Me failing to fix the radiology volume crisis
Julie Bauml
Diagnostic Radiologist and Clinical Informatician; Medical Director for AI (HOPPR); Adjunct Professor in Biomedical Informatics (Vanderbilt)
For those who can't read my chicken scratch above:
Radiology Volume Crunch
Increased volume per rad
decreased time for non-RVU activities (tumor board, educating trainees, consulting with patients or clinicians)-->useful and career satisfying
HIGH BURNOUT: RSIs, early retirements, unfilled positions
Where to claw back time?
-AI/ML/tech efficiency gains
Negatives: upper limit of visual/cognitive burden of human brain (burnout could worsen: increase efficiency trap into increase volume temptation); safety concerns
Positives: reduce RSIs
-More radiologists-->residency spots have seen no significant increase in decades
Positives: less volume per rad; Negatives: decrease salaries, would be pulling docs from other specialties like primary care
-non-MD mid level readers
Positives: less volume per rad; Negatives: safety concerns/liability: who takes on the malpractice?
-Decrease imaging volume overall
Positives: less vol/rad, less cost to overall system, less patient radiation; Negatives: decreased salaries; perception of care rationing
OBSTACLES:
1. Fee for service incentive structure
2. Current AI/HIT market focus on binary diagnostic tools (not workflow)
3. High malpractice costs/fear of lawsuits
SOLUTIONS:
1. Slay the fee for service dragon: single payer? capitated payments for imaging/total payments per diagnosis? allow billing of the currently non-RVU generating tasks?
2. United front of rads to developers: We want integrated hands free workflow tools for rote tasks leveraging NLP, NOT binary dx tool add-ons OR build them ourselves
3. Tort reform; nationalized standards on malpractice.
Clear rules for percentage of liability for any novel solutions: rads, mid-levels, vendors, government, hospital
The above problems and solutions are riddled with issues and complications. I'm not advocating them, I'm just thinking out loud. I'd love to hear YOUR thoughts!
Aurora Radiology Consultants Northwestern Advocate
1 年On the right track—- go get em!
Senior AI Product Manager @ Sirona Medical | Evaluation and validation of generative-AI and Deep-Learning models
1 年This is great! On the last point, I have heard that malpractice law varies quite a bit across states, with places like Texas where the bar is a lot higher a lawsuit. Is there any evidence that these differences result in less radiologist burnout or higher radiologist satisfaction?
Healthcare Executive | Technology Innovation & Strategy | AI + Imaging + Clinical Informatics
1 年I love this! We all need more whiteboard / brainstorming in our lives.