Health care utilization patterns in early symptomatic AD prior to enrollment into Phase 2/3 clinical trial AMARANTH

Abstract

AbstractBackgroundIn Alzheimer’s disease (AD), pathogenic changes occur decades before clinical symptoms and diagnosis. Pre‐diagnostic medical history and healthcare resource utilization (HCRU) patterns of patients with early symptomatic AD may aid patient identification earlier in the disease. This study examined HCRU in a subgroup of clinical trial patients to detect early indicators of cognitive, functional and/or behavioral decline.MethodsAMARANTH (NCT02245737) was a double‐blind, placebo‐controlled multi‐center trial that enrolled amyloid‐positive patients with early symptomatic AD (mild cognitive impairment (MCI) due to AD or mild AD dementia). This addendum study analyzed baseline characteristics and Medicare claims of a subgroup of participants compared to two larger age‐matched cohorts of Medicare beneficiaries (Cognitive impairment (CI)/AD and non‐CI/AD matched controls) (Figure 1). AMARANTH participants consented to access of 8 years of prior Medicare claims data and had baseline data available from the parent study. Participants were stratified by APOE4 carrier status, disease, and gender. Descriptive statistics and regression models assessed differences in participant costs across strata and the diagnostic histories between participants and controls. Variables considered in stepwise selection of model covariates included observation months, age, frailty, geography, and comorbidities.ResultsAddendum participants were representative of the AMARANTH cohort but were older and less functionally and cognitively impaired at study entry. Medicare costs did not differ across APOE4 or disease status, but costs for men were lower. Pre‐diagnosis, participant and control costs were comparable, while an increase at diagnosis was observed in the CI/AD controls. A summary of diagnostic histories between participants and controls is shown in Table 1. Post‐diagnostic drug utilization was higher in participants. They were more likely than CI/AD controls to have used psychotherapeutic and neurological agents, and more likely than non‐CI/AD controls to have used antidepressants/anticonvulsants and analgesics/anti‐inflammatory/genitourinary agents.ConclusionsPost‐diagnostic HCRU patterns and costs were higher in CI/AD controls than participants, who did not differ from non‐CI/AD controls; however, participants may not be representative of the broader Medicare CI/AD population. This study provides descriptive HCRU patterns of participants years prior to diagnosis; future analysis with larger cohorts may refine and better identify diagnostic indicators in early symptomatic AD patients.