Changes in Medicare's Chronic Condition Files and Disparities in Dementia Identification

Area of Science:

• Gerontology
• Public Health
• Health Services Research

Background:

• The Chronic Condition Warehouse (CCW) updated its Chronic Condition file from a 27-condition (CC27) to a 30-condition (CC30) version in 2022.
• The CC30 file introduced changes to dementia identification criteria, including a shortened look-back period and increased claim requirements.
• These modifications may impact the identification of dementia cases, particularly among vulnerable populations.

Purpose of the Study:

• To quantify the impact of the CC27 to CC30 transition on dementia identification across diverse demographic and geographic subgroups.
• To examine individual-level factors (race/ethnicity, dual status, age, gender) and area-level factors (provider shortages, social deprivation) associated with dementia identification discrepancies.
• To assess the effect of updated CCW data on Alzheimer's disease (AD) and Alzheimer's disease and related dementias (ADRD) case identification.

Main Methods:

• An observational study utilizing Medicare Beneficiary Summary Files (2018-2021) linked with publicly available data.
• Identification of beneficiaries with AD or ADRD using both CC27 and CC30 criteria.
• Logistic regression analysis to determine factors associated with disagreement in dementia identification between CC27 and CC30.

Main Results:

• The CC30 file identified only 70-72% of AD and ADRD cases previously identified by CC27, adding minimal new cases (0.04%).
• Discrepancies in dementia identification were significantly associated with patient and community characteristics.
• Asian and Hispanic beneficiaries, rural residents, and those in socioeconomically deprived areas showed higher odds of disagreement in ADRD identification.

Conclusions:

• The transition from CC27 to CC30 led to a reduction in identified AD and ADRD cases, disproportionately affecting specific subpopulations.
• These changes necessitate careful consideration by researchers and policymakers when interpreting dementia trends and developing care interventions.
• The findings highlight potential disparities in dementia identification due to data update modifications.