EPtalk by Dr. Jayne 8/29/24

I’ve written previously about telehealth and the math that is needed to try to prove that it will result in overall savings to the healthcare economy. A recent JAMA viewpoint article further dissects the impact of telehealth on care delivery spending, calling out the “iron triangle” of tradeoffs where a service may improve only two of the three elements of quality, cost, and access.

The authors point out that telehealth may lead to more care, especially if preventive procedures are recommended as part of encounters. This will inherently increase spending, making suspect the claims that telehealth will reduce healthcare costs. On the other hand, that increased spending should drive value, but that value is often realized well down the line.

The authors propose that reframing the discussion around value might lead to different choices, such as only expanding telehealth services that address the areas of highest value. An example given is funding telehealth visits for federally qualified health centers in the hopes of improving quality and equity. Only time will tell, and we’ll have to wait to see what happens with federal telehealth funding in the US.

From Podcast Schmodcast: “Re: your pet peeve of being forced to use your phone for webinar links. I totally agree. Maybe I’m just a grumpy old man, but I prefer to do most of my work on a 24-inch monitor or larger but will use a laptop in a pinch. My pet peeve is written ’articles’ that are little more than redirects to podcasts, which I can’t stand. I can read much faster than many of the podcasters speak and am very much a visual learner. Podcast creators need to include a transcript so that they aren’t discriminating against what I suspect is a large part of the population that feels similarly.” I’m one of those people, so I agree. For the love of all things, please include a transcript. I’ve found that when I try to listen to recordings, I get too tempted to multitask, which results in my absorbing very little of the spoken content. When I’m consuming written documents, it’s much less likely that I’ll try to surf the internet or do any number of things that will cause me to have to skip back and listen again.

Members of the American Medical Informatics Association (AMIA) received an email this week addressing concerns about the organization’s Annual Symposium that is being hosted at the Hilton San Francisco Union Square. The hotel is apparently subject to a labor dispute and union leaders are discouraging organizations from doing business at the property. Since negotiations are ongoing, AMIA is following closely and hoping for a resolution that doesn’t impact the meeting. The hotel claims to have contingency plans to ensure service delivery in the event of a strike, but having stayed at hotels in several adverse but less-contentious circumstances (including boil orders and weather emergencies), I’m not hopeful for their ability to host the conference during a strike without some level of disruption. I’ll be following this one closely.

Speaking of meetings, Oracle has announced that its CloudWorld headliner will be legendary rock band Journey. I’ve seen them perform and it was a great show, but I’d imagine that a half century on the road might be starting to take its toll on some of the performers. Attendees can purchase a guest ticket for the show for a mere $350 while supplies last. Admission to the Oracle client conference is $2,300 with the price dropping to $1,700 for groups of five or more. The registration site includes a “Convince your boss” section complete with an email template to help workers summarize the costs and benefits of attendance.

I was back flying the friendly skies this week, enjoying some West Coast sunshine before starting a big project. Unfortunately, my seatmate made it a less than fun experience, as she constantly talked to herself, made tsk-tsk sounds when reading her emails, and laughed hysterically while marking up a PowerPoint presentation. I could see everything she was doing since she didn’t have a privacy filter. I could also see her email address and her passwords that were on a sticky note that was covered in tape applied to her laptop. I had half a mind to log in to her Concur account and enter a bunch of bogus expenses, just to prove a point.

The flight attendants had to scold her for failing to put her laptop away as instructed, after which she slammed things around trying to stuff them into her enormous bag. I normally travel with noise canceling earbuds, but somehow they got left at home, which is a mistake I won’t be making again. In fact, I might throw a pair of foam earplugs in my bag as a precaution since they weigh nothing and would have been very welcome in this situation. They say travel is broadening, but I would argue that it’s not always in the way we might want.

I’m as much at risk of being drawn in my clickbait headlines as the next girl, so I admit I was taken in by a discussion of “Death Bots.” I wasn’t even sure what the term referred, to so of course I had to follow along. The article is a transcript (yay!) of a discussion by medical ethicist Art Caplan. The concept is this: patients who know that they will be dying soon might be able to record their voices so that after they pass, family members can converse with a virtual entity that is representative of their loved one. This AI-driven entity would become part of the grief process and might also draw from other materials that are left by the deceased individual such as diaries, writings, videos, and more.

It’s very “Star Trek” to be able to have a conversation with your departed ancestors, but as a physician who has worked with patients and their families through various levels of grieving, it would need to be clear that anyone participating in this is part of an unregulated experiment that deviates from our current scientific understanding of grief. I’d feel a lot better if participants went through some kind of informed consent process, but given the fact that this is already being commercialized, that would be unlikely.

Caplan points out the risks of having an AI version of a departed person “create information that sounds like you, but really isn’t what you have said, despite the effort to glean it from recordings and past information about you. He illustrates the other ways to leave memories for loved ones, including audio / video recordings, diaries, and the like.

Caplan agrees with the need for a consent process as well as safeguards for control of the information and cessation of the service if survivors desire it to end. The comments on the piece are interesting and bring up topics including regulations and their enforcement, the impact of such a service on survivor mental health, and more. Another notes, “Hopefully I have a way to go before I am gone, which gives me a lot of time to try this out and see whether I can even tolerate my own company after I am gone, much less inflict myself on posterity.” There are numerous comments on how this might go awry. My favorite comment is this: “One character in Futurama was Nixon’s head in a bottle of preservative. It spoke, and had opinions. But of course, AI is more sophisticated now.”

What are your thoughts on so-called Death Bots? Would you make one yourself, or want to have one representing a loved one? Leave a comment or email me.

Email Dr. Jayne.