Healthcare AI Daily Briefing

Operations & Workflow

AI use in insurance claims linked to $942 million extra spending over two years.

An analysis by the Blue Cross Blue Shield Association found hospitals’ use of AI tools when submitting insurance claims added $942 million to healthcare spending over two years. The analysis noted a sharp increase in patients documented as having complex conditions but found no evidence of corresponding change in care delivered, indicating a disconnect between coding and treatment. [4]

Research & Evidence

AI-driven phenotyping reveals four HFpEF trajectories with differing mortality risk.

Journal article

Using NLP and latent class analysis on multisite EHRs, researchers identified four reproducible heart failure with preserved ejection fraction phenogroups among 2,223 patients. The cardiovascular-kidney-metabolic group had the highest mortality risk. Current phenogroup status modestly improved mortality discrimination over baseline alone. [2]

Medical Imaging

AI guidance enables rapid, sensitive DVT triage with high prioritization specificity.

Preprint

In a multicenter, double-blinded prospective study of 594 subjects, ThinkSono Guidance achieved diagnostic image quality in 86.83% of scans and triage sensitivity of 92.86% for proximal DVT. Prioritization specificity was 97.96%, allowing standard of care ultrasounds to be avoided in 35.32% of patients. [3]

Clinical Practice

Most health workers say AI makes care feel less human, few find tech helpful.

A Brenau University survey found that 69% of healthcare workers say empathy is the top human skill AI cannot replicate, while 70% say AI is making care feel less human for patients. Separately, Sage data show only 32% of senior care teams report their current technology generally helps their work, despite 63% being open to AI support. [1]

Policy & Governance

Liability for AI errors in medicine remains unsettled as clinicians bear oversight burden.

Healthcare attorney Amanda Hill states that when AI influences clinical decision making, the clinician using it remains accountable for the care provided, emphasizing that meaningful oversight requires more than scanning AI-generated notes. A July 2026 lawsuit against Mayo Clinic alleges the hospital fired an AI compliance lead after she raised concerns about rushed AI incorporation without proper safeguards. [5]