*.*
News7News 7
HomeHealthCan App Data Link Sleep and Chronic Disease?; AI Model to Manage Diabetes

Can App Data Link Sleep and Chronic Disease?; AI Model to Manage Diabetes

by News7

TTHealthWatch is a weekly podcast from Texas Tech. In it, Elizabeth Tracey, director of electronic media for Johns Hopkins Medicine in Baltimore, 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 topics include a monoclonal antibody for ulcerative colitis, the impact of cash payments on healthcare utilization by low-income people, an artificial intelligence (AI) model to manage diabetes, and sleep and chronic disease.

Program notes:

0:38 Cash payments and healthcare utilization

1:38 Biomarker measurement

2:38 No effect on COVID vaccination

3:38 Enable healthier choices

4:38 Didn’t offset costs of the program

5:03 AI and primary diabetes care

6:04 Compared with primary care physicians (PCPs) and endocrinology residents

7:04 Only 4 weeks of follow-up

7:25 Sleep patterns and risk of chronic disease

8:25 Fractured sleep positively associated

9:25 Longer follow-up needed

10:25 Irregularity and duration

11:01 Antibodies to treat ulcerative colitis

12:05 IV use of antibody

13:28 End

Transcript:

Elizabeth: Can app data help inform the relationship between sleep and chronic disease?

Rick: Using antibodies to treat ulcerative colitis?

Elizabeth: What’s the impact of a cash benefit on healthcare utilization and health?

Rick: An image-based deep learning and language model for diabetes care.

Elizabeth: That’s what we’re 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: 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: Rick, I’d like to turn straight to JAMA and take a look at this issue of, gosh, if we give people money, can we impact their utilization of healthcare? It’s an idea that, in fact, we have talked about at least a couple of times in our decades, by the way of podcasting. This purports to be, however, a unique study in examining this issue.

What they did is these were people who were low-income in Massachusetts. They gave them a cash benefit via a debit card of up to $400 per month for 9 months. This is the first time that any study has looked at that provision of cash benefits for a more prolonged period of time. They also had their control group who did not get this cash benefit over the 9-month period.

Their primary outcome measure was emergency department [ED] visits, but they also looked at outpatient use overall and by specialty, COVID-19 vaccination, and biomarkers such as cholesterol levels. They had 2,880 people who applied for a lottery to be randomized into these groups. There were ultimately 746 participants who were randomized to receive the cash benefits. Their mean age was 45 years, and 77% were female.

Those folks who got the cash benefit had 217 versus 317 ED visits per 1,000 people over this time period. This included reductions in ED visits related to behavioral health, substance use, and those that resulted in hospitalization.

Between these two groups, there was no statistically significant effect on outpatient visits, and outpatient visits to other subspecialties were higher in the cash-benefit group. They note that that was particularly true for people who did not own a car. They saw no significant effect on COVID-19 vaccination, blood pressure, body weight, hemoglobin A1C, or cholesterol level.

What they say is, gosh, this cash benefit does result in significantly fewer ED visits. It’s probably worth looking at this even more rigorously in the future.

Rick: It’s really an interesting study, because there was concern that if you just put cash in people’s hands, especially people in the low-income status, or people that have behavioral issues and/or drug abuse issues, there would be more drug abuse.

What this particular study showed in this unique population — I want to stress that — is that there were no more increased visits for drug use, but actually decreased visits to the emergency room for behavioral issues.

The authors surmise, in fact, what this does is it decreases the mental health issues or the financial strain that occurs in individuals who don’t have money. That results in the decrease of emergency room visits, especially for behavioral health issues.

Elizabeth: They also postulate that this money may have enabled them to choose healthier things, healthier foods — not just the reduction in mental distress that may be related to economic distress.

I thought the study was fascinating in that they stated that they obtained electronic health record data for 3 years from the three major health systems that were all around this area where they were doing this study. There were 15 hospitals, 30 health centers, and many additional outpatient clinics that they obtained this data from. They were really able to get this comprehensive look, and that makes it pretty compelling for me.

Rick: It does, because most of the other studies that have looked at this haven’t had the totality of looking at all of the medical records across the geographical region, because people can go to different hospitals; you might not capture all that data. I think this is a pretty robust study.

Elizabeth, I would be remiss if I didn’t say they did a cost analysis and said that the cash benefit would have resulted in net savings of approximately $450 per person over 9 months. That would have covered approximately one-sixth to one-seventh of what the cost of the program was.

Elizabeth: Clearly a concern. I myself am in favor of this idea of providing cash benefits to people who are at low income and I would absolutely be in favor of continuing this practice.

Rick: Again, this patient population is very specific. It’s in Chelsea, Massachusetts. The individuals were primarily Latino. There were less uninsured. The question is, can we extend this to different or larger populations to get the same results? I hope future studies will tell us about that.

Elizabeth, since you talked about AI, can we turn to Nature Medicine and talk about how that could affect primary diabetes care?

Elizabeth: Yes, of course.

Rick: As you’re aware, and all our listeners should be aware, there are now more than 500 million people who have diabetes worldwide. Eighty percent of those live in low- and middle-income countries; many of them don’t even have access to primary care. One of the major complications that affects 30% to 40% of people with diabetes is diabetic retinopathy, an eye condition that is a leading cause of blindness in economically active, working-age adults.

Can we use our current AI looking at large language models and also our image systems? Can we fine-tune those, combine them together, and apply them to individuals to improve care of diabetes?

