Harlan Krumholz, MD, SM - Outcomes Research and Y-Weight: Research to Optimize the Patient Outcomes in the Era of Highly Effective Anti-Obesity Medications

March 07, 2024

ID11435

To CiteDCA Citation Guide

00:00All right, everyone.
00:01We're going to go ahead and get
00:04started again and and continue on.
00:06And it is my pleasure to introduce our next
00:11speaker who like all of our speakers today,
00:13really doesn't need an introduction.
00:15Dr. Krumholtz graduated from Yale
00:17College and earned his medical
00:19degree from Harvard Medical School.
00:21He completed internship and residency
00:22programs in medicine at the University
00:24of California, San Francisco,
00:26and did a fellowship in cardiovascular
00:28medicine at Beth Israel in Boston.
00:30And he earned a master's degree at
00:33Harvard School of Public Health in 1995.
00:35In 1995, in his third year at year,
00:39he founded the Center for
00:42Outcomes Research and Evaluation,
00:44and in 2005, he was named the
00:47Harold Hines Junior Professor.
00:49Yeah, I have to follow the script.
00:52OK Was that my
00:53phone or yours?
00:54Well, that's your phone.
00:55OK, good. Cool. Great job.
00:57Wow, what a pleasure to be
00:58able to speak with you guys.
00:59And what an amazing Dean's workshop.
01:00Thank you to Dean Brown for
01:02setting this up and for Anya,
01:04one of my heroes and is doing such
01:06a great job in the era of OBESI.
01:07I think the leading,
01:09the internationally leading
01:11figure in obesity medicine today,
01:14both because of the quality of her science,
01:16the strength of her voice,
01:18and her ability to inspire
01:19all those around her.
01:20But she really is worth.
01:26We are fortunate to have her here. So my
01:33OK disclosures.
01:37So there's AI would say fledgling
01:40team of outcomes researchers.
01:41And I'm going to recruit you all to join us,
01:43but at the center to really support
01:46Anya's vision about how we're at a central
01:48juncture in the treatment of obesity in
01:51this country and around the world with
01:53tools that are emerging at a dizzying pace,
01:56putting us in a position to do things that
01:59were unimaginable even five years ago.
02:02So the question will be will the
02:05evidence generation keep pace with
02:06our needs to help this in this trans,
02:08the transformation.
02:10And again, you know, Anya,
02:12I think is at the center of this,
02:13but Rohan, Keira, you and Lou, Erica,
02:15Spatz and many others, Mona Sharifi,
02:18others I should have put on this slide
02:20who are doing important work in this
02:22area and I think we're gonna grow is
02:24a facet of what Y weight is about,
02:26a facet of this work.
02:28So what is outcomes research?
02:29Many of you may not be familiar with this.
02:31I think it's a more of a basic
02:33science orientation.
02:33So it's really science that
02:35concerns itself with the result.
02:37We sort of say the end result,
02:38what are we really achieving
02:39at the end of the day?
02:41How do we tangibly affect people's lives?
02:43What can we do to improve their outcomes?
02:45Not just about declaring victory
02:47because we've had a paper or there's
02:49a breakthrough or there's a new study,
02:51but at the end of the day,
02:51have we really affected population?
02:53Not that we affected individuals health.
02:55So what should we do exactly?
02:57And not only the what,
02:59but how should we do it in ways that
03:01we know we can actually ensure,
03:03ensure that this is being adopted
03:05broadly and appropriately monitoring
03:07that adoption and ensuring again
03:09that individuals are benefiting.
03:10And we'll we're focused on in this
03:12kind of research on effectiveness,
03:14efficiency, equity, patient centeredness,
03:17safety and timeliness.
03:19So you know, what's the moment?
03:21Obesity is endemic and it's
03:24causing much suffering and cost.
03:26And and by the way,
03:26because suffering and cost means there's
03:28the prospect that actually treating it
03:31will lead to economic incentives and
03:33motivations 'cause sometimes we have
03:35innovations that can be beneficial to people,
03:37but there's not an an economic reason
03:40for the healthcare system to reorient.
03:42But as we go to value based cares,
03:44people have become increasingly
03:45interested in population health.
03:46There's a strong motivation here.
