Related Experiment Video
Updated: Jan 18, 2026
07:45
The Use of Mixed Reality in Custom-Made Revision Hip Arthroplasty: A First Case Report
Published on: August 4, 2022
3.8K
Use of Natural Language Processing Algorithms to Identify Common Data Elements in Operative Notes for Total Hip
Cody C Wyles1, Meagan E Tibbo1, Sunyang Fu1
1Departments of Orthopedic Surgery (C.C.W., M.E.T., D.J.B., D.G.L., and H.M.-K.) and Health Sciences Research (S.F., Y.W., S.S., W.K.K., and H.M.-K.), Mayo Clinic, Rochester, Minnesota.
The Journal of Bone and Joint Surgery. American Volume
|October 1, 2019
Summary
Natural language processing (NLP) accurately extracts key data from total hip arthroplasty (THA) operative notes, offering a faster alternative to manual chart review. This technology shows promise for clinical research and data registries.
Area of Science:
- Orthopaedic surgery
- Medical informatics
- Natural Language Processing (NLP)
Background:
- Manual chart review for extracting data from operative notes is time-consuming and requires specialized expertise.
- Natural Language Processing (NLP) tools can efficiently extract critical information from unstructured electronic health record (EHR) text.
- This study explores NLP's capability to identify key elements in total hip arthroplasty (THA) operative reports.
Purpose of the Study:
- To assess the accuracy of NLP algorithms in identifying specific data points within THA operative notes.
- To evaluate NLP's potential as an automated method for data extraction in orthopaedic research.
- To demonstrate the external validity of NLP algorithms across different institutions.
Main Methods:
- Developed and tested three NLP algorithms targeting operative approach, fixation method, and bearing surface category from THA operative reports (2000-2015).
- Algorithms were trained on a sample of reports and validated on a separate test set, including data from outside facilities.
- Accuracy was determined by comparing NLP extraction against manual chart review as the gold standard.
Main Results:
- The operative approach algorithm achieved 99.2% accuracy.
- The fixation technique algorithm demonstrated 90.7% accuracy.
- The bearing surface algorithm showed 95.8% accuracy, with comparable performance on external data, confirming validity.
Conclusions:
- NLP-enabled algorithms offer a viable and efficient alternative to manual chart review for extracting data from orthopaedic operative notes.
- This proof-of-concept study supports the use of NLP in clinical research and registry development for expeditious and cost-effective data extraction.
- NLP technology can reliably capture essential data elements, streamlining research processes in orthopaedics.
