EPtalk by Dr. Jayne 9/12/24

AI alert. One of the topics around the public health informatics virtual water cooler this week was about whether Google’s AI Overviews are negatively impacting patients.

Physicians have had a longstanding love / hate relationship with the so-called Dr. Google as a source of health information. Although many health systems spend a lot of time and money providing high-quality patient education materials, it’s just so convenient to type a clinical question into the Google search bar and hope for the best.

One of my colleagues noted that when you get an AI Overview for certain clinical topics, there’s a disclaimer that says, “This is for informational purposes only. For medical advice or diagnosis, consult a professional. Generative AI is experimental.” When I replicated the topic he mentioned on my PC with 24-inch monitor, the disclaimer scrolled off the bottom of the window, so I doubt that people who are using devices with smaller form factors see it easily.

Health literacy is woefully low in the US, with the Office of the Surgeon General reporting that only 12% of US adults possess proficient health literacy skills. Many can’t understand drug labels or understand how to identify and access healthcare resources, so it’s not surprising that they’re going to turn to consumer-level resources. For care delivery organizations that have robust patient education solutions and consumer resources, I’m challenging you to double down on those and increase their visibility so that patients know how to access them and when to use them. It doesn’t have to be a complicated omnichannel campaign – it can be as simple as having signs in exam rooms and waiting rooms, or even those paper table tents we used to see in the hospital cafeteria in the olden days.

A recent KLAS Arch Collaborative report shows that despite interoperability advances, clinicians are still struggling with synthesizing information from disparate systems. Almost half of the 33,000 clinicians surveyed said that they found it difficult to find key patient information from outside sources, with the same proportion noting that they are challenged with addressing duplicate data.

I’m sad to say that I’ll become part of the problem in a couple of weeks when I show up for a subspecialist visit with paper copies of critical records, because I don’t trust the various providers to share what needs to be shared in a timely manner. I’ve already tried to send digital copies of a pathology report to my care team and they were rejected, so I’ll be there with my manila folder in hand.

From Jersey Collector: “Re: branding. I know this has been a hot topic for you. Hospitals and health systems are getting into the act with the WNBA, which makes sense since women make the majority of healthcare decisions for their families.” That’s certainly a valid statistic, but I’m still not sure how much seeing a hospital or health system logo on a professional athlete’s uniform impacts someone’s choice of healthcare providers. I would say that the number one driver would be insurance coverage, followed by recommendations, ratings and reviews, and also the acuity of a problem.

If a loved one is having a significant issue, people tend to go to the closest facility that accepts their insurance. They don’t care  who they sponsor or what celebrity might endorse a given hospital. Some of the jersey deals run in the millions and I’m certainly glad to see women’s sports receiving sponsorships, but I can’t help but think that nurses who are looking for raises or families who are struggling with medical bills might be less than impressed.

It’s been a while since I saw a major healthcare bombshell reported, but reports out of the University of Virginia certainly meet that description. The Cavalier Daily reports that faculty have called for the immediate removal of UVA Health’s CEO as well as the dean of the school of medicine at the University of Virginia. The letter is signed by 128 members of the faculty, who accuse the two of creating a toxic work environment that compromises patient safety and has led to “an ongoing exodus of experience and expertise.”

Additional allegations include retaliation against physicians who raised safety concerns by denying promotion, encouraging staff to bypass safety processes, harassment and bullying of trainees, and financial mismanagement. One call-out in the letter mentions “disregarding valid reports of fraudulent billing and requests by senior leaders to fraudulently modify patient records in order to obfuscate adverse outcomes and boost productivity metrics.” CMS doesn’t look too kindly on this sort of thing, so I hope the institution has its compliance auditors and attorneys on standby.

The non-profit Emergency Care Research Institute (ECRI) has released a report showing that the vast majority of diagnostic errors occur during the testing process. They reviewed 3,000 patient safety adverse events and near misses. Leading issues include technical and processing errors, lack of skill in performing a test, sample mix-ups, wrong-patient issues, and communication failures. The report specifically calls out “productivity pressures that prevent providers from exploring all investigative options or from consulting other providers” as a factor in causing diagnostic error. Time pressure is also a factor when test results aren’t reviewed quickly or when results aren’t appropriately communicated to patients.

Those of us on the informatics side should take note of their findings with regard to health equity, where women and underrepresented populations can be at greater risk for diagnostic errors. They specifically call out the potential for race-based biases in medical algorithms and communication barriers, both of which can be significantly improved by thoughtful application of healthcare technologies. For organizations looking at artificial intelligence solutions, it’s going to be critical that they appraise how systems handle these biases and how the potential for hallucinations might contribute to additional opportunities for diagnostic errors.

Speaking of patient safety, World Patient Safety Day is right around the corner on September 17th. I have yet to see anyone who I regularly interact with, including my own clinical employer, making plans to mark the day. This year’s theme is focused around improving the safety of the diagnostic process, with the slogan “Get it right, make it safe!” Although this seems like a simple concept, we’ve learned that it can be more complicated than one can imagine.

My own loved one presented for a scheduled surgery this week to find that it had been booked for the wrong side of the body, leading to confusion and delay as well as stress to the family. Fortunately, the patient safety processes in place at the hospital worked and a wrong-side surgery was avoided, but it’s staggering to know that this is still a risk in 2024.

What is your institution doing to mark World Patient Safety Day? Leave a comment or email me.

Email Dr. Jayne.