03:48And let me just say clearly and of course,
03:51Anya is a great influence of for me on this,
03:54but really obesity treatment's
03:55not about appearance,
03:56but it's about health.
03:57And I think it's the idea that
04:00we've got these medications to
04:02treat obesity and reduce risk.
04:04I sort of think about the weight loss
04:07tongue in cheek as a side effect.
04:08Actually,
04:09it's a really good side effect
04:10because it helps us with compliance.
04:11People actually like this side effect.
04:13So they're going to continue
04:14with the medication.
04:15But as physicians,
04:16our central drive is to improve health
04:18and reduce risk, advance global health.
04:20And I think This is why we're
04:22going to be able to see that.
04:23That's why I think it's a historic juncture.
04:26So you may have seen this,
04:27this graphic and others.
04:29This week Lancet came out with a sort
04:32of landmark non communicable disease
04:34groups publication on obesity worldwide.
04:37This is kind of a cool figure.
04:39I just like the way it looks.
04:40I don't know what it means,
04:41but I'm just joking.
04:43But the it's 1990 on the left
04:47side in in 2022 on the right side
04:50for every country in the world.
04:52And if you look at the red,
04:53you know there's some percentages
04:55of people with obesity and
04:56you can see what's happening.
04:58I could have shown you figures
04:59from throughout this article.
05:00It shows what you already know that
05:02there's been a great degree of growth.
05:04You know,
05:04so we we're making such progress in
05:06cardiovascular disease for decades.
05:07It's slowed.
05:08In 2022,
05:08CDC reported that we actually had an
05:11uptick in cardiovascular mortality.
05:12I attributed it to this decade long,
05:15decades long increase in obesity
05:17that's coming to roost now.
05:19And we were really only treating
05:21sort of the manifestations,
05:22lipids and blood pressure,
05:23but not getting to the root
05:24cause that many people,
05:25which was the obesity itself.
05:27This is what's changing.
05:29This is again showing 1990
05:31to 2022 across the world.
05:32It doesn't show any surprise.
05:33I'm just doing it to emphasize
05:35that this is a pressing need.
05:37There's an urgent issue that's
05:40affecting population health
05:41throughout the world and now
05:43we have ability to treat this.
05:45So you know my my view is that
05:47again to reinforce this,
05:49we've been treating manifestations
05:51of obesity as population health has
05:53steadily declined but after after
05:55this period of marked improvement.
05:57But we failed on the root cause and
05:59most of our armmentarium up until now
06:01has lacked safety and effectiveness.
06:03We now strategies that address obesity
06:05and can improve health and like I said,
06:07I consider weight loss.
06:07But but what we have now
06:10is evidence and questions.
06:11I mean,
06:12as compelling evidence comes out
06:14from Phrase 3 clinical trials,
06:16it really starts to open up a
06:18wider range of questions that are
06:21needed if we're to understand how
06:23to optimize the use of these new
06:25medications And if we're able to
06:27ensure the proper implementation
06:28and application of this new
06:30knowledge in ways that will tangibly
06:33show by improvements in health.
06:35And.
06:36And so our strategies on the outcomes
06:38research side is to answer these questions.
06:40And what we're trying to do
06:41is to assemble a range of,
06:43you know if they're people here from the lab,
06:44I'll say these are our reagents,
06:46our data reagents in order to
06:48do data experiments in order to
06:50generate knowledge that will fuel
06:52the proper application of of these
06:54new strategies that are going to
06:56again come out at a dizzying pace.
06:59It's not just the two meds that we have now,
07:01it's going to, they're going to be 10s,
07:03twenty different kinds of medications and
07:05choices and we're going to have to parse this
07:07or make challenges around access and cost.
07:09They're going to be a question of
07:11who's this best in, and it's going to
07:13be a question of how to optimize it
07:15for anyone for whom it is effective.
07:17And so we'll use federal databases and
07:19public registries like UK Biobank,
07:21clinical trial databases,
07:23international data repositories
07:25like the Odyssey trials,
07:28prospective decentralized registries,
07:29regulatory science analysis and preview
07:32simulation, decentralized trials.
