Identification of Cognitively Impaired Patients At Risk For Development of Alzheimer's Disease: An Analysis of US Medicare Claims Data

Document Type



Media is loading

Publication Date




J McAna, Jefferson College of Population Health, Thomas Jefferson University, Philadelphia, PA.



It has been reported that roughly 10% of patients with a medical claim for cognitive subsequently have a claim for AD. Identifying the factors associated with conversion of cognitive impairment to AD (or dementia) would be a powerful tool for physicians to initiate an AD or dementia assessment and eventually plan for patient management of the disease if diagnosed. The aim of this study was to identify patient factors present upon a diagnosis of cognitive impairment which may indicate future AD or dementia.


This study utilized a multivariate logistic regression approach to identify potential determinants of progression to AD or dementia amongst a population of cognitively impaired individuals from the Medicare Advantage patient database. Variables were identified from the claims data and used to defined potential parameters of investigation. Two cognitively impaired cohorts were evaluated, one with a broad definition of cognitive impairment and one defined as MCI, so stated (ICD-9 code 331.83).


Although the use of antipsychotics had low prevalence around the time of the CI diagnosis, their use was associated with 2.5x risk of that patient ultimately getting an AD diagnosis. A 75-year old with CI or MCI increases the odds by 70% (OR 1.7) that the patient will have a subsequent AD diagnosis relative to a 65 year old. Diagnosis of CI by neurologist increases odds of a subsequent AD diagnosis. Although indicators of comorbidities were spotty across the data sets, the presence of hypertension medication increased the risk of AD by 25% (OR 1.25). Overall the presence of medication to treat comorbid conditions was a better indicator of risk than the coding for the comorbidity. Protective factors for progression were observed: use of anxiolytics reduced odds of progressing to AD or dementia among the broader defined CI cohorts (OR 0.76); being diagnosed as an inpatient was associated with ½ the chance of a later AD diagnosis; history of stroke reduced odds slightly (OR 0.87) . No risk factors differentiated a subsequent AD vs dementia diagnosis.


Several factors were identified as indicating increased risk of a subsequent AD or dementia diagnosis. By identifying patients at risk for AD, better health care and family planning may be undertaken to improve patient and caregiver outcomes.

Presentation: 40:20

This document is currently not available here.