Information Extraction From Electronic Health Records to Predict Readmission Following Acute Myocardial Infarction: Does Natural Language Processing Using Clinical Notes Improve Prediction of Readmission?
Background Social risk factors influence rehospitalization rates yet are challenging to incorporate into prediction models. Integration of social risk factors using natural language processing (NLP) and machine learning could improve risk prediction of 30‐day readmission following an acute myocardia...
| Main Authors: | , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2022-04-01
|
| Series: | Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease |
| Subjects: | |
| Online Access: | https://www.ahajournals.org/doi/10.1161/JAHA.121.024198 |
| _version_ | 1797849648980819968 |
|---|---|
| author | Jeremiah R. Brown Iben M. Ricket Ruth M. Reeves Rashmee U. Shah Christine A. Goodrich Glen Gobbel Meagan E. Stabler Amy M. Perkins Freneka Minter Kevin C. Cox Chad Dorn Jason Denton Bruce E. Bray Ramkiran Gouripeddi John Higgins Wendy W. Chapman Todd MacKenzie Michael E. Matheny |
| author_facet | Jeremiah R. Brown Iben M. Ricket Ruth M. Reeves Rashmee U. Shah Christine A. Goodrich Glen Gobbel Meagan E. Stabler Amy M. Perkins Freneka Minter Kevin C. Cox Chad Dorn Jason Denton Bruce E. Bray Ramkiran Gouripeddi John Higgins Wendy W. Chapman Todd MacKenzie Michael E. Matheny |
| author_sort | Jeremiah R. Brown |
| collection | DOAJ |
| description | Background Social risk factors influence rehospitalization rates yet are challenging to incorporate into prediction models. Integration of social risk factors using natural language processing (NLP) and machine learning could improve risk prediction of 30‐day readmission following an acute myocardial infarction. Methods and Results Patients were enrolled into derivation and validation cohorts. The derivation cohort included inpatient discharges from Vanderbilt University Medical Center between January 1, 2007, and December 31, 2016, with a primary diagnosis of acute myocardial infarction, who were discharged alive, and not transferred from another facility. The validation cohort included patients from Dartmouth‐Hitchcock Health Center between April 2, 2011, and December 31, 2016, meeting the same eligibility criteria described above. Data from both sites were linked to Centers for Medicare & Medicaid Services administrative data to supplement 30‐day hospital readmissions. Clinical notes from each cohort were extracted, and an NLP model was deployed, counting mentions of 7 social risk factors. Five machine learning models were run using clinical and NLP‐derived variables. Model discrimination and calibration were assessed, and receiver operating characteristic comparison analyses were performed. The 30‐day rehospitalization rates among the derivation (n=6165) and validation (n=4024) cohorts were 15.1% (n=934) and 10.2% (n=412), respectively. The derivation models demonstrated no statistical improvement in model performance with the addition of the selected NLP‐derived social risk factors. Conclusions Social risk factors extracted using NLP did not significantly improve 30‐day readmission prediction among hospitalized patients with acute myocardial infarction. Alternative methods are needed to capture social risk factors. |
| first_indexed | 2024-04-09T18:47:28Z |
| format | Article |
| id | doaj.art-8eeaa93660b34a4eb7587d857b0b09d8 |
| institution | Directory Open Access Journal |
| issn | 2047-9980 |
| language | English |
| last_indexed | 2024-04-09T18:47:28Z |
| publishDate | 2022-04-01 |
| publisher | Wiley |
| record_format | Article |
| series | Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease |
| spelling | doaj.art-8eeaa93660b34a4eb7587d857b0b09d82023-04-10T11:57:34ZengWileyJournal of the American Heart Association: Cardiovascular and Cerebrovascular Disease2047-99802022-04-0111710.1161/JAHA.121.024198Information Extraction From Electronic Health Records to Predict Readmission Following Acute Myocardial Infarction: Does Natural Language Processing Using Clinical Notes Improve Prediction of Readmission?Jeremiah R. Brown0Iben M. Ricket1Ruth M. Reeves2Rashmee U. Shah3Christine A. Goodrich4Glen Gobbel5Meagan E. Stabler6Amy M. Perkins7Freneka Minter8Kevin C. Cox9Chad Dorn10Jason Denton11Bruce E. Bray12Ramkiran Gouripeddi13John Higgins14Wendy W. Chapman15Todd MacKenzie16Michael E. Matheny17Departments of Epidemiology and Biomedical Data Science Dartmouth Geisel School of Medicine Hanover NHDepartments of Epidemiology and Biomedical Data Science Dartmouth Geisel School of Medicine Hanover NHDepartment of Biomedical Informatics Vanderbilt University Medical Center Nashville TNDivision of Cardiovascular Medicine University of Utah School of Medicine Salt Lake City UTDepartments of Epidemiology and Biomedical Data Science Dartmouth Geisel School of Medicine Hanover NHDepartment of Biomedical Informatics Vanderbilt University Medical Center Nashville TNDepartments of Epidemiology and Biomedical Data Science Dartmouth Geisel School of Medicine Hanover NHGeriatric Research Education and Clinical Care Center Tennessee Valley Healthcare System VA Nashville TNDepartment of Biomedical Informatics Vanderbilt University Medical Center Nashville TNDepartments of Epidemiology and Biomedical Data Science Dartmouth Geisel School of Medicine Hanover NHDepartment of Biomedical Informatics Vanderbilt University Medical Center Nashville TNDepartment of Biomedical Informatics Vanderbilt University Medical Center Nashville TNDivision of General Internal Medicine Vanderbilt University Medical Center Nashville TNDepartment of Biomedical Informatics University of Utah School of Medicine Salt Lake City UTDepartments of Epidemiology and Biomedical Data Science Dartmouth Geisel School of Medicine Hanover NHCentre for Digital Transformation of Health University of Melbourne Melbourne Victoria