07:33The thing about outcomes research
07:34is we have a broad toolkit because
07:37we're motivated by the questions,
07:38not by having a singular approach
07:40with regard to how we answer it,
07:42not by having a singular assay
07:44or type of analysis,
07:45but by being able to approach
07:47us in many different ways.
07:48And I want to say we've been
07:49at this for a while.
07:50Rohan had this you know,
07:51piece in JAMA.
07:52You'll see that this we're,
07:54this group is not new to the issue
07:56of obesity.
07:57This paper actually we're writing a long
07:59time ago but after low carbohydrate
08:02diets and the obesity paradox was
08:04something we published in Heart
08:06Failure a decade and a half ago.
08:08Body mass index and mortality and
08:11acute micro infarction patients.
08:12I do self and parent reported
08:14dietary physical activity and
08:15sedentary behaviors predict worsening
08:17obesity in children.
08:18This was a PhD thesis from
08:20someone in investigative medicine,
08:21Karen Dorsey,
08:21who has focused her thesis on this
08:23and applying practice recommendations
08:25for prevention and treatment of
08:27obesity in children and adolescents.
08:29Obesity prevalence and risk.
08:30We did this internationally.
08:31We looked in China in a million
08:34persons project that we designed
08:36in order to understand risk within
08:38large scale populations in China and
08:40we published this in German network
08:42open about body mass index with
08:43blood pressure in 1.7 million Chinese adults.
08:46We we were looking at the issues
08:48around disparities and barriers to
08:50access looking at racial and ethnic
08:53disparities and and and financial
08:55barriers and overweight and obese
08:57adults eligible for Smeglitide in
08:59the US by you and Lou And another
09:02one that you and did with,
09:04we did with Anya looking at what
09:06were the implications for the select
09:07trial with regard to the population
09:09that might be eligible for it.
09:11So I'm,
09:11I'm only just saying this because
09:13we've been at this for a while,
09:15but now we've got a center,
09:16now we've got the world's leading
09:18expert in obesity medicine.
09:19I think we're poised to kind of
09:21organize these efforts that have
09:23been a little disparate and not
09:25necessarily concentrated in a way
09:27that really positions Yale as a as a
09:28real leader and as a pillar of what
09:30this center is going to be about.
09:32Obviously there's other science
09:33you've been discussing today.
09:35There's a wide range of great
09:36science at Yale in this area.
09:38We want outcomes research to have
09:40a a big according to that.
09:41And I just said what we don't
09:43know is enormous and I was
09:44just just throwing these downs.
09:45How do we optimize the safety and effects,
09:47not just writing the prescription.
09:49What's different to the people who
09:51have success with the prescription
09:52versus people who don't?
09:54How can we understand the
09:55context of the lives?
09:55What should we be telling them behaviorally?
09:57How do we set them up for success?
09:59What does it mean between
10:00those who succeed in failure?
10:01What are the range and magnitude of benefits?
10:04Who benefits and why?
10:05Who incurs safety issues and why?
10:08Who should we prioritize?
10:09You know these trials,
10:10they they haven't included a large
10:12number of minoritized populations,
10:13They haven't includes a large
10:15number of elderly populations.
10:16They haven't included a large
10:17number of younger populations.
10:19If people are going to be
10:19on this for 10 years,
10:20none of them have gone
10:21beyond three years so far.
10:23So what what happens?
10:23What happens when people stop and start,
10:25what happens when people
10:27switch types of medications.
10:29These are real world questions that
10:30clinicians and their patients are
10:31going to need to know if they're
10:33going to be making informed choices.
10:34In the end is a cost effective.
10:36Can we make the case?
10:37Because the benefit,
10:38interestingly,
10:39in select,
10:40when people were treated with some agglutide,
10:43the benefit accrued almost immediately
10:45before you could discern the weight loss
10:47so that the curves continue to depart.
10:50But that benefit was very early.
10:52Can that manifest as a cost saving?
10:54Even people talk about
10:55this bankrupting Medicare,
10:56bankrupting the health system,
10:59but maybe it'll actually turn that on
11:01its head because of its health effects,
11:03orthopaedic procedures, cancers,
11:05as well as cardiovascular.
11:07We need to look at all this stuff.
11:09So the question will be what are the
11:11real world implications for this?
