Deep Learning Models to Predict Diagnostic and Billing Codes Following Visits to a Family Medicine Practice: Development and Validation Study

Abstract BackgroundDespite significant time spent on billing, family physicians routinely make errors and miss billing opportunities. In other disciplines, machine learning models have predicted Current Procedural Terminology codes with high accuracy. ObjectiveOur...

Full description

Saved in:
Bibliographic Details
Main Authors: Akshay Rajaram, Michael Judd, David Barber
Format: Article
Language:English
Published: JMIR Publications 2025-03-01
Series:JMIR AI
Online Access:https://ai.jmir.org/2025/1/e64279
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849706599615037440
author Akshay Rajaram
Michael Judd
David Barber
author_facet Akshay Rajaram
Michael Judd
David Barber
author_sort Akshay Rajaram
collection DOAJ
description Abstract BackgroundDespite significant time spent on billing, family physicians routinely make errors and miss billing opportunities. In other disciplines, machine learning models have predicted Current Procedural Terminology codes with high accuracy. ObjectiveOur objective was to derive machine learning models capable of predicting diagnostic and billing codes from notes recorded in the electronic medical record. MethodsWe conducted a retrospective algorithm development and validation study involving an academic family medicine practice. Visits between July 1, 2015, and June 30, 2020, containing a physician-authored note and an invoice in the electronic medical record were eligible for inclusion. We trained 2 deep learning models and compared their predictions to codes submitted for reimbursement. We calculated accuracy, recall, precision, F1 ResultsOf the 245,045 visits eligible for inclusion, 198,802 (81%) were included in model development. Accuracy was 99.8% and 99.5% for the diagnostic and billing code models, respectively. Recall was 49.4% and 70.3% for the diagnostic and billing code models, respectively. Precision was 55.3% and 76.7% for the diagnostic and billing code models, respectively. The area under the receiver operating characteristic curve was 0.983 for the diagnostic code model and 0.993 for the billing code model. ConclusionsWe developed models capable of predicting diagnostic and billing codes from electronic notes following visits to a family medicine practice. The billing code model outperformed the diagnostic code model in terms of recall and precision, likely due to fewer codes being predicted. Work is underway to further enhance model performance and assess the generalizability of these models to other family medicine practices.
format Article
id doaj-art-7505f22944c34bcba6a764d3a54ea4e7
institution DOAJ
issn 2817-1705
language English
publishDate 2025-03-01
publisher JMIR Publications
record_format Article
series JMIR AI
spelling doaj-art-7505f22944c34bcba6a764d3a54ea4e72025-08-20T03:16:08ZengJMIR PublicationsJMIR AI2817-17052025-03-014e64279e6427910.2196/64279Deep Learning Models to Predict Diagnostic and Billing Codes Following Visits to a Family Medicine Practice: Development and Validation StudyAkshay Rajaramhttp://orcid.org/0000-0002-5310-7843Michael Juddhttp://orcid.org/0009-0009-0574-0212David Barberhttp://orcid.org/0000-0002-8824-6966 Abstract BackgroundDespite significant time spent on billing, family physicians routinely make errors and miss billing opportunities. In other disciplines, machine learning models have predicted Current Procedural Terminology codes with high accuracy. ObjectiveOur objective was to derive machine learning models capable of predicting diagnostic and billing codes from notes recorded in the electronic medical record. MethodsWe conducted a retrospective algorithm development and validation study involving an academic family medicine practice. Visits between July 1, 2015, and June 30, 2020, containing a physician-authored note and an invoice in the electronic medical record were eligible for inclusion. We trained 2 deep learning models and compared their predictions to codes submitted for reimbursement. We calculated accuracy, recall, precision, F1 ResultsOf the 245,045 visits eligible for inclusion, 198,802 (81%) were included in model development. Accuracy was 99.8% and 99.5% for the diagnostic and billing code models, respectively. Recall was 49.4% and 70.3% for the diagnostic and billing code models, respectively. Precision was 55.3% and 76.7% for the diagnostic and billing code models, respectively. The area under the receiver operating characteristic curve was 0.983 for the diagnostic code model and 0.993 for the billing code model. ConclusionsWe developed models capable of predicting diagnostic and billing codes from electronic notes following visits to a family medicine practice. The billing code model outperformed the diagnostic code model in terms of recall and precision, likely due to fewer codes being predicted. Work is underway to further enhance model performance and assess the generalizability of these models to other family medicine practices.https://ai.jmir.org/2025/1/e64279
spellingShingle Akshay Rajaram
Michael Judd
David Barber
Deep Learning Models to Predict Diagnostic and Billing Codes Following Visits to a Family Medicine Practice: Development and Validation Study
JMIR AI
title Deep Learning Models to Predict Diagnostic and Billing Codes Following Visits to a Family Medicine Practice: Development and Validation Study
title_full Deep Learning Models to Predict Diagnostic and Billing Codes Following Visits to a Family Medicine Practice: Development and Validation Study
title_fullStr Deep Learning Models to Predict Diagnostic and Billing Codes Following Visits to a Family Medicine Practice: Development and Validation Study
title_full_unstemmed Deep Learning Models to Predict Diagnostic and Billing Codes Following Visits to a Family Medicine Practice: Development and Validation Study
title_short Deep Learning Models to Predict Diagnostic and Billing Codes Following Visits to a Family Medicine Practice: Development and Validation Study
title_sort deep learning models to predict diagnostic and billing codes following visits to a family medicine practice development and validation study
url https://ai.jmir.org/2025/1/e64279
work_keys_str_mv AT akshayrajaram deeplearningmodelstopredictdiagnosticandbillingcodesfollowingvisitstoafamilymedicinepracticedevelopmentandvalidationstudy
AT michaeljudd deeplearningmodelstopredictdiagnosticandbillingcodesfollowingvisitstoafamilymedicinepracticedevelopmentandvalidationstudy
AT davidbarber deeplearningmodelstopredictdiagnosticandbillingcodesfollowingvisitstoafamilymedicinepracticedevelopmentandvalidationstudy