I spent Friday evening following the return of the Artemis II mission, breathing a sigh of relief when the crew was safely recovered in good condition.
I viewed the NASA live stream and watched the circling helicopters on Flightradar24 for the full experience. The stunning contrast between the parachute, the sky, and the ocean was one more addition to a week of surreal experiences.
I chatted with friends who were following the the mission closely. We were struck by how the world was riveted to an event that didn’t include influencers or celebrities. The idea that we would be fascinated by a jar of Nutella stealing the show made us chuckle, as did the fact that Commander Reid Wiseman deliberately boosted the Rise zero-gravity indicator plushy when he had been instructed to leave it in the capsule.
While double-checking to make sure that NASA hadn’t changed its photo use policies, I noticed some AI-related additions to the NASA media usage guidelines. NASA specifically prohibits the attribution of information to NASA once it has been incorporated into large language models, due to the difficulty of verifying accuracy in that situation. It bans the use of NASA insignia appearing with AI-generated images or in training AI tools.
Developers can note that they included NASA materials, but they are not allowed to imply any review or permission by NASA. Use of text that implies that something is “according to NASA” in AI outputs is also prohibited. Image Credit: NASA/Bill Ingalls.
I knew that Medicare was doing a demonstration project using AI, but I was unaware of the details until I saw a recent press release. The Electronic Frontier Foundation (EFF) has filed suit under the Freedom of Information Act seeking details about the WISeR (Wasteful and Inappropriate Service Reduction) pilot program and its use of AI to evaluate prior authorization requests for care that is needed by Medicare beneficiaries.
Prior authorization has been used in Medicare Advantage plans for some time, but it is rare in traditional Medicare. That adds to the concerns.
The six-state pilot program may apply to six million patients. Critics are concerned that algorithms may be biased and that delays or denials of care may be discriminatory. They are asking for transparency around the WISeR algorithms and what training data was used to create the model.
The presence of safeguards is also important, as is an analysis of situations where care is denied and patients experience negative outcomes. Vendors are compensated based on how many services they deny, so it’s easy to understand why patient advocacy organizations are concerned.
A provider fact sheet on the CMS website says that the pilot is limited to a subset of procedures. This includes implanted nerve stimulators, epidural steroid injections for pain management, treatments for vertebral compression fractures, spinal fusions, knee surgeries, sleep apnea treatments, incontinence therapies, spinal decompression for spinal stenosis, and skin and tissue substitutes. These services are pricy, so I understand their inclusion.
I would be interested to hear if any organizations have built rules or alerts in their EHRs to make impacted physicians aware of the new process to streamline ordering and approval.
In addition to understanding the AI being used, the Electronic Frontier Foundation has also asked for copies of vendor agreements, data on AI performance and hallucinations, and audit results. They filed the lawsuit when the earlier Freedom of Information Act request was not addressed. Props to EFF for also including a link to the filing so that I didn’t have to go hunting for it.
I was pulled into the role of family health navigator again this weekend. A health system where several of my relatives receive care has decided that many of its employed primary care physicians will no longer be able to see patients in the hospital. Patients have not been formally notified as far as I know, but my family members heard about it from their physicians during regularly scheduled visits.
It sounds as though the change was mandated by leadership for all members of the group. I’m not surprised that the health system wouldn’t put this in writing. It certainly interferes with the practice of medicine, although system-owned medical groups are notorious for controlling what their physicians can and can’t do.
As a family physician, I made the difficult decision earlier in my career to stop seeing patients in the hospital, for several reasons. My office wasn’t near the hospital, so it was a 50-minute round trip to see a patient, plus the actual hospital care. It was challenging to fit that into my schedule given my other responsibilities. My office staff didn’t have experience working for a physician who also saw patients in the hospital, so a call from the hospital during the day caused chaos.
We didn’t have a hospital EHR then, and changes to policies on verbal orders made things more complicated and sometimes necessitated additional trips to the hospital.
I also realized that I had relatively low admitting volume. The quality numbers that the hospitalist groups posted led me to conclude that my patients would likely have better outcomes if I made the change. A typical hospitalist was carrying eight to 10 patients on their service back then, which is far fewer than they carry now.
I’m not sure about quality trends, but I know that my hospitalist physician friends are significantly more stressed by their patient load than they were when they started out. I hope that organizations are monitoring quality aggressively and adjusting staffing accordingly, but the way some facilities are running makes that a challenge.
Still, it’s one thing for physicians to make a conscious decision to stop rounding at the hospital, but another one entirely to decide to ban the practice across the board.
I don’t know what the physician contracts look like in this particular organization. I have put some feelers out to see if I can get an official update on what led to the decision and what the health system hopes to accomplish with it.
I reassured my relatives by explaining how hospitalist care is supposed to work and its benefits, but I could tell that they weren’t confident about the change.
Have any of your organizations decided to restrict medical staff privileges to no longer allow primary care physicians to care for hospitalized patients? Leave a comment or email me.
Email Dr. Jayne.
Briefly googling, preventive and preventative both mean the same thing and have both been used and accepted for centuries although…