Patient-tailored prioritization for a pediatric care decision support system through machine learning

Jeffrey G. Klann, Vibha Anand, Stephen M. Downs

Research output: Contribution to journalArticle

12 Citations (Scopus)

Abstract

Objective Over 8 years, we have developed an innovative computer decision support system that improves appropriate delivery of pediatric screening and care. This system employs a guidelines evaluation engine using data from the electronic health record (EHR) and input from patients and caregivers. Because guideline recommendations typically exceed the scope of one visit, the engine uses a static prioritization scheme to select recommendations. Here we extend an earlier idea to create patient-tailored prioritization. Materials and methods We used Bayesian structure learning to build networks of association among previously collected data from our decision support system. Using area under the receiver-operating characteristic curve (AUC) as a measure of discriminability (a sine qua non for expected value calculations needed for prioritization), we performed a structural analysis of variables with high AUC on a test set. Our source data included 177 variables for 29 402 patients. Results The method produced a network model containing 78 screening questions and anticipatory guidance (107 variables total). Average AUC was 0.65, which is sufficient for prioritization depending on factors such as population prevalence. Structure analysis of seven highly predictive variables reveals both face-validity (related nodes are connected) and non-intuitive relationships. Discussion We demonstrate the ability of a Bayesian structure learning method to 'phenotype the population' seen in our primary care pediatric clinics. The resulting network can be used to produce patient-tailored posterior probabilities that can be used to prioritize content based on the patient's current circumstances. Conclusions This study demonstrates the feasibility of EHR-driven population phenotyping for patient-tailored prioritization of pediatric preventive care services.

Original languageEnglish (US)
Pages (from-to)e267-e274
JournalJournal of the American Medical Informatics Association
Volume20
Issue numberE2
DOIs
StatePublished - Dec 19 2013

Fingerprint

Pediatrics
Area Under Curve
Electronic Health Records
Learning
Guidelines
Population
Preventive Medicine
Aptitude
Information Storage and Retrieval
Feasibility Studies
Reproducibility of Results
ROC Curve
Caregivers
Machine Learning
Primary Health Care
Phenotype

ASJC Scopus subject areas

  • Health Informatics

Cite this

Patient-tailored prioritization for a pediatric care decision support system through machine learning. / Klann, Jeffrey G.; Anand, Vibha; Downs, Stephen M.

In: Journal of the American Medical Informatics Association, Vol. 20, No. E2, 19.12.2013, p. e267-e274.

Research output: Contribution to journalArticle

@article{94044fd7fe2f4b70b1329ef0e58bf96f,
title = "Patient-tailored prioritization for a pediatric care decision support system through machine learning",
abstract = "Objective Over 8 years, we have developed an innovative computer decision support system that improves appropriate delivery of pediatric screening and care. This system employs a guidelines evaluation engine using data from the electronic health record (EHR) and input from patients and caregivers. Because guideline recommendations typically exceed the scope of one visit, the engine uses a static prioritization scheme to select recommendations. Here we extend an earlier idea to create patient-tailored prioritization. Materials and methods We used Bayesian structure learning to build networks of association among previously collected data from our decision support system. Using area under the receiver-operating characteristic curve (AUC) as a measure of discriminability (a sine qua non for expected value calculations needed for prioritization), we performed a structural analysis of variables with high AUC on a test set. Our source data included 177 variables for 29 402 patients. Results The method produced a network model containing 78 screening questions and anticipatory guidance (107 variables total). Average AUC was 0.65, which is sufficient for prioritization depending on factors such as population prevalence. Structure analysis of seven highly predictive variables reveals both face-validity (related nodes are connected) and non-intuitive relationships. Discussion We demonstrate the ability of a Bayesian structure learning method to 'phenotype the population' seen in our primary care pediatric clinics. The resulting network can be used to produce patient-tailored posterior probabilities that can be used to prioritize content based on the patient's current circumstances. Conclusions This study demonstrates the feasibility of EHR-driven population phenotyping for patient-tailored prioritization of pediatric preventive care services.",
author = "Klann, {Jeffrey G.} and Vibha Anand and Downs, {Stephen M.}",
year = "2013",
month = "12",
day = "19",
doi = "10.1136/amiajnl-2013-001865",
language = "English (US)",
volume = "20",
pages = "e267--e274",
journal = "Journal of the American Medical Informatics Association : JAMIA",
issn = "1067-5027",
publisher = "Oxford University Press",
number = "E2",

}

TY - JOUR

T1 - Patient-tailored prioritization for a pediatric care decision support system through machine learning

AU - Klann, Jeffrey G.

