Blog

About a year ago at Erlanger Baroness, Chattanooga’s largest hospital, anesthesia staff noticed a nurse was slurring her speech and struggling to stay awake while on duty at the surgery center, according to a Tennessee Board of Nursing Consent Order.

In the following days, the nurse tested positive for drugs and was fired, according to the order. The nurse later admitted that for months she had stolen and abused fentanyl left over from surgeries, sometimes daily, according to the warrant.

Under most circumstances, this would be a routine case of what is known as “drug diversion,” the illegal taking of controlled substances from health care facilities, which is believed to be so widespread that it occurs in nearly every hospital in the United States.

But the Erlanger case stands out because a high-tech watchdog was supposed to be on guard.

The hospital uses the newest line of defense against drug diversion: Sentri7, medication tracking software powered by artificial intelligence and designed to detect missing medications faster than any human. But for months at Erlanger, Sentri7 failed to raise alarms, missing missing medications and other “inconsistencies” that “should have been flagged,” the nursing board order states.

Erlanger’s case, which has not been previously reported, offers a rare glimpse into an apparent flaw in AI drug diversion software used at hundreds of U.S. hospitals with little transparency or oversight. Healthcare facilities are not required to disclose the implementation of this type of software or inform anyone about its malfunction, so there is no complete overview of how widely these programs are used or how often they fail.

Erlanger Baroness, also known as Erlanger Medical Center, declined to comment on its use of Sentri7 or the diverted medications. André Rebelo, a spokesman for the health division of Wolters Kluwer, the Dutch technology company behind Sentri7, declined to answer questions about what happened at Erlanger, but said the company remained “confident in our software.”

Little transparency

David Rastall, a Johns Hopkins Medicine neurologist and AI researcher, said that because AI technology is heavily proprietary and hospital officials often don’t understand how it works, this lack of transparency allows errors to be buried rather than corrected. That means the mistakes could be repeated at other hospitals, he said.

“The ideal thing for patients, caregivers and hospital systems would be,” Rastall said, “when an AI is found to be making some type of error, that becomes very transparent and public.”

The Drug Enforcement Administration requires hospitals to confidentially report lost or stolen medications. Hospitals can also report stolen medications to state health agencies, which license medical professionals and investigate irregularities.

But these reports are not required to include details about any artificial intelligence software involved, according to interviews with three drug diversion prevention experts. In interviews, all said they had never seen a publicly documented AI failure like the one apparent at Erlanger.

“I’ve never seen these technologies mentioned in that specific way,” Jacob Smith, a pharmacist in charge of medication safety at Johns Hopkins Medicine, said of Sentri7’s apparent failure. “It doesn’t make sense to me how you can miss it.”

Smith and other experts said the Erlanger case also raises questions because the theft of leftover drugs is one of the most well-known diversion methods. And fentanyl, a painkiller that can be 50 times stronger than heroin, is one of the most common targets.

Terri Vidals, the founder of Rxpert Solutionsquestioned whether the Erlanger case was the result of user error and not a malfunction.

“These are the basics of this software,” Vidals said. “I find it interesting that they say the software didn’t flag it. I think maybe there’s more to that story.”

The apparent failure of Sentri7 in Erlanger was revealed by the Tennessee Department of Health in a routine release from state disciplinary orders in December. Among those records was the Board of Nursing order, summarizing a state investigation into nurse anesthetist John Stevenson, who settled the case against him by signing the document in November.

Stevenson declined to comment through his attorney. He has not been charged with any crime related to the Erlanger case. The nursing board placed his license on probation while he attended drug counseling.

Bill Christian, spokesman for the Department of Health and the Board of Nursing, declined to comment on the Erlanger or Sentri7 case. In response to public records requests, the Department of Health and the Tennessee Health Facilities Commission said they had no other documents about the apparent Sentri7 failure in Erlanger.

Erlanger spokesman Charlie Milburn said earlier this year that the hospital had prepared a written statement about the use of Sentri7 in response to questions from KFF Health News.

That statement was never published.

“Our legal team is discussing whether this is something we want to talk about,” Milburn said in a March email, before declining to answer questions.

