Cognitive measures lacking in EHR prior to dementia or Alzheimer's disease diagnosis

Predicting Risk of Alzheimer's Diseases and Related Dementias with AI Foundation Model on Electronic Health Records

Early identification of Alzheimer's disease (AD) and AD-related dementias (ADRD) has high clinical significance, both because of the potential to slow decline through initiating FDA-approved therapies and managing modifiable risk factors, and to help persons living with dementia and their families to plan before cognitive loss makes doing so challenging. However, substantial racial and ethnic disparities in early diagnosis currently lead to additional inequities in care, urging accurate and inclusive risk assessment programs. In this study, we trained an artificial intelligence foundation model to represent the electronic health records (EHR) data with a vast cohort of 1.2 million patients within a large health system. Building upon this foundation EHR model, we developed a predictive Transformer model, named TRADE, capable of identifying risks for AD/ADRD and mild cognitive impairment (MCI), by analyzing the past sequential visit records. Amongst individuals 65 and older, our model was able to generate risk predictions for various future timeframes. On the held-out validation set, our model achieved an area under the receiver operating characteristic (AUROC) of 0.772 (95% CI: 0.770, 0.773) for identifying the AD/ADRD/MCI risks in 1 year, and AUROC of 0.735 (95% CI: 0.734, 0.736) in 5 years. The positive predictive values (PPV) in 5 years among individuals with top 1% and 5% highest estimated risks were 39.2% and 27.8%, respectively. These results demonstrate significant improvements upon the current EHR-based AD/ADRD/MCI risk assessment models, paving the way for better prognosis and management of AD/ADRD/MCI at scale.

Which factors preceding dementia identification impact future healthcare use trajectories: multilevel analyses in administrative data

BACKGROUND

Healthcare use patterns preceding a diagnosis of Alzheimer's Disease and Related Diseases (ADRD) may be associated with the quality of healthcare use trajectories (HUTs) after diagnosis. We aimed to identify determinants of future favorable HUTs, notably healthcare use preceding ADRD identification.

METHODS

This nationwide retrospective observational study was conducted on subjects with incident ADRD identified in 2012 in the French health insurance database. We studied the 12-month healthcare use ranging between 18 and 6 months preceding ADRD identification. The five-year HUTs after ADRD identification were qualified by experts as favorable or not. In order to take into account geographical differences in healthcare supply, we performed mixed random effects multilevel multivariable logistic regression model to identify determinants of future favorable HUTs. Analyses were stratified by age group (65-74, 75-84, ≥ 85).

RESULTS

Being a woman, and preventive and specialist care preceding ADRD identification increased the probability of future favorable HUT, whereas institutionalization, comorbidities, medical transportation and no reimbursed drug during [-18;-6] months decreased it. Besides, some specificities appeared according to age groups. Among the 65-74 years subjects, anxiolytic dispensing preceding ADRD identification decreased the probability of future favorable HUT. In the 75-84 years group, unplanned hospitalization and emergency room visit preceding ADRD identification decreased this probability. Among subjects aged 85 and older, short hospitalization preceding ADRD identification increased the probability of future favorable HUTs.

CONCLUSION

Regular healthcare use with preventive and specialist care preceding ADRD identification increased the probability of future favorable HUTs whereas dependency decreased it.

Expanding the use of brief cognitive assessments to detect suspected early-stage cognitive impairment in primary care

INTRODUCTION

Mild cognitive impairment remains substantially underdiagnosed, especially in disadvantaged populations. Failure to diagnose deprives patients and families of the opportunity to treat reversible causes, make necessary life and lifestyle changes and receive disease-modifying treatments if caused by Alzheimer's disease. Primary care, as the entry point for most, plays a critical role in improving detection rates.

METHODS

We convened a Work Group of national experts to develop consensus recommendations for policymakers and third-party payers on ways to increase the use of brief cognitive assessments (BCAs) in primary care.

RESULTS

The group recommended three strategies to promote routine use of BCAs: providing primary care clinicians with suitable assessment tools; integrating BCAs into routine workflows; and crafting payment policies to encourage adoption of BCAs.

DISSCUSSION

Sweeping changes and actions of multiple stakeholders are necessary to improve detection rates of mild cognitive impairment so that patients and families may benefit from timely interventions.

Current Diagnostic Pathways for Alzheimer's Disease: A Cross-Sectional Real-World Study Across Six Countries

Background

Diagnostic pathways for patients presenting with cognitive complaints may vary across geographies.

Objective

To describe diagnostic pathways of patients presenting with cognitive complaints across 6 countries.

Methods

This real-world, cross-sectional study analyzed chart-extracted data from healthcare providers (HCPs) for 6,744 patients across China, France, Germany, Spain, UK, and the US.

