AI raised some healthcare costs
Reporting suggests AI transcription and related automation have increased rather than reduced some healthcare costs, as new tools add workflow complexity and review burdens. The piece argues automation can create more downstream work—like verification and error correction—before any efficiency gains appear. That dynamic implies caution when adding AI layers that might multiply operational steps. (futurism.com)
Hospitals bought artificial intelligence note-taking tools to save time, but insurers and providers now say the software is also pushing some medical bills higher. (statnews.com) The tools are called ambient scribes: software that listens to a doctor-patient visit and drafts the clinical note for the electronic health record. A March 25, 2025 report from the Peterson Health Technology Institute said early adopters saw less documentation burden, but the financial effect was still unclear. (phti.org) By April 8, 2026, STAT reported that insurers, hospitals, investors and providers privately agreed the tools were increasing “coding intensity,” meaning visits were documented in ways that supported higher-paying billing codes. Caroline Pearson, the Peterson institute’s executive director, said participants at a recent roundtable were in agreement on that point. (statnews.com) The mechanism is simple: better notes can capture more diagnoses, more detail and more complexity, which can translate into higher reimbursement. A December 24, 2025 policy brief in *npj Digital Medicine* said early evidence linked ambient scribes to higher billing and stronger risk-adjustment coding. (nature.com) That paper pointed to concrete examples. Riverside Health in Virginia saw an 11% rise in physician work relative value units and a 14% increase in documented Hierarchical Condition Category diagnoses per encounter, while Northwestern Medicine clinicians using Nuance DAX billed more high-level Evaluation and Management visits on average. (nature.com) The cost story is colliding with a real workflow problem in medicine. A February 19, 2025 *JAMA Network Open* study of 46 outpatient clinicians in Philadelphia found ambient scribe use cut note time per appointment by 20.4% and after-hours work by 30.0%. (jamanetwork.com) A larger 2026 study summarized by *Healthcare IT News* found the gains were more modest in broader practice: clinicians using ambient documentation spent 13 fewer minutes a day in the electronic health record and 16 fewer minutes a day on documentation, with the biggest benefits among frequent users. (healthcareitnews.com) Health systems are still scaling the tools quickly. A September 24, 2025 comment in *npj Digital Medicine* said about 30% of physician practices were already using artificial intelligence scribes, and one large system had deployed them to more than 7,000 physicians across 2.5 million patient encounters in 14 months. (nature.com) Supporters say the software is doing work doctors were already supposed to do and may be correcting years of under-documentation. Critics say every artificial intelligence draft still has to be checked, and the same systems that save keyboard time can add review, error-correction and compliance work before a note is safe to sign. (ama-assn.org) (nature.com) The fight now is less about whether ambient scribes save some clinician time than about who pays when cleaner documentation produces higher bills. That is why insurers are talking about downcoding and recalibrating payments even as hospitals keep rolling the tools out. (nature.com)
