AI-Generated Study Titles; A Hip-Hop Song and Suicidality; Surgeons Square Off

TTHealthWatch is a weekly podcast from Texas Tech. In it, Elizabeth Tracey, director of electronic media for Johns Hopkins Medicine, and Rick Lange, MD, president of the Texas Tech University Health Sciences Center in El Paso, look at the top medical stories of the week.

This week’s topic is the annual BMJ Christmas issue.

Program notes:

0:33 BMJ Christmas issue

1:00 Artificial intelligence and titles of papers

2:00 Real title or AI-generated?

3:00 Times of day or night for dental appointments

3:17 Aging trajectories of superheroes

4:17 Undertake HIIT

5:18 Like how they did study

5:35 Rocket science or brain surgery

6:35 Compared engineers to neurosurgeons

7:15 A hip-hop song and suicidality

8:17 Lifeline received excess calls

9:18 Protective effect of stories of hope and recovery

10:48 End

Transcript:

Elizabeth Tracey: How well do superheroes age?

Rick Lange, MD: Generating Christmas research articles using artificial intelligence.

Elizabeth: Can a hip-hop song reduce suicidality?

Rick: And what do you mean it’s not rocket science?

Elizabeth: That’s what we are talking about this week on TT HealthWatch, your weekly look at the medical headlines from Texas Tech University Health Sciences Center in El Paso. I’m Elizabeth Tracey, a Baltimore-based medical journalist.

Rick: And I’m Rick Lange, president of Texas Tech University Health Sciences Center in El Paso, where I’m also dean of the Paul L. Foster School of Medicine.

Elizabeth: For those of our listeners who may not remember, we frequently, but not always, take on this issue of the BMJ, which is somewhat tongue-in-cheek. It’s their annual Christmas issue and a lot of the research, and it brings a smile to our faces. Which of these studies would you like to start with?

Rick: Elizabeth, let’s start with artificial intelligence. These are tongue-in-cheek articles, but they are actual studies, just a little bit of a twist on them. The BMJ has been doing this for over 25 years.

So here is the interesting thing. We have talked about the use of artificial intelligence, but it’s actually kind of revolutionized medical practice over the past decade. It’s using computer process and available datasets to arrive at plausible hypotheses. And in medicine, we use it to actually read X-rays.

Here is what they said, “Okay. We now have many years of titles of actual research articles in the BMJ. Let’s use artificial intelligence to generate research article titles and see how well they match up. Are they as plausible? Are they as attractive? Are they as enjoyable? Are they of educational value?”

To use the AI, they had over 175 billion varied items of text collected from things like Wikipedia and from a collection of books and websites. They generated research articles. They sent them to 25 different healthcare providers and compared them blindly to actual titles. They said, “Which of these do you believe are real titles? Which ones are artificial intelligence-generated? Which of these are really the most interesting to you?”

Elizabeth, I am going to give you some of the titles and I want you to tell me: real or generated by artificial intelligence? Here is one right here. “The clinical effectiveness of lollipops as a treatment for sore throats: a randomized controlled trial.” Real or artificial?

Elizabeth: I think that one’s real.

Rick: Artificial. Here is one. “The survival times of chocolate in the hospital wards: a covert observational study.” Real or artificial?

Elizabeth: I’m going to go with real again.

Rick: It is absolutely real. The last one, “The association between belief in conspiracy theories and the willingness to receive vaccinations.” Real or artificial?

Elizabeth: Well, it’s a real phenomenon. That’s for sure.

Rick: Artificial. You’re absolutely wrong. In fact, what they discovered was that when individuals chose the single most plausible title, 40% of the time they chose the one that was generated by AI. The one that you just chose was the one that was thought to be most plausible.

Elizabeth: Certainly, as I said, a real phenomenon that’s going on right now, so it’s not surprising. I bet it will be an upcoming one that we will talk about.

Rick: Here is one that I thought was really cute. “Is Jack Frost nipping at your nose? The observational study of the times of day and night people make emergency dental appointments.” Artificial, nevertheless, very enjoyable.

Elizabeth: Very much fun. We will recommend to folks that they take a look at that particular article, and let’s move to another one in the same issue.

This is anticipating the aging trajectories of superheroes in the Marvel Cinematic Universe. I think it’s really great that they even took this on. They said, “Okay, we’re going to take a look at these superheroes.” They eliminated outliers and that was Spider Man, who just graduated from high school, and Thor, who has been alive for several millennia. They just selected the cohort of superheroes that were estimated to be between 30 and 55 years of age. They said, “Okay. What they are doing now, how is that going to predict how they actually are going to age over the long haul?”

Positive characteristics that the superheroes have is that they are physically active, they are socially engaged, they are optimistic, they have high educational attainment, and with one notable exception, they are all of a healthy weight. Obviously, these are things that are associated with really good aging possibilities.

