Data‐driven
            discovery of probable Alzheimer's disease and related dementia subphenotypes using electronic health records

IntroductionEffects of early life conditions on Alzheimer’s disease (AD) and related dementia (RD) risk have been hypothesized. However, prospective study is potentially cost prohibitive. Retrospective studies using routinely collected health records in large cohorts may be a feasible way to carry out such research, but diagnostic accuracy should be determined. Objectives and ApproachWe aim to determine accuracy of AD/RD diagnoses in electronic health records (EHR) (inpatient, ambulatory surgery, and Medicare) and death certificates (DC) compared to gold standard. The Cache County Study on Memory in Aging (CACHE, 1995–2008) enrolled 90% of the county’s residents age ≥ 65 years (N=5092). Over the course of 12 years/4 triennial waves of thorough dementia ascertainment, 942 persons (18.5%) were identified with dementia. Prevalence of AD or AD comorbid with other dementia (AD mixed) was 12.8% and for RD alone, 5.7%. We used the Utah Population Database, linking EMR/DCs (1995–2008) to CACHE participants (98% linkage). ResultsThe prevalence of AD/AD mixed and RD in EHR/DCs was 12.2% and 35.8%. Among linked CACHE participants diagnosed with AD or AD mixed (n=628), 505 (80%) were captured by EHR/DCs as having some form of dementia (AD, AD mixed, or RD) with 301 (60%) correctly classified as having AD or AD mixed. Among those with RD (n=399), 275 (69%) were captured by EHR/DCs as having some form of dementia, with 163 (60%) correctly classified as having RD. Sensitivity, specificity, positive and negative predictive values, and area under the curve (AUC) were 48%, 93%, 49%, 93%, and 0.70 for AD or AD mixed; and 67%, 67%, 15%, 96%, and 0.67 for RD. Overall dementia agreement between CACHE diagnoses and EHR/DCs was fair (Cohen's κ = 0.34). Conclusion/ImplicationsIn this will characterized cohort, routinely collected health record diagnoses of AD/AD mixed and RD have only fair correlation with carefully phenotyped diagnoses. Determining additional features of a person’s medical record that may be predictive of AD/RD via formal classification modeling is warranted.

Download Full-text

Identifying and evaluating clinical subtypes of Alzheimer’s disease in care electronic health records using unsupervised machine learning

BMC Medical Informatics and Decision Making ◽

10.1186/s12911-021-01693-6 ◽

2021 ◽

Vol 21 (1) ◽

Author(s):

Nonie Alexander ◽

Daniel C. Alexander ◽

Frederik Barkhof ◽

Spiros Denaxas

Keyword(s):

Alzheimer’S Disease ◽

Alzheimer's Disease ◽

Electronic Health Records ◽

Latent Class Analysis ◽

Latent Class ◽

Affinity Propagation ◽

Class Analysis ◽

Clustering Methods ◽

Health Records ◽

Electronic Health

Abstract Background Alzheimer’s disease (AD) is a highly heterogeneous disease with diverse trajectories and outcomes observed in clinical populations. Understanding this heterogeneity can enable better treatment, prognosis and disease management. Studies to date have mainly used imaging or cognition data and have been limited in terms of data breadth and sample size. Here we examine the clinical heterogeneity of Alzheimer's disease patients using electronic health records (EHR) to identify and characterise disease subgroups using multiple clustering methods, identifying clusters which are clinically actionable. Methods We identified AD patients in primary care EHR from the Clinical Practice Research Datalink (CPRD) using a previously validated rule-based phenotyping algorithm. We extracted and included a range of comorbidities, symptoms and demographic features as patient features. We evaluated four different clustering methods (k-means, kernel k-means, affinity propagation and latent class analysis) to cluster Alzheimer’s disease patients. We compared clusters on clinically relevant outcomes and evaluated each method using measures of cluster structure, stability, efficiency of outcome prediction and replicability in external data sets. Results We identified 7,913 AD patients, with a mean age of 82 and 66.2% female. We included 21 features in our analysis. We observed 5, 2, 5 and 6 clusters in k-means, kernel k-means, affinity propagation and latent class analysis respectively. K-means was found to produce the most consistent results based on four evaluative measures. We discovered a consistent cluster found in three of the four methods composed of predominantly female, younger disease onset (43% between ages 42–73) diagnosed with depression and anxiety, with a quicker rate of progression compared to the average across other clusters. Conclusion Each clustering approach produced substantially different clusters and K-Means performed the best out of the four methods based on the four evaluative criteria. However, the consistent appearance of one particular cluster across three of the four methods potentially suggests the presence of a distinct disease subtype that merits further exploration. Our study underlines the variability of the results obtained from different clustering approaches and the importance of systematically evaluating different approaches for identifying disease subtypes in complex EHR.

Download Full-text