Assessing the similarity of surface linguistic features related to epilepsy across pediatric hospitals

Abstract

ObjectiveThe constant progress in computational linguistic methods provides amazing opportunities for discovering information in clinical text and enables the clinical scientist to explore novel approaches to care. However, these new approaches need evaluation. We describe an automated system to compare descriptions of epilepsy patients at three different organizations: Cincinnati Children’s Hospital, the Children’s Hospital Colorado, and the Children’s Hospital of Philadelphia. To our knowledge, there have been no similar previous studies.Materials and methodsIn this work, a support vector machine (SVM)-based natural language processing (NLP) algorithm is trained to classify epilepsy progress notes as belonging to a patient with a specific type of epilepsy from a particular hospital. The same SVM is then used to classify notes from another hospital. Our null hypothesis is that an NLP algorithm cannot be trained using epilepsy-specific notes from one hospital and subsequently used to classify notes from another hospital better than a random baseline classifier. The hypothesis is tested using epilepsy progress notes from the three hospitals.ResultsWe are able to reject the null hypothesis at the 95% level. It is also found that classification was improved by including notes from a second hospital in the SVM training sample.Discussion and conclusionWith a reasonably uniform epilepsy vocabulary and an NLP-based algorithm able to use this uniformity to classify epilepsy progress notes across different hospitals, we can pursue automated comparisons of patient conditions, treatments, and diagnoses across different healthcare settings.

Automated detection of sudden unexpected death in epilepsy risk factors in electronic medical records using natural language processing

“…Although only minor revisions were required to achieve generalizability, this may be a practical barrier to dissemination. 38 There is growing interest in advancing information technologies and data-mining techniques to expand clinical knowledge and improve patient care. Furthermore, to provide clinical value, we need NLP tools to be integrated in the EMR and translatable between different EMR systems.…”

“…Hospitals may not currently have the necessary support for implementation and maintenance of regular expressions. 39,40 Steps to increase data utility for epilepsy research include (1) continued collaboration between institutions to share data for epilepsy research, 41,42 (2) building structured data elements into physician note templates to improve documentation at the point of care for later use in epilepsy research, (3) continued development of epilepsy-specific NLP tools to extract valuable information from unstructured clinical narratives in EMRs, [28][29][30][31]38 (4) increased adoption of standard epilepsy classification systems 43,44 to improve consistency of language used in EMRs, and (5) continued development of epilepsy-specific ontologies 45 and vocabularies 37 for epilepsy researchers using EMR data. We performed data analysis on a research server, after data were extracted from the EMR.…”

“…30 There are a handful of epilepsy researchers who have developed promising NLP tools, including techniques to extract clinical information from EMU discharge summaries, 30 and methods to identify epilepsy surgery candidates, 28 possible first time febrile seizures, 29 nonepileptic seizures, 31 and generalized versus focal epilepsy. 38 There is growing interest in advancing information technologies and data-mining techniques to expand clinical knowledge and improve patient care. 39,40 Steps to increase data utility for epilepsy research include (1) continued collaboration between institutions to share data for epilepsy research, 41,42 (2) building structured data elements into physician note templates to improve documentation at the point of care for later use in epilepsy research, (3) continued development of epilepsy-specific NLP tools to extract valuable information from unstructured clinical narratives in EMRs, [28][29][30][31]38 (4) increased adoption of standard epilepsy classification systems 43,44 to improve consistency of language used in EMRs, and (5) continued development of epilepsy-specific ontologies 45 and vocabularies 37 for epilepsy researchers using EMR data.…”

Building a Natural Language Processing Tool to Identify Patients With High Clinical Suspicion for Kawasaki Disease from Emergency Department Notes

“…An NLP system to categorize patients into low‐, high‐, and equivocal‐risk categories for acute appendicitis had a sensitivity of 89.7% and positive predicted value of 95.2% . NLP systems have also determined disease severity in epilepsy and asthma status and have been applied to multiple sources of clinical notes including radiology reports, ED notes, discharge summaries, clinical visits, and progress notes …”

Natural Language Processing for Cohort Discovery in a Discharge Prediction Model for the Neonatal ICU

“…In pediatric settings, NLP has been used to identify asthma status [13], celiac disease [14], and epilepsy [15]. These studies demonstrate that NLP can be used to accurately identify patients belonging to certain cohorts; however, to our knowledge, we are presenting the first study using NLP in predicting discharge dates in an NICU.…”