Unique Clinical Language Patterns Among Expert Vestibular Providers Can Predict Vestibular Diagnoses. Otol Neurotol 2018 Oct;39(9):1163-1171
Date
08/07/2018Pubmed ID
30080764DOI
10.1097/MAO.0000000000001930Scopus ID
2-s2.0-85056607740 (requires institutional sign-in at Scopus site) 5 CitationsAbstract
OBJECTIVE: To identify novel language usage by expert providers predictive of specific vestibular conditions.
STUDY DESIGN: Retrospective chart review and natural language processing. Level IV.
SETTING: Tertiary referral center.
PATIENTS: Patients seen for vestibular complaint.
INTERVENTION(S): Natural language processing and machine learning analyses of semantic and syntactic patterns in clinical documentation from vestibular patients.
MAIN OUTCOME MEASURE: Accuracy of Naïve Bayes predictive models correlating language usage with clinical diagnoses.
RESULTS: Natural language analyses on 866 physician-generated histories from vestibular patients found 3,286 unique examples of language usage of which 614 were used 10 or greater times. The top 15 semantic types represented only 11% of all Unified Medical Language System semantic types but covered 86% of language used in vestibular patient histories. Naïve Bayes machine learning algorithms on a subset of 255 notes representing benign paroxysmal positional vertigo, vestibular migraine, anxiety-related dizziness and central dizziness generated strong predictive models showing an average sensitivity rate of 93.4% and a specificity rate of 98.2%. A binary model for assessing whether a subject had a specific diagnosis or not had an average AUC for the receiver operating characteristic curves of .995 across all conditions.
CONCLUSIONS: These results indicate that expert providers utilize unique language patterns in vestibular notes that are highly conserved. These patterns have strong predictive power toward specific vestibular diagnoses. Such language elements can provide a simple vocabulary to aid nonexpert providers in formulating a differential diagnosis. They can also be incorporated into clinical decision support systems to facilitate accurate vestibular diagnosis in ambulatory settings.
Author List
Luo J, Erbe C, Friedland DRAuthors
David R. Friedland MD Associate Director, Director, Chief, Professor in the Otolaryngology department at Medical College of WisconsinJake Luo Ph.D. Associate Professor; Director, Center for Biomedical Data and Language Processing (BioDLP) in the Health Informatics & Administration department at University of Wisconsin - Milwaukee
MESH terms used to index this publication - Major topics in bold
AdultAlgorithms
Bayes Theorem
Female
Humans
Male
Medical History Taking
Middle Aged
Natural Language Processing
Otolaryngologists
Practice Patterns, Physicians'
ROC Curve
Retrospective Studies
Sensitivity and Specificity
Vestibular Diseases