Results

Most common symptoms at presentation were cognitive (memory/amnestic; 89.86%), followed by physical/behavioral (87.13%). Clinical/cognitive tests were used in > 95%, with Mini-Mental State Examination being the most common cognitive test (79.0%). Blood tests for APOE ɛ4/other mutations, or to rule out treatable causes, were used in half of the patients. Clinical and cognitive tests were used at higher frequency at earlier visits, and amyloid PET/CSF biomarker testing at higher frequency at later visits. The latter were ordered at low rates even by specialists (across countries, 5.7% to 28.7% for amyloid PET and 5.0% to 27.3% for CSF testing). Approximately half the patients received a diagnosis (52.1% of which were Alzheimer's disease [AD]). Factors that influenced risk of not receiving a diagnosis were HCP type (higher for primary care physicians versus specialists) and region (highest in China and Germany).

Conclusion

These data highlight variability in AD diagnostic pathways across countries and provider types. About 45% of patients are referred/told to 'watch and wait'. Improvements can be made in the use of amyloid PET and CSF testing. Efforts should focus on further defining biomarkers for those at risk for AD, and on dismantling barriers such low testing capacity and reimbursement challenges.

Assess the documentation of cognitive tests and biomarkers in electronic health records via natural language processing for Alzheimer's disease and related dementias

BACKGROUND

Cognitive tests and biomarkers are the key information to assess the severity and track the progression of Alzheimer's' disease (AD) and AD-related dementias (AD/ADRD), yet, both are often only documented in clinical narratives of patients' electronic health records (EHRs). In this work, we aim to (1) assess the documentation of cognitive tests and biomarkers in EHRs that can be used as real-world endpoints, and (2) identify, extract, and harmonize the different commonly used cognitive tests from clinical narratives using natural language processing (NLP) methods into categorical AD/ADRD severity.

METHODS

We developed a rule-based NLP pipeline to extract the cognitive tests and biomarkers from clinical narratives in AD/ADRD patients' EHRs. We aggregated the extracted results to the patient level and harmonized the cognitive test scores into severity categories using cutoffs determined based on both relevant literature and domain knowledge of AD/ADRD clinicians.

RESULTS

We identified an AD/ADRD cohort of 48,912 patients from the University of Florida (UF) Health system and identified 7 measurements (6 cognitive tests and 1 biomarker) that are frequently documented in our data. Our NLP pipeline achieved an overall F1-score of 0.9059 across the 7 measurements. Among the 6 cognitive tests, we were able to harmonize 4 cognitive test scores into severity categories, and the population characteristics of patients with different severity were described. We also identified several factors related to the availability of their documentation in EHRs.

CONCLUSION

This study demonstrates that our NLP pipelines can extract cognitive tests and biomarkers of AD/ADRD accurately for downstream studies. Although, the documentation of cognitive tests and biomarkers in EHRs appears to be low, RWD is still an important resource for AD/ADRD research. Nevertheless, providing standardized approach to document cognitive tests and biomarkers in EHRS are also warranted.

Hierarchical Two-Stage Cost-Sensitive Clinical Decision Support System for Screening Prodromal Alzheimer's Disease and Related Dementias

BACKGROUND

The detection of subtle cognitive impairment in a clinical setting is difficult. Because time is a key factor in small clinics and research sites, the brief cognitive assessments that are relied upon often misclassify patients with very mild impairment as normal.

OBJECTIVE

In this study, we seek to identify a parsimonious screening tool in one stage, followed by additional assessments in an optional second stage if additional specificity is desired, tested using a machine learning algorithm capable of being integrated into a clinical decision support system.

METHODS

The best primary stage incorporated measures of short-term memory, executive and visuospatial functioning, and self-reported memory and daily living questions, with a total time of 5 minutes. The best secondary stage incorporated a measure of neurobiology as well as additional cognitive assessment and brief informant report questionnaires, totaling 30 minutes including delayed recall. Combined performance was evaluated using 25 sets of models, trained on 1,181 ADNI participants and tested on 127 patients from a memory clinic.

RESULTS

The 5-minute primary stage was highly sensitive (96.5%) but lacked specificity (34.1%), with an AUC of 87.5% and diagnostic odds ratio of 14.3. The optional secondary stage increased specificity to 58.6%, resulting in an overall AUC of 89.7% using the best model combination of logistic regression and gradient-boosted machine.

CONCLUSION

The primary stage is brief and effective at screening, with the optional two-stage technique further increasing specificity. The hierarchical two-stage technique exhibited similar accuracy but with reduced costs compared to the more common single-stage paradigm.

Interpretable deep learning translation of GWAS and multi-omics findings to identify pathobiology and drug repurposing in Alzheimer's disease

Translating human genetic findings (genome-wide association studies [GWAS]) to pathobiology and therapeutic discovery remains a major challenge for Alzheimer's disease (AD). We present a network topology-based deep learning framework to identify disease-associated genes (NETTAG). We leverage non-coding GWAS loci effects on quantitative trait loci, enhancers and CpG islands, promoter regions, open chromatin, and promoter flanking regions under the protein-protein interactome. Via NETTAG, we identified 156 AD-risk genes enriched in druggable targets. Combining network-based prediction and retrospective case-control observations with 10 million individuals, we identified that usage of four drugs (ibuprofen, gemfibrozil, cholecalciferol, and ceftriaxone) is associated with reduced likelihood of AD incidence. Gemfibrozil (an approved lipid regulator) is significantly associated with 43% reduced risk of AD compared with simvastatin using an active-comparator design (95% confidence interval 0.51-0.63, p < 0.0001). In summary, NETTAG offers a deep learning methodology that utilizes GWAS and multi-genomic findings to identify pathobiology and drug repurposing in AD.

