Evidence-Based Assessment From Simple Clinical Judgments to Statistical Learning: Evaluating a Range of Options Using Pediatric Bipolar Disorder as a Diagnostic Challenge

Eric A. Youngstrom, Tate F. Halverson, Jennifer K. Youngstrom, Oliver Lindhiem, Robert L Findling

Research output: Contribution to journalArticle

Abstract

Reliability of clinical diagnoses is often low. There are many algorithms that could improve diagnostic accuracy, and statistical learning is becoming popular. Using pediatric bipolar disorder as a clinically challenging example, we evaluated a series of increasingly complex models ranging from simple screening to a supervised LASSO (least absolute shrinkage and selection operation) regression in a large (N = 550) academic clinic sample. We then externally validated models in a community clinic (N = 511) with the same candidate predictors and semistructured interview diagnoses, providing high methodological consistency; the clinics also had substantially different demography and referral patterns. Models performed well according to internal validation metrics. Complex models degraded rapidly when externally validated. Naive Bayesian and logistic models concentrating on predictors identified in prior meta-analyses tied or bettered LASSO models when externally validated. Implementing these methods would improve clinical diagnostic performance. Statistical learning research should continue to invest in high-quality indicators and diagnoses to supervise model training.

Original languageEnglish (US)
Pages (from-to)243-265
Number of pages23
JournalClinical Psychological Science
Volume6
Issue number2
DOIs
StatePublished - Mar 1 2018

Fingerprint

Bipolar Disorder
Learning
Pediatrics
Meta-Analysis
Referral and Consultation
Logistic Models
Demography
Interviews
Research

Keywords

  • bipolar disorder
  • diagnostic accuracy
  • evidence-based assessment
  • open data
  • sensitivity and specificity

ASJC Scopus subject areas

  • Clinical Psychology

Cite this

Evidence-Based Assessment From Simple Clinical Judgments to Statistical Learning : Evaluating a Range of Options Using Pediatric Bipolar Disorder as a Diagnostic Challenge. / Youngstrom, Eric A.; Halverson, Tate F.; Youngstrom, Jennifer K.; Lindhiem, Oliver; Findling, Robert L.

In: Clinical Psychological Science, Vol. 6, No. 2, 01.03.2018, p. 243-265.

Research output: Contribution to journalArticle

@article{9326cd86fd084dad8efcde36c4e96af3,
title = "Evidence-Based Assessment From Simple Clinical Judgments to Statistical Learning: Evaluating a Range of Options Using Pediatric Bipolar Disorder as a Diagnostic Challenge",
abstract = "Reliability of clinical diagnoses is often low. There are many algorithms that could improve diagnostic accuracy, and statistical learning is becoming popular. Using pediatric bipolar disorder as a clinically challenging example, we evaluated a series of increasingly complex models ranging from simple screening to a supervised LASSO (least absolute shrinkage and selection operation) regression in a large (N = 550) academic clinic sample. We then externally validated models in a community clinic (N = 511) with the same candidate predictors and semistructured interview diagnoses, providing high methodological consistency; the clinics also had substantially different demography and referral patterns. Models performed well according to internal validation metrics. Complex models degraded rapidly when externally validated. Naive Bayesian and logistic models concentrating on predictors identified in prior meta-analyses tied or bettered LASSO models when externally validated. Implementing these methods would improve clinical diagnostic performance. Statistical learning research should continue to invest in high-quality indicators and diagnoses to supervise model training.",
keywords = "bipolar disorder, diagnostic accuracy, evidence-based assessment, open data, sensitivity and specificity",
author = "Youngstrom, {Eric A.} and Halverson, {Tate F.} and Youngstrom, {Jennifer K.} and Oliver Lindhiem and Findling, {Robert L}",
year = "2018",
month = "3",
day = "1",
doi = "10.1177/2167702617741845",
language = "English (US)",
volume = "6",
pages = "243--265",
journal = "Clinical Psychological Science",
issn = "2167-7026",
publisher = "SAGE Publications Inc.",
number = "2",

}

TY - JOUR

T1 - Evidence-Based Assessment From Simple Clinical Judgments to Statistical Learning

T2 - Evaluating a Range of Options Using Pediatric Bipolar Disorder as a Diagnostic Challenge

AU - Youngstrom, Eric A.

AU - Halverson, Tate F.

AU - Youngstrom, Jennifer K.

AU - Lindhiem, Oliver

AU - Findling, Robert L

PY - 2018/3/1

Y1 - 2018/3/1

N2 - Reliability of clinical diagnoses is often low. There are many algorithms that could improve diagnostic accuracy, and statistical learning is becoming popular. Using pediatric bipolar disorder as a clinically challenging example, we evaluated a series of increasingly complex models ranging from simple screening to a supervised LASSO (least absolute shrinkage and selection operation) regression in a large (N = 550) academic clinic sample. We then externally validated models in a community clinic (N = 511) with the same candidate predictors and semistructured interview diagnoses, providing high methodological consistency; the clinics also had substantially different demography and referral patterns. Models performed well according to internal validation metrics. Complex models degraded rapidly when externally validated. Naive Bayesian and logistic models concentrating on predictors identified in prior meta-analyses tied or bettered LASSO models when externally validated. Implementing these methods would improve clinical diagnostic performance. Statistical learning research should continue to invest in high-quality indicators and diagnoses to supervise model training.

AB - Reliability of clinical diagnoses is often low. There are many algorithms that could improve diagnostic accuracy, and statistical learning is becoming popular. Using pediatric bipolar disorder as a clinically challenging example, we evaluated a series of increasingly complex models ranging from simple screening to a supervised LASSO (least absolute shrinkage and selection operation) regression in a large (N = 550) academic clinic sample. We then externally validated models in a community clinic (N = 511) with the same candidate predictors and semistructured interview diagnoses, providing high methodological consistency; the clinics also had substantially different demography and referral patterns. Models performed well according to internal validation metrics. Complex models degraded rapidly when externally validated. Naive Bayesian and logistic models concentrating on predictors identified in prior meta-analyses tied or bettered LASSO models when externally validated. Implementing these methods would improve clinical diagnostic performance. Statistical learning research should continue to invest in high-quality indicators and diagnoses to supervise model training.

KW - bipolar disorder

KW - diagnostic accuracy

KW - evidence-based assessment

KW - open data

KW - sensitivity and specificity

UR - http://www.scopus.com/inward/record.url?scp=85042481385&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85042481385&partnerID=8YFLogxK

U2 - 10.1177/2167702617741845

DO - 10.1177/2167702617741845

M3 - Article

C2 - 30263876

AN - SCOPUS:85042481385

VL - 6

SP - 243

EP - 265

JO - Clinical Psychological Science

JF - Clinical Psychological Science

SN - 2167-7026

IS - 2

ER -