the National Institute on Aging, is a consortium of the memory disorders clinics at 23 major tertiary care medical centers, including all the Alzheimer's Disease Centers [6]. Eligibility criteria for entry into CERAD included age 50 or greater, mild to moderate AD with no other disorders that could affect the dementing condition, independently mobile, a score of greater than 0 on at least two of the five
Descriptive results
Table 1 presents the number of AD and ADRD claims for the 417 patients examined. Roughly eighty (78.9%) percent had at least one Medicare claim during the study period that listed ICD-9-CM code 331.0 (Alzheimer's disease) as a diagnosis code. When using a broader definition of dementia that included additional ICD-9-CM codes, just under 90% (87%) of all persons had at least one Medicare claim with a diagnosis code signifying dementia. Many subjects had multiple Medicare claims that carried
Discussion
There was an undercount of patients with AD in our study. Our results, however, provide some support for using Medicare claims records to identify persons with AD, when Medicare claims allow examination of important research topics that could not otherwise be addressed.
When 3 years of Medicare claims data were used, approximately 75% of CERAD cases who were known to have AD were identified as having AD in Medicare claims; the identification rate rose to over 80% when we used a broader set of
Acknowledgements
This work was supported by NIH Grants No. AG06790, AG08937, and 1PO1-AG-17937. The authors thank Bercedis Peterson for help in data management and comments on the manuscript, and Albert Heyman and Carl Pieper for comments on the manuscript.
References (21)
- et al.
Alzheimer's disease under managed careimplications from Medicare utilization and expenditure patterns
J Am Geriatr Soc
(1999) - General Accounting Office. Alzheimer's disease: estimates of prevalence in the United States. Washington, DC: U.S....
- et al.
Cost of Alzheimer's disease and related dementia in managed medicine
J Am Geriatr Soc
(1999) - et al.
Misclassification and selection bias when identifying Alzheimer's Disease solely from Medicare claims records
J Am Geriatr Soc
(1999) - et al.
How much do persons with Alzheimer's disease cost Medicare?
J Am Geriatr Soc
(2000) - et al.
Consortium to establish a registry for Alzheimer's Diseasethe CERAD experience
Neurology
(1997) - et al.
The consortium to establish a registry for Alzheimer's disease (CERAD). Part 1. Clinical and neuropsychological assessment of Alzheimer's disease
Neurology
(1989) - et al.
Clinical diagnosis of Alzheimer's diseasereport of the NINCDS-ADRDA Work Group under the auspices of the Department of Health and Human Services Task Force on Alzheimer's Disease
Neurology
(1984) - et al.
A new clinical scale for the staging of dementia
Br J Psychiatry
(1982) The clinical dementia rating (CDR)current version and scoring rules
Neurology
(1993)
Cited by (200)
The disproportionate burden of Alzheimer's disease and related dementias (ADRD) in diverse older adults diagnosed with cancer
2023, Journal of Geriatric OncologyForecasting prevalence and mortality of Alzheimer's disease using the partitioning models
2023, Experimental GerontologyPredicting Alzheimer's Disease and Related Dementias in Heart Failure and Atrial Fibrillation
2023, American Journal of Medicine