Cognitive measures lacking in EHR prior to dementia or Alzheimer's disease diagnosis

Introduction

The extent that cognitive measures are documented in electronic health records (EHR) is important for quality care and addressing disparities in timely diagnosis of dementia or Alzheimer's disease (AD).

Methods

Analysis of U.S. EHR data to describe the frequency and factors associated with cognitive measures prior to diagnosis of dementia (N = 111,125) or AD (N = 30,203).

Results

Only 11% of dementia patients and 24% of AD patients had a cognitive measure documented in the 5 years prior to diagnosis. Black race, older age, non‐commercial health insurance, lower mean neighborhood income, greater in‐patient stays, and fewer out‐patient visits were associated with lacking cognitive measures.

Discussion

Extensive missing cognitive data and differences in the availability of cognitive measures by race, age, and socioeconomic factors hinder patient care and limit utility of EHR for dementia research. Structured fields and prompts for cognitive data inputs at the point of care may help address these gaps.

A Modified CAIDE Risk Score as a Screening Tool for Cognitive Impairment in Older Adults

BACKGROUND

Although an efficacious dementia-risk score system, Cardiovascular Risk Factors, Aging, and Dementia (CAIDE) was derived using midlife risk factors in a population with low educational attainment that does not reflect today's US population, and requires laboratory biomarkers, which are not always available.

OBJECTIVE

Develop and validate a modified CAIDE (mCAIDE) system and test its ability to predict presence, severity, and etiology of cognitive impairment in older adults.

METHODS

Population consisted of 449 participants in dementia research (N = 230; community sample; 67.9±10.0 years old, 29.6%male, 13.7±4.1 years education) or receiving dementia clinical services (N = 219; clinical sample; 74.3±9.8 years old, 50.2%male, 15.5±2.6 years education). The mCAIDE, which includes self-reported and performance-based rather than blood-derived measures, was developed in the community sample and tested in the independent clinical sample. Validity against Framingham, Hachinski, and CAIDE risk scores was assessed.

RESULTS

Higher mCAIDE quartiles were associated with lower performance on global and domain-specific cognitive tests. Each one-point increase in mCAIDE increased the odds of mild cognitive impairment (MCI) by up to 65%, those of AD by 69%, and those for non-AD dementia by > 85%, with highest scores in cases with vascular etiologies. Being in the highest mCAIDE risk group improved ability to discriminate dementia from MCI and controls and MCI from controls, with a cut-off of ≥7 points offering the highest sensitivity, specificity, and positive and negative predictive values.

CONCLUSION

mCAIDE is a robust indicator of cognitive impairment in community-dwelling seniors, which can discriminate well between dementia severity including MCI versus controls. The mCAIDE may be a valuable tool for case ascertainment in research studies, helping flag primary care patients for cognitive testing, and identify those in need of lifestyle interventions for symptomatic control.

Medical and Social Determinants of Brain Health and Dementia in a Multicultural Community Cohort of Older Adults

BACKGROUND

Socioeconomic status (SES), race, ethnicity, and medical comorbidities may contribute to Alzheimer's disease and related disorders (ADRD) health disparities.

OBJECTIVE

Analyze effects of social and medical determinants on cognition in 374 multicultural older adults participating in a community-based dementia screening program.

METHODS

We used the Montreal Cognitive Assessment (MoCA) and AD8 as measures of cognition, and a 3-way race/ethnicity variable (White, African American, Hispanic) and SES (Hollingshead index) as predictors. Potential contributors to health disparities included: age, sex, education, total medical comorbidities, health self-ratings, and depression. We applied K-means cluster analyses to study medical and social dimension effects on cognitive outcomes.

RESULTS

African Americans and Hispanics had lower SES status and cognitive performance compared with similarly aged Whites. We defined three clusters based on age and SES. Cluster #1 and #3 differed by SES but not age, while cluster #2 was younger with midlevel SES. Cluster #1 experienced the worse health outcomes while cluster #3 had the best health outcomes. Within each cluster, White participants had higher SES and better health outcomes, African Americans had the worst physical performance, and Hispanics had the most depressive symptoms. In cross-cluster comparisons, higher SES led to better health outcomes for all participants.

CONCLUSION

SES may contribute to disparities in access to healthcare services, while race and ethnicity may contribute to disparities in the quality and extent of services received. Our study highlights the need to critically address potential interactions between race, ethnicity, and SES which may better explain disparities in ADRD health outcomes.