Jianlin Shi1, John F Hurdle1, Stacy A Johnson2, Jeffrey P Ferraro3, David E Skarda4, Samuel R G Finlayson5, Matthew H Samore3, Brian T Bucher6. 1. Department of Biomedical Informatics, University of Utah School of Medicine, Salt Lake City, UT. 2. Department of Medicine, University of Utah School of Medicine, Salt Lake City, UT; Thrombosis Service, University of Utah Health, Salt Lake City, UT. 3. Department of Medicine, University of Utah School of Medicine, Salt Lake City, UT; IDEAS Center 2.0, VA Salt Lake City Health Care System, Salt Lake City, UT. 4. Intermountain Healthcare, Salt Lake City, UT; Department of Surgery, University of Utah School of Medicine, Salt Lake City, UT. 5. Department of Surgery, University of Utah School of Medicine, Salt Lake City, UT. Electronic address: https://twitter.com/srgfin. 6. Department of Biomedical Informatics, University of Utah School of Medicine, Salt Lake City, UT; Department of Surgery, University of Utah School of Medicine, Salt Lake City, UT. Electronic address: brian.bucher@utah.edu.
Abstract
BACKGROUND: The objective of this study was to develop a portal natural language processing approach to aid in the identification of postoperative venous thromboembolism events from free-text clinical notes. METHODS: We abstracted clinical notes from 25,494 operative events from 2 independent health care systems. A venous thromboembolism detected as part of the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) was used as the reference standard. A natural language processing engine, easy clinical information extractor-pulmonary embolism/deep vein thrombosis (EasyCIE-PEDVT), was trained to detect pulmonary embolism and deep vein thrombosis from clinical notes. International Classification of Diseases (ICD) discharge diagnosis codes for venous thromboembolism were used as baseline comparators. The classification performance of EasyCIE-PEDVT was compared with International Classification of Diseases codes using sensitivity, specificity, area under the receiver operating characteristic curve, using an internal and external validation cohort. RESULTS: To detect pulmonary embolism, EasyCIE-PEDVT had a sensitivity of 0.714 and 0.815 in internal and external validation, respectively. To detect deep vein thrombosis, EasyCIE-PEDVT had a sensitivity of 0.846 and 0.849 in internal and external validation, respectively. EasyCIE-PEDVT had significantly higher discrimination for deep vein thrombosis compared with International Classification of Diseases codes in internal validation (area under the receiver operating characteristic curve: 0.920 vs 0.761; P < .001) and external validation (area under the receiver operating characteristic curve: 0.921 vs 0.794; P < .001). There was no significant difference in the discrimination for pulmonary embolism between EasyCIE-PEDVT and ICD codes. CONCLUSION: Accurate surveillance of postoperative venous thromboembolism may be achieved using natural language processing on clinical notes in 2 independent health care systems. These findings suggest natural language processing may augment manual chart abstraction for large registries such as NSQIP.
BACKGROUND: The objective of this study was to develop a portal natural language processing approach to aid in the identification of postoperative venous thromboembolism events from free-text clinical notes. METHODS: We abstracted clinical notes from 25,494 operative events from 2 independent health care systems. A venous thromboembolism detected as part of the American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) was used as the reference standard. A natural language processing engine, easy clinical information extractor-pulmonary embolism/deep vein thrombosis (EasyCIE-PEDVT), was trained to detect pulmonary embolism and deep vein thrombosis from clinical notes. International Classification of Diseases (ICD) discharge diagnosis codes for venous thromboembolism were used as baseline comparators. The classification performance of EasyCIE-PEDVT was compared with International Classification of Diseases codes using sensitivity, specificity, area under the receiver operating characteristic curve, using an internal and external validation cohort. RESULTS: To detect pulmonary embolism, EasyCIE-PEDVT had a sensitivity of 0.714 and 0.815 in internal and external validation, respectively. To detect deep vein thrombosis, EasyCIE-PEDVT had a sensitivity of 0.846 and 0.849 in internal and external validation, respectively. EasyCIE-PEDVT had significantly higher discrimination for deep vein thrombosis compared with International Classification of Diseases codes in internal validation (area under the receiver operating characteristic curve: 0.920 vs 0.761; P < .001) and external validation (area under the receiver operating characteristic curve: 0.921 vs 0.794; P < .001). There was no significant difference in the discrimination for pulmonary embolism between EasyCIE-PEDVT and ICD codes. CONCLUSION: Accurate surveillance of postoperative venous thromboembolism may be achieved using natural language processing on clinical notes in 2 independent health care systems. These findings suggest natural language processing may augment manual chart abstraction for large registries such as NSQIP.
Authors: Paul D Rozeboom; Michael R Bronsert; Catherine G Velopulos; William G Henderson; Kathryn L Colborn; Karl E Hammermeister; Anne Lambert-Kerzner; Mark G Hall; Robert C McIntyre; Robert A Meguid Journal: Surgery Date: 2020-09-06 Impact factor: 3.982
Authors: Jianlin Shi; Siru Liu; Liese C C Pruitt; Carolyn L Luppens; Jeffrey P Ferraro; Adi V Gundlapalli; Wendy W Chapman; Brian T Bucher Journal: AMIA Annu Symp Proc Date: 2020-03-04
Authors: Karl Y Bilimoria; Jeanette Chung; Mila H Ju; Elliott R Haut; David J Bentrem; Clifford Y Ko; David W Baker Journal: JAMA Date: 2013-10-09 Impact factor: 56.272
Authors: Anthony D Yang; Daniel Brock Hewitt; Eddie Blay; Lindsey J Kreutzer; Christopher M Quinn; Kimberly A Cradock; Vivek Prachand; Karl Y Bilimoria Journal: Ann Surg Date: 2020-06 Impact factor: 13.787