They developed an integrated image language system so that they could give specialized, personalized, diabetes management care instructions to the primary care physicians. What they discovered is that when they compared the large language model to the PCPs and endocrinology residents, both in English and in Chinese, is that these large language models outperform the PCPs and had comparable performance to endocrinology residents, especially in Chinese.

When they identified diabetic retinopathy, the PCPs’ accuracy was about 81%, but this combined large language and image model was accurate to 92%. For individuals who receive these individualized, personalized recommendations, they were more likely to adhere to diabetes management, they had better self-management, and they were more likely to actually see the ophthalmologist than individuals who just saw the PCP.

By the way, they also looked at empathy. Did the people who received these feel like they received empathetic care? The answer was yes.

Elizabeth: When you talked about the adherence, how would a large language model help to get folks to adopt those recommendations?

Rick: That’s a good question. The first thing I would say is they only looked at this out to 4 weeks. Now, whether that will translate to better adherence later, I’m not really sure. It’s interesting because they thought it was more empathetic than the PCPs maybe because it was more personalized.

Elizabeth: That is definitely something I would want to explore a little bit further and also the durability of adherence.

Rick: Yeah. We like objective clinical outcomes. For example, what did the hemoglobin A1C do?

Elizabeth: Unquestionably more of this coming. Remaining in Nature Medicine then, let’s take a look at another technology-driven study. This one is associating sleep patterns and risk of chronic disease. It’s using participants in the All of Us Research Program.

They identified 6,785 participants in whom they could get data from Fitbits that would actually identify what was going on with their sleep and long-term follow-up with regard to some chronic diseases. A total of 71% of the participants were female, 84% self-identified as white, and 71% had a college degree — so a very prescribed cohort here. Their median age was 50 years. Their sleep monitoring period was 4.5 years.

They were able to find that, if you had sleep irregularity, you were at increased odds of incident obesity, hyperlipidemia, hypertension, major depressive disorder, and generalized anxiety disorder. They observed J-shaped associations between average daily sleep duration and hypertension, major depressive disorder, and generalized anxiety disorder.

Some of this is unsurprising, of course. There has been a long-term research effort trying to identify the relationships between sleep and chronic disease. This study says, hey, yeah, and we have a lot of robust and validated data with these Fitbits and therefore we can make some better associations.

My major objections to this are this cohort: educated, white women who had the wherewithal to purchase Fitbits and then upload that data. Does this really pertain to anybody else? Also, it’s a very young cohort and I would like to see this go out for a much longer period of time and really delineate the relationships with chronic disease.

Rick: Yeah. Elizabeth, we are going to hear a lot more results from this All of Us Research Program. This is an NIH [National Institutes of Health]-funded initiative to gather health data from more than one million diverse people living in the United States.

The remarkable thing about this particular study is they have information monitoring for about 4.5 years. Now, you can’t do sleep studies in an individual for 4.5 years. I agree with you — individuals who enroll in this digitally and stay up with it may differ somewhat from people that are the other demographics of the United States. But you and I would agree that there is no other way one could get this information in almost 7,000 individuals over about 4.5 years.

We say people need an average of 7 to 9 hours of sleep. By the way, they confirm that about 7 hours is about the right amount of time to associate with the decreased incidence of chronic diseases. But the fact that not only sleep duration, but sleep irregularity also matters as well — that’s new information.

Elizabeth: It still, though, begs the question of what do we do about this? Because, as we know, this sleep, both irregularity and duration, are kind of complex things that do not respond well to medicines, for example.

Rick: We have talked before about that there are non-pharmacologic things that people could do to improve their sleep patterns: regularity, decreasing distractions, and eating patterns.

Elizabeth: I would like to note that they taught me something in this paper. They say that they employed something that’s called phenome-wide association studies, PheWAS, to identify some of these associations and that was something that I was not familiar with.

Rick: I’m going to turn to our last topic, and that’s the use of antibodies to treat ulcerative colitis. That’s in JAMA.

Ulcerative colitis is a chronic inflammatory bowel disease that’s initiated by our immune system — mostly with diarrhea and rectal bleeding, bowel urgency — and it affects about 1.5 million people in North America. About one in five of those will be hospitalized within 5 years of diagnosis and about 7% will actually require a colectomy — i.e., part of their colon be removed.

We have therapies that are anti-inflammatory. There are a lot of different immune therapies that are available and monoclonal antibodies. This particular study looked at a monoclonal antibody, risankizumab. It specifically targets one of the inflammatory proteins called a cytokine.

They looked at 975 individuals with an average age of 42. They would try to induce a remission and in the others where they would try to maintain the remission. They started with about 975 patients, half of whom received usual care, and the other half received an IV administration of the antibody at 0, 4, and 8 weeks.

What they discovered was that the clinical remission rate was 20% versus 6% for placebo. Then when they looked at maintenance, those who received the antibody continued to receive it — subcutaneous injections over the course of several months afterwards. At 1 year, 40% of them had remission versus 25% for placebo. There were no adverse events in those who received the antibody.

The FDA has approved it for ulcerative colitis, but where it fits now into the scheme of the other drugs we won’t know until we have head-to-head comparisons.

Elizabeth: I think also that ulcerative colitis, like an awful lot of other chronic diseases that we talked about, that’s a clinical entity and my suspicion is that there is a lot of things that are individual that underpin that. It may very likely require some other assessment of why is it that you are presenting with ulcerative colitis and what is the best strategy for you.

Rick: Absolutely. Because of these different advanced therapies, they target different pathways that cause ulcerative colitis, and that may be a strategy that we have to look at.

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

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

Source : MedPageToday

You may also like

12345678..........................%%%...*...........................................$$$$$$$$$$$$$$$$$$$$--------------------.....