They also note that they often undertake high intensity interval training, or HIIT, which as we know has become a very popular thing to engage in right now if you have limited time to engage in exercise. Their negative behaviors, however, are that they are exposed to very loud noises, sometimes extra-terrestrial loud noises, and abundant air pollution. They have many multiple head injuries and those are precipitated by high-risk physical activities, which could result in sudden death, of course. The multiple head injuries, we know chronic traumatic encephalopathy can be the result of that. But in toto, they think that in general, superheroes are probably on a pretty positive trajectory to live for a long time. What are your thoughts, Rick?

Rick: Yeah. There are a couple things you didn’t mention. They are obviously exposed to interplanetary collisions and explosions as well. You mentioned the one that was overweight and this is The Hulk. He has got a BMI of about 120.

The suggestion is that they need to focus on challenges like how to provide better-quality healthcare to themselves and social care across a large aging population, especially a superhuman population as well.

I like how they did their studies. They did audio-visual material was reviewed by all authors. They ranged from eight to 50 hours of watching 24 different Marvel movies released between 2008 and 2021. Next time I want to sign up for that study.

Elizabeth: Haha. Good luck with that one. Why don’t we turn to, what about rocket science and rocket scientists?

Rick: We have all heard the phrase, “It’s not rocket science,” or, “It’s not brain surgery.” That’s to imply that it must be a pretty simple task. That implies, of course, that rocket scientists and brain surgeons are actually smarter than the rest of us.

Well, you know what? Nobody has really ever tested that theory. That’s exactly what these authors did. They compared the cognitive testing scores in neurosurgeons and aerospace engineers to help settle the age-old question: “It’s not brain surgery” or “It’s not rocket science” — is it really a deserved statement?

They asked for a number of these individuals from the United Kingdom, Europe, the United States, and Canada. In the final analysis, 329 aerospace engineers and 72 neurosurgeons were compared to over 250,000 people that had taken the Great British Intelligence Test. This test things like cognition, planning and reasoning, working memory, attention, and emotion-processing abilities.

First of all, they compared the engineers to the neuroscientists. They found that the aerospace engineers and neurosurgeons, although they were equally matched, aerospace engineers showed better mental manipulation abilities, but neurosurgeons were better at semantic problem-solving.

However, when they compared to them to the general population, there really wasn’t any significant difference, implying that although we give them great credit, they really don’t deserve it. I personally think that credit deserves to cardiologists. We should say “It’s not an interventional cardiology procedure.” Instead of saying, “It’s not rocket science. It’s not brain surgery,” we’re going to say instead, “It’s just a walk in the park.”

Elizabeth: That’s what I’m going to say to you, of course, the next time such a rejoinder might become appropriate.

Let’s turn to our last one, which is really one that’s rather serious, even though it’s in this particular issue of the BMJ. Actually, it’s been covered on all kinds of national media. That’s this idea that, is there an association between Logic’s hip-hop song “1-800-273-8255” with Lifeline calls and suicides in the United States? — an interrupted time series analysis.

What they did was take a look at these calls to the National Suicide Prevention Lifeline in the United States and daily call volumes, so a huge data set. They did this from the 1^st of January 2010 to December 2018. Then they also looked at Twitter posts after the release of this particular song and estimated the amount and time of attention that the song received.

In this 34-day period, there were three events with the strongest public attention when the song was released, the MTV Video Music Awards in 2017 and the Grammy Awards in 2018. During that time, Lifeline received an excess of almost 10,000 calls, so an increase of almost 7% in the calls to people who were contemplating suicide or at least were thinking about some aspect of it and were motivated to call this number. They also, with that, saw almost a 6%, a 5.5% reduction in that same period of the actual numbers of suicides that would have been expected.

While this is an interesting association, is it causal? We don’t know the answer to that, but this is a pretty powerful thing.

Rick: For those listeners who may not be familiar, a hip-hop artist named Logic composed a song called “1-800-273-8255.” That’s the number to Lifeline calls. As you said, after it was released, and after it was MTV Video shown, and after the Grammys, the calls went up and the number of suicides went down.

By the way, this is called the Papageno effect and it’s actually a protective effect of stories of hope and recovery from suicidal crisis. That’s what this song and video was about. It was a story of promoting hope and recovery featured in the song in someone that had suicidal ideation.

That’s just the opposite of what’s called the Werther effect, when notable individuals like Anthony Bourdain or Kate Spade committed suicide, and that increases the suicide risk and suicide intentions over the subsequent reporting of this. Instead of reporting those things, let’s report hope and recovery. What a positive effect! What we should be able to do is take that and leverage that, and say, we need to be doing this, especially around times of the holidays when some people are very sad.

Elizabeth: It’s also interesting; they note that there was about a 10% uptick in Google searches for Lifeline in the 28 days after the song was released. By the end of 2020, the song had surpassed 1 billion streams on Spotify.

Rick: If we tried to advertise the Lifeline calls and suicide prevention, we couldn’t get that kind of press. It just shows how important both social media is and how much music and arts can influence health. It’s an unusual study, very positive. We need to have that message of hope and recovery out there for everybody.

Elizabeth: On that note, then, that’s a look at this week’s medical headlines from Texas Tech. I’m Elizabeth Tracey.

Rick: And I’m Rick Lange. Y’all listen up and make healthy choices.