Identifying Acute Low Back Pain Episodes in Primary Care Practice From Clinical Notes: Observational Study
Riccardo Miotto1,2,3, Bethany L Percha2,3, Benjamin S Glicksberg1,2,3
1Hasso Plattner Institute for Digital Health at Mount Sinai, Icahn School of Medicine at Mount Sinai, New York, NY, United States.
JMIR Medical Informatics
|March 5, 2020
Summary
This study shows that deep learning models can automatically distinguish acute low back pain (LBP) from clinical notes, improving care guidelines. This method offers a path to better LBP management and billing accuracy.
Area of Science:
- Medical Informatics
- Natural Language Processing
- Clinical Decision Support
Background:
- Acute and chronic low back pain (LBP) share an ICD-10 code (M54.5), hindering distinct treatment and billing.
- Current differentiation relies on manual chart reviews, impeding data-driven guidelines.
Purpose of the Study:
- To assess the feasibility of automatically differentiating acute LBP episodes using free-text clinical notes.
- To develop an automated method for identifying acute LBP for improved clinical management.
Main Methods:
- Utilized 17,409 clinical notes, with 891 manually annotated for acute LBP.
- Compared supervised (logistic regression, deep learning ConvNet) and unsupervised (topic modeling) methods.
- Trained models using manual annotations and ICD-10 codes.
Main Results:
- A Convolutional Neural Network (ConvNet) trained on manual annotations achieved an AUC of 0.98 and F-score of 0.70.
- ConvNet performance remained robust with fewer annotated documents.
- Topic models outperformed ICD-10 code-based methods when manual annotations were absent.
Conclusions:
- Clinical notes can be leveraged to systematically learn therapeutic strategies for acute LBP.
- This approach facilitates improved billing guidelines and point-of-care management options.
- Automated LBP acuity identification supports data-driven healthcare recommendations.
Related Concept Videos
01:11Observational Studies
10.6K
Observational studies are a type of analytical study where researchers observe events without any interventions. In other words, the researcher does not influence the response variable or the experiment's outcome.
There are three types of observational studies – Prospective, retrospective, and cross-sectional.
Prospective Study
Prospective studies, also known as longitudinal or cohort studies, are carried out by collecting future data from groups sharing similar characteristics. One...
There are three types of observational studies – Prospective, retrospective, and cross-sectional.
Prospective Study
Prospective studies, also known as longitudinal or cohort studies, are carried out by collecting future data from groups sharing similar characteristics. One...
10.6K
01:26Guidelines for Nursing Documentation II
1.5K
Effective documentation is an integral part of nursing practice. Here are some essential guidelines to follow when documenting patient care:
Timely documentation is crucial to ensure continuity of care for patients. Any delays in recording or reporting medical information can result in medical errors and even adverse patient outcomes. From medication administration to diagnostic test results, every detail must be accurately and promptly documented to provide the best possible care for patients.
Timely documentation is crucial to ensure continuity of care for patients. Any delays in recording or reporting medical information can result in medical errors and even adverse patient outcomes. From medication administration to diagnostic test results, every detail must be accurately and promptly documented to provide the best possible care for patients.
1.5K