Increasing and disparate use of neuroimaging for adults and children with non-traumatic headaches in the US emergency departments: Opportunities for improvement

Optimizing Advanced Imaging of the Pediatric Patient in the Emergency Department: Technical Report

Advanced diagnostic imaging modalities, including ultrasonography, computed tomography, and magnetic resonance imaging, are key components in the evaluation and management of pediatric patients presenting to the emergency department. Advances in imaging technology have led to the availability of faster and more accurate tools to improve patient care. Notwithstanding these advances, it is important for physicians, physician assistants, and nurse practitioners to understand the risks and limitations associated with advanced imaging in children and to limit imaging studies that are considered low value, when possible. This technical report provides a summary of imaging strategies for specific conditions where advanced imaging is commonly considered in the emergency department. As an accompaniment to the policy statement, this document provides resources and strategies to optimize advanced imaging, including clinical decision support mechanisms, teleradiology, shared decision-making, and rationale for deferred imaging for patients who will be transferred for definitive care.

Optimizing Advanced Imaging of the Pediatric Patient in the Emergency Department: Technical Report

Advanced diagnostic imaging modalities, including ultrasonography, computed tomography, and magnetic resonance imaging (MRI), are key components in the evaluation and management of pediatric patients presenting to the emergency department. Advances in imaging technology have led to the availability of faster and more accurate tools to improve patient care. Notwithstanding these advances, it is important for physicians, physician assistants, and nurse practitioners to understand the risks and limitations associated with advanced imaging in children and to limit imaging studies that are considered low value, when possible. This technical report provides a summary of imaging strategies for specific conditions where advanced imaging is commonly considered in the emergency department. As an accompaniment to the policy statement, this document provides resources and strategies to optimize advanced imaging, including clinical decision support mechanisms, teleradiology, shared decision-making, and rationale for deferred imaging for patients who will be transferred for definitive care.

Performance of an Open-Source Large Language Model in Extracting Information from Free-Text Radiology Reports

Purpose To assess the performance of a local open-source large language model (LLM) in various information extraction tasks from real-life emergency brain MRI reports. Materials and Methods All consecutive emergency brain MRI reports written in 2022 from a French quaternary center were retrospectively reviewed. Two radiologists identified MRI scans that were performed in the emergency department for headaches. Four radiologists scored the reports' conclusions as either normal or abnormal. Abnormalities were labeled as either headache-causing or incidental. Vicuna (LMSYS Org), an open-source LLM, performed the same tasks. Vicuna's performance metrics were evaluated using the radiologists' consensus as the reference standard. Results Among the 2398 reports during the study period, radiologists identified 595 that included headaches in the indication (median age of patients, 35 years [IQR, 26-51 years]; 68% [403 of 595] women). A positive finding was reported in 227 of 595 (38%) cases, 136 of which could explain the headache. The LLM had a sensitivity of 98.0% (95% CI: 96.5, 99.0) and specificity of 99.3% (95% CI: 98.8, 99.7) for detecting the presence of headache in the clinical context, a sensitivity of 99.4% (95% CI: 98.3, 99.9) and specificity of 98.6% (95% CI: 92.2, 100.0) for the use of contrast medium injection, a sensitivity of 96.0% (95% CI: 92.5, 98.2) and specificity of 98.9% (95% CI: 97.2, 99.7) for study categorization as either normal or abnormal, and a sensitivity of 88.2% (95% CI: 81.6, 93.1) and specificity of 73% (95% CI: 62, 81) for causal inference between MRI findings and headache. Conclusion An open-source LLM was able to extract information from free-text radiology reports with excellent accuracy without requiring further training. Keywords: Large Language Model (LLM), Generative Pretrained Transformers (GPT), Open Source, Information Extraction, Report, Brain, MRI Supplemental material is available for this article. Published under a CC BY 4.0 license. See also the commentary by Akinci D'Antonoli and Bluethgen in this issue.

Evolving Radiology Trainee Neuroimaging Workloads: A National Medicare Claims-based Analysis

RATIONALE AND OBJECTIVES

While radiology training programs aim to prepare trainees for clinical practice, the relationship between trainee, and national radiology workforce demands is unclear. This study assesses changing radiology trainee neuroimaging workloads nationwide for neuroimaging studies.

MATERIALS AND METHODS

Using aggregate Medicare claims files from 2002 to 2018, we identified all computed tomography (CT) and magnetic resonance (MR) examinations of the brain, head and neck, and spine (hereafter "neuroimaging") in Medicare fee-for-service beneficiaries nationwide. Using separate Medicare files, we calculated population utilization rates, and work relative value unit (wRVU) weights of all diagnostic neuroradiology services. Using claims modifiers, we identified services rendered by radiology trainees. Using separate national trainee enrollment files, we calculated mean annual per trainee wRVUs.

RESULTS

Between 2002 and 2018, total Medicare neuroimaging claims increased for both radiologists overall (86.1%) and trainees (162.5%), including increases in both CT (102.9% vs 196.8%), and MR (59.9% vs 106.6%). The national percentage of all radiologist neuroimaging wRVUs rendered by trainees increased 46.1% (3.8% of all wRVUs nationally in 2002 to 5.6% in 2018). National trainee increases were present across all neuroimaging services but greatest for head and neck CT (+86.5%). Mean annual per radiology trainee neuroimaging Medicare wRVUs increased +174.9% (42.1 per trainee in 2002 to 115.70 in 2018). Mean per trainee wRVU increases were greatest for spine CT (+394.2%) but present across all neuroimaging services.

CONCLUSION

As neuroimaging utilization in Medicare beneficiaries has grown, radiology trainee neuroimaging workloads have increased disproportionately.