AustraliaDepartments of Epidemiology and Biomedical Data Science Dartmouth Geisel School of Medicine Hanover NHDepartment of Biomedical Informatics Vanderbilt University Medical Center Nashville TNBackground Social risk factors influence rehospitalization rates yet are challenging to incorporate into prediction models. Integration of social risk factors using natural language processing (NLP) and machine learning could improve risk prediction of 30‐day readmission following an acute myocardial infarction. Methods and Results Patients were enrolled into derivation and validation cohorts. The derivation cohort included inpatient discharges from Vanderbilt University Medical Center between January 1, 2007, and December 31, 2016, with a primary diagnosis of acute myocardial infarction, who were discharged alive, and not transferred from another facility. The validation cohort included patients from Dartmouth‐Hitchcock Health Center between April 2, 2011, and December 31, 2016, meeting the same eligibility criteria described above. Data from both sites were linked to Centers for Medicare & Medicaid Services administrative data to supplement 30‐day hospital readmissions. Clinical notes from each cohort were extracted, and an NLP model was deployed, counting mentions of 7 social risk factors. Five machine learning models were run using clinical and NLP‐derived variables. Model discrimination and calibration were assessed, and receiver operating characteristic comparison analyses were performed. The 30‐day rehospitalization rates among the derivation (n=6165) and validation (n=4024) cohorts were 15.1% (n=934) and 10.2% (n=412), respectively. The derivation models demonstrated no statistical improvement in model performance with the addition of the selected NLP‐derived social risk factors. Conclusions Social risk factors extracted using NLP did not significantly improve 30‐day readmission prediction among hospitalized patients with acute myocardial infarction. Alternative methods are needed to capture social risk factors.https://www.ahajournals.org/doi/10.1161/JAHA.121.024198electronic health recordsmachine learningmyocardial infarctionnatural language processingpatient readmission |
| spellingShingle | Jeremiah R. Brown Iben M. Ricket Ruth M. Reeves Rashmee U. Shah Christine A. Goodrich Glen Gobbel Meagan E. Stabler Amy M. Perkins Freneka Minter Kevin C. Cox Chad Dorn Jason Denton Bruce E. Bray Ramkiran Gouripeddi John Higgins Wendy W. Chapman Todd MacKenzie Michael E. Matheny Information Extraction From Electronic Health Records to Predict Readmission Following Acute Myocardial Infarction: Does Natural Language Processing Using Clinical Notes Improve Prediction of Readmission? Journal of the American Heart Association: Cardiovascular and Cerebrovascular Disease electronic health records machine learning myocardial infarction natural language processing patient readmission |
| title | Information Extraction From Electronic Health Records to Predict Readmission Following Acute Myocardial Infarction: Does Natural Language Processing Using Clinical Notes Improve Prediction of Readmission? |
| title_full | Information Extraction From Electronic Health Records to Predict Readmission Following Acute Myocardial Infarction: Does Natural Language Processing Using Clinical Notes Improve Prediction of Readmission? |
| title_fullStr | Information Extraction From Electronic Health Records to Predict Readmission Following Acute Myocardial Infarction: Does Natural Language Processing Using Clinical Notes Improve Prediction of Readmission? |
| title_full_unstemmed | Information Extraction From Electronic Health Records to Predict Readmission Following Acute Myocardial Infarction: Does Natural Language Processing Using Clinical Notes Improve Prediction of Readmission? |
| title_short | Information Extraction From Electronic Health Records to Predict Readmission Following Acute Myocardial Infarction: Does Natural Language Processing Using Clinical Notes Improve Prediction of Readmission? |
| title_sort | information extraction from electronic health records to predict readmission following acute myocardial infarction does natural language processing using clinical notes improve prediction of readmission |
| topic | electronic health records machine learning myocardial infarction natural language processing patient readmission |
| url | https://www.ahajournals.org/doi/10.1161/JAHA.121.024198 |
| work_keys_str_mv | AT jeremiahrbrown informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT ibenmricket informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT ruthmreeves informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT rashmeeushah informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT christineagoodrich informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT glengobbel informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT meaganestabler informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT amymperkins informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT frenekaminter informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT kevinccox informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT chaddorn informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT jasondenton informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT bruceebray informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT ramkirangouripeddi informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT johnhiggins informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT wendywchapman informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT toddmackenzie informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission AT michaelematheny informationextractionfromelectronichealthrecordstopredictreadmissionfollowingacutemyocardialinfarctiondoesnaturallanguageprocessingusingclinicalnotesimprovepredictionofreadmission |