Passive Digital Signature for Early Identification of Alzheimer's Disease and Related Dementia

Malaz Boustani, Anthony J. Perkins, Rezaul Karim Khandker, Stephen Duong, Paul R. Dexter, Richard Lipton, Christopher M. Black, Vasu Chandrasekaran, Craig A. Solid, Patrick Monahan

Research output: Contribution to journalArticle

Abstract

OBJECTIVES: Developing scalable strategies for the early identification of Alzheimer's disease and related dementia (ADRD) is important. We aimed to develop a passive digital signature for early identification of ADRD using electronic medical record (EMR) data. DESIGN: A case-control study. SETTING: The Indiana Network for Patient Care (INPC), a regional health information exchange in Indiana. PARTICIPANTS: Patients identified with ADRD and matched controls. MEASUREMENTS: We used data from the INPC that includes structured and unstructured (visit notes, progress notes, medication notes) EMR data. Cases and controls were matched on age, race, and sex. The derivation sample consisted of 10 504 cases and 39 510 controls; the validation sample included 4500 cases and 16 952 controls. We constructed models to identify early 1- to 10-year, 3- to 10-year, and 5- to 10-year ADRD signatures. The analyses included 14 diagnostic risk variables and 10 drug classes in addition to new variables produced from unstructured data (eg, disorientation, confusion, wandering, apraxia, etc). The area under the receiver operating characteristics (AUROC) curve was used to determine the best models. RESULTS: The AUROC curves for the validation samples for the 1- to 10-year, 3- to 10-year, and 5- to 10-year models that used only structured data were.689,.649, and.633, respectively. For the same samples and years, models that used both structured and unstructured data produced AUROC curves of.798,.748, and.704, respectively. Using a cutoff to maximize sensitivity and specificity, the 1- to 10-year, 3- to 10-year, and 5- to 10-year models had sensitivity that ranged from 51% to 62% and specificity that ranged from 80% to 89%. CONCLUSION: EMR-based data provide a targeted and scalable process for early identification of risk of ADRD as an alternative to traditional population screening.

Original languageEnglish (US)
JournalJournal of the American Geriatrics Society
DOIs
StateAccepted/In press - Jan 1 2019

Fingerprint

Dementia
Alzheimer Disease
Electronic Health Records
ROC Curve
Confusion
Patient Care
Apraxias
Case-Control Studies
Sensitivity and Specificity
Pharmaceutical Preparations
Population

Keywords

  • Alzheimer's disease
  • dementia
  • risk factors

ASJC Scopus subject areas

  • Geriatrics and Gerontology

Cite this

Passive Digital Signature for Early Identification of Alzheimer's Disease and Related Dementia. / Boustani, Malaz; Perkins, Anthony J.; Khandker, Rezaul Karim; Duong, Stephen; Dexter, Paul R.; Lipton, Richard; Black, Christopher M.; Chandrasekaran, Vasu; Solid, Craig A.; Monahan, Patrick.

In: Journal of the American Geriatrics Society, 01.01.2019.

Research output: Contribution to journalArticle

Boustani, Malaz ; Perkins, Anthony J. ; Khandker, Rezaul Karim ; Duong, Stephen ; Dexter, Paul R. ; Lipton, Richard ; Black, Christopher M. ; Chandrasekaran, Vasu ; Solid, Craig A. ; Monahan, Patrick. / Passive Digital Signature for Early Identification of Alzheimer's Disease and Related Dementia. In: Journal of the American Geriatrics Society. 2019.
@article{b91ee5a903c84482963be2482963b311,
title = "Passive Digital Signature for Early Identification of Alzheimer's Disease and Related Dementia",
abstract = "OBJECTIVES: Developing scalable strategies for the early identification of Alzheimer's disease and related dementia (ADRD) is important. We aimed to develop a passive digital signature for early identification of ADRD using electronic medical record (EMR) data. DESIGN: A case-control study. SETTING: The Indiana Network for Patient Care (INPC), a regional health information exchange in Indiana. PARTICIPANTS: Patients identified with ADRD and matched controls. MEASUREMENTS: We used data from the INPC that includes structured and unstructured (visit notes, progress notes, medication notes) EMR data. Cases and controls were matched on age, race, and sex. The derivation sample consisted of 10 504 cases and 39 510 controls; the validation sample included 4500 cases and 16 952 controls. We constructed models to identify early 1- to 10-year, 3- to 10-year, and 5- to 10-year ADRD signatures. The analyses included 14 diagnostic risk variables and 10 drug classes in addition to new variables produced from unstructured data (eg, disorientation, confusion, wandering, apraxia, etc). The area under the receiver operating characteristics (AUROC) curve was used to determine the best models. RESULTS: The AUROC curves for the validation samples for the 1- to 10-year, 3- to 10-year, and 5- to 10-year models that used only structured data were.689,.649, and.633, respectively. For the same samples and years, models that used both structured and unstructured data produced AUROC curves of.798,.748, and.704, respectively. Using a cutoff to maximize sensitivity and specificity, the 1- to 10-year, 3- to 10-year, and 5- to 10-year models had sensitivity that ranged from 51{\%} to 62{\%} and specificity that ranged from 80{\%} to 89{\%}. CONCLUSION: EMR-based data provide a targeted and scalable process for early identification of risk of ADRD as an alternative to traditional population screening.",
keywords = "Alzheimer's disease, dementia, risk factors",
author = "Malaz Boustani and Perkins, {Anthony J.} and Khandker, {Rezaul Karim} and Stephen Duong and Dexter, {Paul R.} and Richard Lipton and Black, {Christopher M.} and Vasu Chandrasekaran and Solid, {Craig A.} and Patrick Monahan",
year = "2019",
month = "1",
day = "1",
doi = "10.1111/jgs.16218",
language = "English (US)",
journal = "Journal of the American Geriatrics Society",
issn = "0002-8614",
publisher = "Wiley-Blackwell",

