Re: Health Brief: When AI is taking notes, hospitals charge more (fwd)

[robert <http://dummy.us.eu.org/robert>]

 > From: Noelle <noelle>
 > Date: Thu, 12 Mar 2026 14:39:59 -0700 (PDT)
 >
 > AI scribes programmed for upcoding?

Well, AI generating more complex health visit notes was completely
predictable since, as we all know, AI seems to be very verbose in almost
everything it does, including composing emails and writing computer code
(the latter making the lives of software engineers very, very difficult).

 >  > From: WP Intelligence <http://www.washingtonpost.com/~wpintelligence>
 >  > Date: Thu, 12 Mar 2026 21:35:14 +0000
 >  > 
 >  > Health Brief from The Washington Post
 >  > 
 >  > <
 >  > https://sli.washingtonpost.com/click?s=234952&li=health202&m=6e1c8b48186107df
 >  > cba979ecd71fbefc&p=69b3315c8c1e360585f0bcd5
 >  > >
 >  > 
 >  > <
 >  > https://s2.washingtonpost.com/car-ln-tr/44a9caf/69b3315c8c1e360585f0bcd5/5977ca66ae7e8a6816e764cb/4/54/69b3315c8c1e360585f0bcd5
 >  > >
 >  > 
 >  > <
 >  > https://sli.washingtonpost.com/click?s=1069934&li=health202&m=6e1c8b48186107dfcba979ecd71fbefc&p=69b3315c8c1e360585f0bcd5&stpe=default
 >  > >
 >  > 
 >  > <
 >  > https://sli.washingtonpost.com/click?s=1069934&li=health202&m=6e1c8b48186107dfcba979ecd71fbefc&p=69b3315c8c1e360585f0bcd5&stpe=default
 >  > >
 >  > 
 >  > By Megan R. Wilson
 >  > 
 >  > <
 >  > https://s2.washingtonpost.com/car-ln-tr/46fcaf3/69b3315c8c1e360585f0bcd5/5977ca66ae7e8a6816e764cb/8/54/69b3315c8c1e360585f0bcd5
 >  > >
 >  > Listen to this briefing <
 >  > https://s2.washingtonpost.com/car-ln-tr/46fcaf4/69b3315c8c1e360585f0bcd5/5977ca66ae7e8a6816e764cb/9/54/69b3315c8c1e360585f0bcd5
 >  > >
 >  > 
 >  > Did someone forward this newsletter to you? Sign up here to get it in your 
 >  > inbox. <
 >  > https://s2.washingtonpost.com/car-ln-tr/44ee640/69b3315c8c1e360585f0bcd5/5977ca66ae7e8a6816e764cb/10/54/69b3315c8c1e360585f0bcd5
 >  > >
 >  > 
 >  > More providers are using AI scribes during visits with patients to help with 
 >  > administrative burnout. 
 >  > 
 >  >  
 >  > 
 >  > The Lead Brief
 >  > 
 >  > Hospitals are adopting so-called AI scribes, artificial intelligence devices 
 >  > that listen to doctor-patient conversations and automatically generate 
 >  > clinical notes. The technology is meant to reduce paperwork and clinician 
 >  > burnout by documenting visits in real time.
 >  > 
 >  > But a new analysis by Trilliant Health is adding to a growing body of data 
 >  > that shows  the technology  is also increasing how much providers are paid compared to 
 >  > when it’s not in use.
 >  > 
 >  > In the Trilliant Health report, researchers found a steady shift toward higher-intensity medical 
 >  > codes among six large U.S. health systems that have adopted AI-powered 
 >  > scribing tools.
 >  > 
 >  > → This means that the outpatient visits at hospitals and affiliated 
 >  > clinics were increasingly being billed as more complex — which largely 
 >  > translates into being more expensive. The analysis examined claims data from 
 >  > between 2018 and 2024.
 >  > 
 >  > “These patterns were observed across geographically diverse and 
 >  > organizationally distinct health systems, all of which have adopted ambient 
 >  > AI scribing in recent years, suggesting external forces rather than isolated 
 >  > institutional behavior are catalyzing these changes.”
 >  > - Trilliant Health report
 >  > 
 >  > “While the health of Americans has consistently declined for years, 
 >  > especially during the covid-19 pandemic, the scale, uniformity and 
 >  > persistence of these changes suggest a cause that goes beyond” the types 
 >  > of patients being treated,” according to the report.
 >  > 
 >  > Why it matters: Even small differences between billing codes — sometimes 
 >  > $10 to $40 per visit — can add up to millions of dollars when spread 
 >  > across large hospital systems. The findings contribute to questions for 
 >  > policymakers about how AI-assisted documentation may influence billing and 
 >  > increase overall health care spending .
 >  > 
 >  > What the study found:
 >  > Higher-level billing increased for both new and returning patients across 
 >  > the six health systems. For new patient visits, high-intensity codes rose by 
 >  > about 12 to 20 percentage points over the seven-year period studied by 
 >  > researchers. At one hospital system, 80 percent of new patient visits were 
 >  > billed at the higher level. For established patients, the codes for a more 
 >  > complex visit increased by about 7 to 12 percentage points.   The shift 
 >  > appeared across most types of diagnoses, including respiratory diseases, 
 >  > mental and behavioral health issues, and digestive diseases. The largest 
 >  > increases in both new and existing patients were seen in the “factors 
 >  > influencing health status ” diagnosis codes meant to 
 >  > encapsulate circumstances that impact a patient’s health, and which are 
 >  > not captured elsewhere as diseases or injuries.   The increase in more 
 >  > complex billing codes doesn’t necessarily indicate fraud. Researchers 
 >  > argue that, because visit codes are partly based on the time spent with 
 >  > patients, better documentation may allow providers to justify higher-level 
 >  > codes that previously went under-billed. It’s an argument
 >  > that some health systems who’ve 
 >  > adopted the technology have used.   But if there is fraud stemming from 
 >  > AI-enabled “over-coding,” the report argues, it would be easy to 
 >  > diagnose because of the detailed and time-stamped digital record of the 
 >  > patient visit created by the AI scribe, so insurers could perform an audit.
 >  > The American Hospital Association pushed back, arguing that an “uptick in 
 >  > coding intensity is not necessarily driven by a more aggressive coding 
 >  > process or use of ambient scribes.”
 >  > 
 >  > Ben Teicher, a spokesperson for the industry group, said that multiple other 
 >  > factors — such as a U.S. population that’s aging and sicker overall — 
 >  > contribute to the trend.
 >  > 
 >  > → He added that most hospitals have auditing and compliance programs to 
 >  > ensure they’re accurately coding complexity levels, regardless of whether 
 >  > a facility has AI tools.
 >  > 
 >  > “Ambient scribes can reduce the administrative burden on health care 
 >  > providers, allowing them to focus more on patient care,” Teicher said. “
 >  > Ultimately, these tools can reduce provider documentation time, increase 
 >  > appointment capacity and improve staff satisfaction.”
 >  > 
 >  > → The AHA doesn’t have data on how widely AI scribes are used in 
 >  > hospitals around the country.
 >  > 
 >  > In their own analyses, some insurers have argued
 >  > that providers are using AI 
 >  > scribes to add complexity to patient billing codes without providing 
 >  > corresponding care. Insurance companies have also deployed AI tools
 >  > to handle medical claims from 
 >  > providers and help with other tasks
 >  > , which has drawn criticism
 >  > from physicians.