Using electronic medical records to enable large-scale studies in psychiatry

Treatment resistant depression as a model

R. H. Perlis, D. V. Iosifescu, V. M. Castro, S. N. Murphy, V. S. Gainer, Jessica Minnier, T. Cai, S. Goryachev, Q. Zeng, P. J. Gallagher, M. Fava, J. B. Weilburg, S. E. Churchill, I. S. Kohane, J. W. Smoller

Research output: Contribution to journalArticle

87 Citations (Scopus)

Abstract

Background Electronic medical records (EMR) provide a unique opportunity for efficient, large-scale clinical investigation in psychiatry. However, such studies will require development of tools to define treatment outcome.Method Natural language processing (NLP) was applied to classify notes from 127 504 patients with a billing diagnosis of major depressive disorder, drawn from out-patient psychiatry practices affiliated with multiple, large New England hospitals. Classifications were compared with results using billing data (ICD-9 codes) alone and to a clinical gold standard based on chart review by a panel of senior clinicians. These cross-sectional classifications were then used to define longitudinal treatment outcomes, which were compared with a clinician-rated gold standard.Results Models incorporating NLP were superior to those relying on billing data alone for classifying current mood state (area under receiver operating characteristic curve of 0.85-0.88 v. 0.54-0.55). When these cross-sectional visits were integrated to define longitudinal outcomes and incorporate treatment data, 15% of the cohort remitted with a single antidepressant treatment, while 13% were identified as failing to remit despite at least two antidepressant trials. Non-remitting patients were more likely to be non-Caucasian (p

Original languageEnglish (US)
Pages (from-to)41-50
Number of pages10
JournalPsychological Medicine
Volume42
Issue number1
DOIs
StatePublished - Jan 2012
Externally publishedYes

Fingerprint

Treatment-Resistant Depressive Disorder
Electronic Health Records
Natural Language Processing
Psychiatry
International Classification of Diseases
Antidepressive Agents
New England
Major Depressive Disorder
ROC Curve
Outpatients
Therapeutics

Keywords

  • Antidepressant
  • classification
  • machine learning
  • natural language processing
  • remission
  • treatment resistant depression

ASJC Scopus subject areas

  • Psychiatry and Mental health
  • Applied Psychology

Cite this

Using electronic medical records to enable large-scale studies in psychiatry : Treatment resistant depression as a model. / Perlis, R. H.; Iosifescu, D. V.; Castro, V. M.; Murphy, S. N.; Gainer, V. S.; Minnier, Jessica; Cai, T.; Goryachev, S.; Zeng, Q.; Gallagher, P. J.; Fava, M.; Weilburg, J. B.; Churchill, S. E.; Kohane, I. S.; Smoller, J. W.

In: Psychological Medicine, Vol. 42, No. 1, 01.2012, p. 41-50.

Research output: Contribution to journalArticle

Perlis, RH, Iosifescu, DV, Castro, VM, Murphy, SN, Gainer, VS, Minnier, J, Cai, T, Goryachev, S, Zeng, Q, Gallagher, PJ, Fava, M, Weilburg, JB, Churchill, SE, Kohane, IS & Smoller, JW 2012, 'Using electronic medical records to enable large-scale studies in psychiatry: Treatment resistant depression as a model', Psychological Medicine, vol. 42, no. 1, pp. 41-50. https://doi.org/10.1017/S0033291711000997
Perlis, R. H. ; Iosifescu, D. V. ; Castro, V. M. ; Murphy, S. N. ; Gainer, V. S. ; Minnier, Jessica ; Cai, T. ; Goryachev, S. ; Zeng, Q. ; Gallagher, P. J. ; Fava, M. ; Weilburg, J. B. ; Churchill, S. E. ; Kohane, I. S. ; Smoller, J. W. / Using electronic medical records to enable large-scale studies in psychiatry : Treatment resistant depression as a model. In: Psychological Medicine. 2012 ; Vol. 42, No. 1. pp. 41-50.
@article{b853321b2b5f40cf9543ae109fba4a65,
title = "Using electronic medical records to enable large-scale studies in psychiatry: Treatment resistant depression as a model",
abstract = "Background Electronic medical records (EMR) provide a unique opportunity for efficient, large-scale clinical investigation in psychiatry. However, such studies will require development of tools to define treatment outcome.Method Natural language processing (NLP) was applied to classify notes from 127 504 patients with a billing diagnosis of major depressive disorder, drawn from out-patient psychiatry practices affiliated with multiple, large New England hospitals. Classifications were compared with results using billing data (ICD-9 codes) alone and to a clinical gold standard based on chart review by a panel of senior clinicians. These cross-sectional classifications were then used to define longitudinal treatment outcomes, which were compared with a clinician-rated gold standard.Results Models incorporating NLP were superior to those relying on billing data alone for classifying current mood state (area under receiver operating characteristic curve of 0.85-0.88 v. 0.54-0.55). When these cross-sectional visits were integrated to define longitudinal outcomes and incorporate treatment data, 15{\%} of the cohort remitted with a single antidepressant treatment, while 13{\%} were identified as failing to remit despite at least two antidepressant trials. Non-remitting patients were more likely to be non-Caucasian (p",
keywords = "Antidepressant, classification, machine learning, natural language processing, remission, treatment resistant depression",
author = "Perlis, {R. H.} and Iosifescu, {D. V.} and Castro, {V. M.} and Murphy, {S. N.} and Gainer, {V. S.} and Jessica Minnier and T. Cai and S. Goryachev and Q. Zeng and Gallagher, {P. J.} and M. Fava and Weilburg, {J. B.} and Churchill, {S. E.} and Kohane, {I. S.} and Smoller, {J. W.}",
year = "2012",
month = "1",
doi = "10.1017/S0033291711000997",
language = "English (US)",
volume = "42",
pages = "41--50",
journal = "Psychological Medicine",
issn = "0033-2917",
publisher = "Cambridge University Press",
number = "1",

