2025 American Academy of Neurology Abstract Website

Yara Alkhodair¹, Ging-Yuek Hsiung¹, Boon Lead Tee², Pei-Chuan Ho⁴, Phoenix Au Yeung¹, Wai Haung Yu⁵, Guerry Peavy⁶, Victor Henderson⁷, Yun-Beom Choi⁸, Dolly Reyes-Dumeyer⁹, Haeok Lee¹⁰, Walter Kukull¹¹, Collin Liu¹², Yian Gu⁹, Gyungah R. Jun¹³, Van Park³, Li-San Wang⁴, Tiffany Chow⁴
¹Neurology Division, Department of Medicine, University of British Columbia, ²Memory and Aging Center, ³School of Nursing, Department of Community Health Systems, University of California San Francisco, ⁴Department of Pathology and Laboratory Medicine, University of Pennsylvania, ⁵Department of Pharmacology and Toxicology, University of Toronto, ⁶Department of Neurosciences, University of California San Diego, ⁷Departments of Epidemiology & Population Health of Neurology & Neurological Sciences, Stanford University, ⁸Department of Neurology, Rutgers New Jersey Medical School, Englewood Health, ⁹Taub Institute, Columbia University, ¹⁰Meyers College of Nursing, New York University, ¹¹Department of Epidemiology, University of Washington, ¹²Department of Neurology, Keck School of Medicine, University of Southern California, ¹³Biomedical Genetics Section, Department of Medicine, Boston University

Objective:

This study assesses the inter-rater reliability of consensus diagnoses for Alzheimer’s disease (AD) across multiple sites in the Asian Cohort for Alzheimer’s Disease (ACAD) study

Background:

The ACAD study is a multi-site collaborative investigation aimed at exploring the genetic and environmental risk factors for AD among Asian Americans and Asian Canadians. With the rising prevalence of AD in these populations, ACAD aims to enhance our understanding of how these factors contribute to the disease's progression. Data collection was designed to minimize participant burden while considering the historical stigma of dementia in Asian cultures. Standardizing diagnostic procedures across sites ensures consistent, high-quality data.

Design/Methods:

We analyzed 42 participants who had previously undergone a consensus-based diagnostic process across seven ACAD recruiting sites. The cohort included a spectrum of cognitive abilities, from cognitively healthy to cognitively impaired. Cognitive assessments were conducted in six languages to accommodate linguistic diversity and preferences among participants. Two independent neurologists re-evaluated the consensus diagnoses. Cohen’s Kappa coefficient measured inter-rater reliability, providing a robust assessment of diagnostic consistency across evaluators and sites.

Results:

Our results showed an observed agreement rate of 88% between the consensus process at the recruiting sites and the two external raters. Cohen’s Kappa value was also calculated at 0.835, indicating a near-perfect agreement. This high level of consistency affirms the reliability of the ACAD study’s diagnostic protocol, demonstrating that standardized procedures across sites, even with language variability, yield consistent diagnostic outcomes.

Conclusions:

The high inter-rater agreement underscores the effectiveness of the ACAD consensus process in delivering reliable diagnoses. These findings emphasize the importance of maintaining standardized diagnostic protocols across multiple research sites. This study's success not only paves a way for large-scale research into Alzheimer's disease in underrepresented groups but also significantly advances global efforts to foster inclusivity in AD research.