AU - Anand, Vibha

AU - Downs, Stephen M.

PY - 2013/12/19

Y1 - 2013/12/19

N2 - Objective Over 8 years, we have developed an innovative computer decision support system that improves appropriate delivery of pediatric screening and care. This system employs a guidelines evaluation engine using data from the electronic health record (EHR) and input from patients and caregivers. Because guideline recommendations typically exceed the scope of one visit, the engine uses a static prioritization scheme to select recommendations. Here we extend an earlier idea to create patient-tailored prioritization. Materials and methods We used Bayesian structure learning to build networks of association among previously collected data from our decision support system. Using area under the receiver-operating characteristic curve (AUC) as a measure of discriminability (a sine qua non for expected value calculations needed for prioritization), we performed a structural analysis of variables with high AUC on a test set. Our source data included 177 variables for 29 402 patients. Results The method produced a network model containing 78 screening questions and anticipatory guidance (107 variables total). Average AUC was 0.65, which is sufficient for prioritization depending on factors such as population prevalence. Structure analysis of seven highly predictive variables reveals both face-validity (related nodes are connected) and non-intuitive relationships. Discussion We demonstrate the ability of a Bayesian structure learning method to 'phenotype the population' seen in our primary care pediatric clinics. The resulting network can be used to produce patient-tailored posterior probabilities that can be used to prioritize content based on the patient's current circumstances. Conclusions This study demonstrates the feasibility of EHR-driven population phenotyping for patient-tailored prioritization of pediatric preventive care services.

AB - Objective Over 8 years, we have developed an innovative computer decision support system that improves appropriate delivery of pediatric screening and care. This system employs a guidelines evaluation engine using data from the electronic health record (EHR) and input from patients and caregivers. Because guideline recommendations typically exceed the scope of one visit, the engine uses a static prioritization scheme to select recommendations. Here we extend an earlier idea to create patient-tailored prioritization. Materials and methods We used Bayesian structure learning to build networks of association among previously collected data from our decision support system. Using area under the receiver-operating characteristic curve (AUC) as a measure of discriminability (a sine qua non for expected value calculations needed for prioritization), we performed a structural analysis of variables with high AUC on a test set. Our source data included 177 variables for 29 402 patients. Results The method produced a network model containing 78 screening questions and anticipatory guidance (107 variables total). Average AUC was 0.65, which is sufficient for prioritization depending on factors such as population prevalence. Structure analysis of seven highly predictive variables reveals both face-validity (related nodes are connected) and non-intuitive relationships. Discussion We demonstrate the ability of a Bayesian structure learning method to 'phenotype the population' seen in our primary care pediatric clinics. The resulting network can be used to produce patient-tailored posterior probabilities that can be used to prioritize content based on the patient's current circumstances. Conclusions This study demonstrates the feasibility of EHR-driven population phenotyping for patient-tailored prioritization of pediatric preventive care services.

UR - http://www.scopus.com/inward/record.url?scp=84890320768&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84890320768&partnerID=8YFLogxK

U2 - 10.1136/amiajnl-2013-001865

DO - 10.1136/amiajnl-2013-001865

M3 - Article

C2 - 23886921

AN - SCOPUS:84890320768

VL - 20

SP - e267-e274

JO - Journal of the American Medical Informatics Association : JAMIA

JF - Journal of the American Medical Informatics Association : JAMIA

SN - 1067-5027

IS - E2

ER -