}

TY - JOUR

T1 - Passive Digital Signature for Early Identification of Alzheimer's Disease and Related Dementia

AU - Boustani, Malaz

AU - Perkins, Anthony J.

AU - Khandker, Rezaul Karim

AU - Duong, Stephen

AU - Dexter, Paul R.

AU - Lipton, Richard

AU - Black, Christopher M.

AU - Chandrasekaran, Vasu

AU - Solid, Craig A.

AU - Monahan, Patrick

PY - 2019/1/1

Y1 - 2019/1/1

N2 - OBJECTIVES: Developing scalable strategies for the early identification of Alzheimer's disease and related dementia (ADRD) is important. We aimed to develop a passive digital signature for early identification of ADRD using electronic medical record (EMR) data. DESIGN: A case-control study. SETTING: The Indiana Network for Patient Care (INPC), a regional health information exchange in Indiana. PARTICIPANTS: Patients identified with ADRD and matched controls. MEASUREMENTS: We used data from the INPC that includes structured and unstructured (visit notes, progress notes, medication notes) EMR data. Cases and controls were matched on age, race, and sex. The derivation sample consisted of 10 504 cases and 39 510 controls; the validation sample included 4500 cases and 16 952 controls. We constructed models to identify early 1- to 10-year, 3- to 10-year, and 5- to 10-year ADRD signatures. The analyses included 14 diagnostic risk variables and 10 drug classes in addition to new variables produced from unstructured data (eg, disorientation, confusion, wandering, apraxia, etc). The area under the receiver operating characteristics (AUROC) curve was used to determine the best models. RESULTS: The AUROC curves for the validation samples for the 1- to 10-year, 3- to 10-year, and 5- to 10-year models that used only structured data were.689,.649, and.633, respectively. For the same samples and years, models that used both structured and unstructured data produced AUROC curves of.798,.748, and.704, respectively. Using a cutoff to maximize sensitivity and specificity, the 1- to 10-year, 3- to 10-year, and 5- to 10-year models had sensitivity that ranged from 51% to 62% and specificity that ranged from 80% to 89%. CONCLUSION: EMR-based data provide a targeted and scalable process for early identification of risk of ADRD as an alternative to traditional population screening.

AB - OBJECTIVES: Developing scalable strategies for the early identification of Alzheimer's disease and related dementia (ADRD) is important. We aimed to develop a passive digital signature for early identification of ADRD using electronic medical record (EMR) data. DESIGN: A case-control study. SETTING: The Indiana Network for Patient Care (INPC), a regional health information exchange in Indiana. PARTICIPANTS: Patients identified with ADRD and matched controls. MEASUREMENTS: We used data from the INPC that includes structured and unstructured (visit notes, progress notes, medication notes) EMR data. Cases and controls were matched on age, race, and sex. The derivation sample consisted of 10 504 cases and 39 510 controls; the validation sample included 4500 cases and 16 952 controls. We constructed models to identify early 1- to 10-year, 3- to 10-year, and 5- to 10-year ADRD signatures. The analyses included 14 diagnostic risk variables and 10 drug classes in addition to new variables produced from unstructured data (eg, disorientation, confusion, wandering, apraxia, etc). The area under the receiver operating characteristics (AUROC) curve was used to determine the best models. RESULTS: The AUROC curves for the validation samples for the 1- to 10-year, 3- to 10-year, and 5- to 10-year models that used only structured data were.689,.649, and.633, respectively. For the same samples and years, models that used both structured and unstructured data produced AUROC curves of.798,.748, and.704, respectively. Using a cutoff to maximize sensitivity and specificity, the 1- to 10-year, 3- to 10-year, and 5- to 10-year models had sensitivity that ranged from 51% to 62% and specificity that ranged from 80% to 89%. CONCLUSION: EMR-based data provide a targeted and scalable process for early identification of risk of ADRD as an alternative to traditional population screening.

KW - Alzheimer's disease

KW - dementia

KW - risk factors

UR - http://www.scopus.com/inward/record.url?scp=85075790761&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85075790761&partnerID=8YFLogxK

U2 - 10.1111/jgs.16218

DO - 10.1111/jgs.16218

M3 - Article

C2 - 31784987

AN - SCOPUS:85075790761

JO - Journal of the American Geriatrics Society

JF - Journal of the American Geriatrics Society

SN - 0002-8614

ER -