11:12What what we've got trials,
11:16very carefully selected groups that
11:19got into well curated and overseen
11:22phase three clinical trials.
11:24What happens in the wild,
11:26what happens when we're really
11:27out in the world?
11:28Who gets access and how does this work now?
11:30I wanted to present just a little
11:32bit of information at what we've
11:34been working on recently,
11:35which is to try to see how can we
11:37get within healthcare systems and
11:39be able to get real time feedback
11:41on performance and the situation
11:44around something like obesity.
11:45Now we've been working with Centara,
11:48an $8 billion healthcare system with
11:50about 22 hospitals in Southern Virginia,
11:53Northern North Carolina.
11:53It turns out our relationship with Centara,
11:55we have greater access to to healthcare
11:57data than we do in the Yale system.
12:00We, we, we actually have to go.
12:02We have to.
12:04I'm just had a Crick in my neck
12:09but it it happens. We have to go
12:10elsewhere to be able to get this.
12:11We're working hard with Daniela and
12:14Lucilla and this will be solved here
12:16and we'll soon be in the same position.
12:19But but we've been able to work with Centaur.
12:21You know it turns out if you just depend
12:23on the problem list or the ICD codes or
12:25the sort of typical structured field
12:27within within the the the medical record,
12:30you can't quite get this.
12:31But we can triangulate on this and
12:33start to see you know for example
12:36this is just looking at you know both
12:38prescription counts for semaglutide here.
12:40We're looking at the prevalence in Centaur,
12:4341% prevalence of obesity and we're
12:45looking at the use of semaglutide
12:48look at this only 2%.
12:50You know people talk about this going
12:52wild actually number total prescriptions
12:53in the country still remain far,
12:55far lower in terms of single digit
12:58percentages like under 5% for
13:00compared to the number of people
13:01who could benefit from this.
13:03So but we're able to show this,
13:04we can identify them.
13:05By the way, if this is for trial recruitment,
13:07immediately we find people,
13:08we're developing the tools so that
13:10we can use the raw data within the
13:12electronic medical record to move
13:14quickly and we can also follow people
13:16over time to say this is what they
13:18were like in in in two periods before.
13:21This is by the way you and Lou
13:22and the group at Centaur,
13:23I really want to shout out you and
13:26that you know can say that in in
13:27sort of the control period before
13:29they start on some gluttitis,
13:30the -3 negative two period zero and
13:32now you can see they're starting
13:34on it and what's their trajectory.
13:36So in the real world what are we
13:38observing and who's benefiting,
13:39who's not, who stays on it,
13:40who doesn't and what kind of health
13:42reduction do you see?
13:44Does it replicate what we see in the trials?
13:46She was showing this in even larger
13:48numbers of periods.
13:48And it what's nice about is when you
13:50start even truncating it into periods,
13:52weight happens to be something
13:54that's very commonly measured within
13:55the health record.
13:56And we can actually show what
13:57we would expect,
13:57which is the longer people were on it,
13:59the more decline.
14:00This is in body mass index.
14:02So you know,
14:03one body mass index is usually, you know,
14:05could be about 10 lbs or something.
14:06So you know this is what you
14:08might have expected from this.
14:09But just to show you we're gaining the tools,
14:11the assays,
14:12the ability to use the real world data
14:14within our own medical records to be
14:15able to ask important questions and be
14:17able to look at this kind of variation.
14:19The last thing I want to say quickly was
14:21we're spending a lot of time thinking
14:22about how AI plays a role in this.
14:24We've got these amazing new capacity
14:26now with artificial intelligence.
14:27Would be crazy not to incorporate
14:29this into our research in ways that
14:31give us entirely new perspectives.
14:33I say despite the transformative advances
14:35in medicine and with these medicine,
14:37medicine itself remains largely
14:39anchored in an older era.
14:41Our labels are antiquated.
14:42I mean just saying this is a person
14:45with obesity without talking
14:46about subclasses, sub cohorts,
14:48really getting to a precision medicine,
14:50understanding what exactly does
14:51that person in front of you have.
14:53Our treatment decisions are largely
14:55based on average effects and our
14:57prognostic methods are quite limited.
14:58AI is game changing for how we diagnose,
15:00predict and treat disease.
