Leveraging Electronic Health Records for Reducing Dementia Screening Disparities in Diverse Communities

Project Summary/Abstract: Alzheimer’s Disease and Alzheimer's Disease related dementias (AD/ADRD) are irreversible conditions affecting over 6 million people in the United States. While the pathogenesis of AD/ADRD is complex, modifiable cardiovascular risk factors contribute to its pathogenesis. These risk factors are unequally distributed, with higher prevalence among socioeconomically disadvantaged and racial and ethnic minority groups, and the problem is compounded by lower rates of cognitive screening and diagnosis across these groups – as documented in Black, Latino, Asian, and Native American populations. While referral to qualified memory specialists (geriatricians and/or neurologists) for evaluation and management is considered gold standard, in practice most diagnosis and management of AD/ADRD occurs in non-specialty (primary care) settings, due to shortage of specialists. Novel cognitive screening programs that leverage electronic health records (EHRs) have the potential to close this gap considerably. In this context, we propose a multicomponent stakeholder- and technology-based framework to improve early detection of MCI and AD/ADRD, to elevate the quality of cognitive screening and secondary prevention within primary care, with a focus on practices that treat patients from socioeconomically and racially diverse backgrounds. We will design and implement an embedded Pragmatic Clinical Trial (ePCT) that involves engaging patients, families and PCPs to refine existing practices on how PCPs should present information on cognitive screening, MCI and AD/ADRD diagnosis, and their linkage with cardiovascular disease. We will leverage an EHR-based clinical decision support tool based on retrospectively validated artificial intelligence (AI) and machine learning (ML) methods for identifying MCI and AD/ADRD. Primary outcome for the ePCT will be the rate of cognitive screening, and secondary outcomes will include diagnosis of dementia, and secondary cardiovascular preventive measures. This ePCT effort will focus on the two NYU Langone Health Brooklyn ambulatory care centers, which include over 70 PCPs and over 10,000 patients age ≥65 years. (With 62% self- identified as non-white). Our Specific Aims are to (Aim 1) Conduct a pilot ePCT to optimize and evaluate the acceptability, usability and adoption of implementing an existing AI/ML tool and PCP decision support in a predominantly racial and ethnic underrepresented minority cohort. (Aim 2) Validate and monitor the ePCT primary outcome of cognitive screening, and secondary outcomes of diagnosis of MCI, AD/ADRD, and cardiovascular disease prevention outcomes. (Aim 3) Perform a sub-analysis of patient outcomes across different racial/ethnic subgroups for differences in uptake of screening, diagnosis, secondary cardiovascular preventative interventions. This pilot ePCT will pave the way for larger scale pragmatic trials that will address an unmet and urgent need in cognitive health management of elderly population, particularly those most vulnerable to the disease.