Kristy Drollinger, an executive at Wolters Kluwer who spoke generally about Sentri7 to KFF Health News in March, said the software is in high demand because many hospitals have had difficulty securing their medications.

Sentri7 monitors about 60 “risk attributions” that identify red flags for hospital employees to investigate further, Drollinger said.

“It’s pretty scary,” Drollinger said of the widespread drug theft. “Every health system, every health center, has had diversion at some point, and probably has it now.”

“The way of the future”

Drug diversion is a widespread challenge in US medical facilities. It can lead to patients not receiving medications or receiving medications contaminated with blood-borne diseases. It is estimated that up to 15% of all healthcare workers divert medications at least once, according to the nonprofit. Healthcare Diversion Network.

The diversion has been linked to at least 13 disease outbreaks (causing more than 200 infections, mostly hepatitis C) since 1985, according to the Centers for Disease Control and Prevention.

To avoid this, hospitals try to track each pill or vial from the time it is dispensed to the time it is administered to the patient, comparing data from electronic medication cabinets and patients’ medical records.

Hospital staff once performed this painstaking process manually, but in the last decade the task has been largely automated using anti-diversion software. After years of mergers and acquisitions, two programs now dominate the industry: Wolters Kluwer’s Sentri7 and Bluesight’s ControlCheck. Both incorporate AI.

“It’s definitely the way of the future,” said Luke Overmire, owner of Diversion specialists.

More than 1,500 hospitals use ControlCheck, according to Bluesight. Another 700 use Sentri7 clinical surveillance programs, which may include its drug diversion software, according to Wolters Kluwer.

No company publishes the price of its software. Smith, the Johns Hopkins drug safety officer, said hospitals buy these “expensive technologies” because a disastrous diversion case could result in a multimillion-dollar fine from the DEA.

“They don’t promise a return on investment,” Smith said. “They promise to avoid costs.”

In 2022, a peer reviewed study funded by the National Institutes of Health found that Sentri7, then known as Flowlytics, could discover drug diversion faster than existing methods. The lead author of the study worked for Invistics, the company that previously owned Sentri7.

According to that study, researchers tested the software by having it review medication data spanning two years and 10 hospitals looking for 22 nurses already known to have diverted medications.

The program not only found them all, the study claims, but it found them faster than humans in as little as a week and as long as a year and a half.

At Erlanger, humans were the first to spot the signs of trouble.

According to the Nursing Board order, co-workers reported that Stevenson appeared disabled “while on duty at the surgical center” around June 30, 2025.

Stevenson “had slurred his speech, appeared extremely tired, was seen standing with his eyes closed and swaying, nodded his head while standing, and appeared to have difficulty keeping his eyes open,” according to the warrant.

When questioned by state investigators, Stevenson admitted that he began diverting “unused fentanyl that would otherwise have been wasted following surgical procedures” in March 2025, according to the warrant. Stevenson said he initially used the fentanyl waste once or twice a week, and then “increased the use to daily” in June of that year, according to the order.

Erlanger audited Stevenson’s dispensing history during those four months. It found approximately five instances in which Sentri7 failed to flag missing medications, according to the order.

It adds that the hospital found “additional inconsistencies between medication dispensing and waste documentation that should have been detected by the automated tracking system.”

One possible explanation is provided by the Board of Nursing, which said in the order that Sentri7 was in its “initial learning phase” at Erlanger, although the board did not provide details.

In an interview, without specifically talking about Erlanger, Drollinger said Sentri7 does not have a “learning phase,” because it is trained on nine to 12 months of historical data when it is deployed at a new hospital.

Johns Hopkins’ Smith had another theory.

In an interview, Smith said his experience with AI medication diversion software had led him to believe that it is effective in monitoring emergency rooms and intensive care units, but not so much in operating rooms, where medications are dispensed and recorded differently.

These areas may be harder for AI to track, Smith said, and therefore require humans to keep a closer eye.

“We have people whose only job is to work with this software,” Smith said. “Software is one part, but if you rely on software to give you all the signals, you’re going to miss things. It’s just not 100%.”

KFF Health News is a national newsroom that produces in-depth journalism on health issues and is one of the main operating programs of KFF — the independent source for health policy research, polling and journalism.