AI-assisted breast cancer screening showed minor clinical benefits over DBT alone but was not cost-effective at standard willingness-to-pay thresholds.
Key Details
- 1Study published in Value in Health evaluated the cost-effectiveness of AI (Saige-DX) in breast cancer screening.
- 2Microsimulation compared digital breast tomosynthesis (DBT) alone and DBT plus AI for women aged 40–74 in biennial screening.
- 3AI reduced false negatives by 2.1 and false positives by 50 per 1,000 women; 0.13 fewer breast cancer deaths per 1,000.
- 4AI-assisted screening led to 3.09 additional QALYs but added $936,430 in lifetime costs per 1,000 women (ICER: $303,279/QALY).
- 5AI was not cost-effective at a $100,000/QALY threshold in 98% of simulations, mainly due to 21% increase in DCIS diagnoses.
- 6Cost-effectiveness did not improve even at very low AI pricing in most scenarios.
Why It Matters
Although AI can modestly improve accuracy in breast cancer screening, this study highlights persistent cost-effectiveness concerns, especially with increased DCIS detection. This underscores an economic and clinical challenge for broader AI adoption in screening workflows.
Source
AuntMinnie
Related News
•AI in Healthcare
Landmark AI Mammography Trial and CMS Launches AI Prior Authorization Pilot
New large randomized trial to test AI in mammography screening, while CMS launches a multi-state pilot using AI for Medicare prior authorizations.
•AuntMinnie
AI Device Manufacturers Paid $39.7M for Radiology Between 2017–2023
AI/ML-enabled medical device firms paid $39.7 million for radiology devices from 2017–2023, raising transparency concerns.
•Radiology Business
AI-Generated Reports Cut Radiology Reading Times and Gain Acceptance
AI-generated reporting significantly reduces radiologists' reading times and increases report acceptability over time.