}

TY - JOUR

T1 - Using electronic medical records to enable large-scale studies in psychiatry

T2 - Treatment resistant depression as a model

AU - Perlis, R. H.

AU - Iosifescu, D. V.

AU - Castro, V. M.

AU - Murphy, S. N.

AU - Gainer, V. S.

AU - Minnier, Jessica

AU - Cai, T.

AU - Goryachev, S.

AU - Zeng, Q.

AU - Gallagher, P. J.

AU - Fava, M.

AU - Weilburg, J. B.

AU - Churchill, S. E.

AU - Kohane, I. S.

AU - Smoller, J. W.

PY - 2012/1

Y1 - 2012/1

N2 - Background Electronic medical records (EMR) provide a unique opportunity for efficient, large-scale clinical investigation in psychiatry. However, such studies will require development of tools to define treatment outcome.Method Natural language processing (NLP) was applied to classify notes from 127 504 patients with a billing diagnosis of major depressive disorder, drawn from out-patient psychiatry practices affiliated with multiple, large New England hospitals. Classifications were compared with results using billing data (ICD-9 codes) alone and to a clinical gold standard based on chart review by a panel of senior clinicians. These cross-sectional classifications were then used to define longitudinal treatment outcomes, which were compared with a clinician-rated gold standard.Results Models incorporating NLP were superior to those relying on billing data alone for classifying current mood state (area under receiver operating characteristic curve of 0.85-0.88 v. 0.54-0.55). When these cross-sectional visits were integrated to define longitudinal outcomes and incorporate treatment data, 15% of the cohort remitted with a single antidepressant treatment, while 13% were identified as failing to remit despite at least two antidepressant trials. Non-remitting patients were more likely to be non-Caucasian (p

AB - Background Electronic medical records (EMR) provide a unique opportunity for efficient, large-scale clinical investigation in psychiatry. However, such studies will require development of tools to define treatment outcome.Method Natural language processing (NLP) was applied to classify notes from 127 504 patients with a billing diagnosis of major depressive disorder, drawn from out-patient psychiatry practices affiliated with multiple, large New England hospitals. Classifications were compared with results using billing data (ICD-9 codes) alone and to a clinical gold standard based on chart review by a panel of senior clinicians. These cross-sectional classifications were then used to define longitudinal treatment outcomes, which were compared with a clinician-rated gold standard.Results Models incorporating NLP were superior to those relying on billing data alone for classifying current mood state (area under receiver operating characteristic curve of 0.85-0.88 v. 0.54-0.55). When these cross-sectional visits were integrated to define longitudinal outcomes and incorporate treatment data, 15% of the cohort remitted with a single antidepressant treatment, while 13% were identified as failing to remit despite at least two antidepressant trials. Non-remitting patients were more likely to be non-Caucasian (p

KW - Antidepressant

KW - classification

KW - machine learning

KW - natural language processing

KW - remission

KW - treatment resistant depression

UR - http://www.scopus.com/inward/record.url?scp=82655165845&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=82655165845&partnerID=8YFLogxK

U2 - 10.1017/S0033291711000997

DO - 10.1017/S0033291711000997

M3 - Article

VL - 42

SP - 41

EP - 50

JO - Psychological Medicine

JF - Psychological Medicine

SN - 0033-2917

IS - 1

ER -