Insurance companies deny claims all day long. They’ll deny doctor-ordered surgeries, medical stays, scans, and weeks of rehab patients need to walk again. Most people read a denial letter and assume it is final.
The reason it’s not is utilization management nurses. When a denial comes in, we read a patient’s chart, compare it to clinical criteria, and write the case for why the care should be covered. Our patients don’t generally know how it happens, they just know they get the treatment they need.
Montefiore recently laid off twelve of these nurses, and replaced us with two doctors and AI software. On Sept. 1, the layoffs went into effect. Datavant, the private-equity-backed firm selling the software, claims it will manage insurance denials with artificial intelligence (AI).
The bet is that software can do a job people spend years learning to do well. Sadly, the cost of that miscalculation will not land on the hospital’s executives. It will land on patients in the Bronx.
Many of our patients live with multiple chronic conditions, and often their charts don’t match any standard template. Software programs aren’t trained to assess these cases. Denials must be reversed by someone with decades of bedside experience navigating complicated cases. This type of judgement can’t be programmed.
It’s dangerous to think software can replace clinical judgement. When AI is left to decide which appeals are worth pursuing, the complicated cases are let go, which can result in a missed diagnosis or a critical surgery that never gets approved.
We also have to ask what exactly Montefiore is buying. Datavant has not published the dollars and cents of what the software recovers. The few public figures in this industry come from its competitors, who describe recovering somewhere between roughly one and four percent of a hospital’s revenue. There is no independent, published evidence that any of this software matches the judgment of an experienced nurse, let alone beats it.
The software holds some of our patients’ most private information, including medical records that can include immigration status. In an area with one of the largest immigrant populations in the country, this is incredibly dangerous. In recent years, Datavant and affiliated companies have disclosed data breaches, including one exposing nearly sixty thousand patient records.
Datavant also partners with Palantir – the company behind the technology that allows ICE to surveil and target immigrants for deportation – on two major projects. How will Montefiore protect patients’ files and personal information, including immigration status?
As members of the New York State Nurses Association, we bargained for contract language to protect patient care and to keep us from being replaced by AI or unproven technology. These layoffs are not “innovation.” They are a short-sighted and small cost cut, poised to do more harm than good.
Although nurses sounded the alarm about the dangers of these layoffs and tried to work with hospital executives to find alternatives, Montefiore moved ahead with the displacement.
Montefiore has a near monopoly in the Bronx. Our patients are poorer and sicker than almost anywhere else in New York, and far more likely to be immigrants, which is the whole reason a human being in their corner matters when an insurer pushes back.
Our ask of Montefiore’s leadership is simple. Keep utilization management nurses where patients can still reach us and talk to us. Use technology to make clinicians faster, not to replace us and our years of advocacy and dedicated service to Bronx patients.
The Bronx needs more nurses, not less. No nurse should lose their job as a result of these short-sighted layoffs, and no patient should be denied care because an algorithm is not a patient advocate. Care in the Bronx has always run on people who refuse to accept no as the final word.
This advocacy in patient care is what we are fighting so hard to protect.
Marilyn Shuler, RN, is a member of the New York State Nurses Association, and has worked for 39-years as a nurse at Montefiore Medical Center.