15:01And I think AI is going to
15:03relate to diagnosis,
15:04therapeutics and prognosis through
15:06these electronic digital signatures.
15:08So in the lab,
15:09you guys are talking about deep immune,
15:11One of the work I'm doing with Akiko,
15:13deep immune phenotyping and she's
15:15developing signatures for different
15:16people based on lab assays.
15:18What we're going to be doing now is
15:20saying like how do we take digital
15:22information that's ubiquitous and to
15:24help us understand what condition
15:25does that person have in front of us?
15:27What's the best intervention that pairs
15:29with exactly who they are and what they need?
15:31And how do we optimize the outcomes
15:33and predict and prognosticate and
15:34then modify what that prediction
15:36might be through not only the
15:38drug that we might use an example
15:40for using pharmacologic therapy,
15:41but how we surround that patient with other,
15:44you know,
15:45outcomes enhancing strategies for
15:47that particular pharmacologic agent
15:49and not really just think about all
15:51we have to do is write the script.
15:53No,
15:53it's a script surrounded by
15:55other information,
15:55particularly in a condition like obesity.
15:58And then I'm saying these data signatures
16:00are really next generation phenotypes that
16:01are going to depend on multimodal inputs.
16:04So honestly,
16:04I'm agnostic actually to what the inputs are.
16:06I mean, as an outcomes researcher,
16:08I don't care.
16:08I want to know that I've got information
16:10coming from different knowledge domains.
16:12So I can use genomic,
16:14proteomic,
16:14clinical, social,
16:15environmental and contextual information.
16:18By contextual,
16:18I mean it may be different at this
16:20health system than somewhere else.
16:21Why are we succeeding more than they are?
16:24What lessons can they glean if
16:26we're doing better than than they
16:28are independent of everything else.
16:29Just saying by the context,
16:31the way we're set up,
16:31the clinics that we have,
16:32the kind of care that we deliver.
16:35So,
16:38So the other thing is we're
16:40developing strategic partnerships
16:41with groups that have aligned values,
16:42data and dissemination channels.
16:43Some of these I hope that we'll
16:45announce relatively soon that I
16:47think they'll blow you away by the
16:49kind of alignments that we're going
16:50to make in the teams that we're
16:51going to work with who want to be
16:52able to have the same goals We are
16:53and are going to help be a force
16:55multiplier effect for our access to
16:57data and channels for dissemination.
16:59And ultimately what we care about most,
17:01impact, impact is what we care about most.
17:04So our goal is to be the preeminent
17:06obesity outcomes research group within
17:08Y wait under Anya's leadership to
17:10optimize the prevention and treatment of
17:13obesity and to improve population health.
17:15Thank you.
17:24Thank you so much, Doctor Krumholz,
17:26Questions for Harlan from the audience.
17:32I've left you spellbound, crystal
17:33clear, crystal clear.
17:35So I'll ask from lessons learned
17:37from other work that you've done.
17:39How do you think we can engage patients
17:42with obesity in this work to help
17:45us better understand their needs,
17:47their experience? What do you think?
17:50Yep. I mean, anya's alluding to the
17:52fact that, but a lot of the work
17:54that I'm doing now is trying to
17:55redesign the way that we do research
17:56in the sense of moving away from a
17:58hierarchical part where the researchers
17:59are on top and we work with subjects.
18:02I don't even use the word
18:03subjects ever anymore.
18:04I mean, I'm talking about partners,
18:06people who we guarantee that anything we do,
18:08we're going to share those
18:09results back with you.
18:10We have town in some of the other work we do.
18:12We have town halls.
18:13We give people access to the investigators.
18:15We let them ask us questions, we give them,
18:17we post them on YouTube when we're done.
18:18So the people who couldn't make
18:20that meeting can find out about
18:22the study and what we're learning.
18:23We we, we really push this agenda
18:26of saying you're our partners,
18:28you know,
18:29we're working together in common
18:30cause no one has more motivation about
18:31trying to find answers than people
18:33who are affected by the conditions.
18:34But so you tell me how is it that we
18:36lose people in trials otherwise is
18:37that people are lost to follow up.
18:39They lose interest. They just follow up.
18:41It's because they get alienated.
18:42They don't feel as if we're actually
18:44attentive to them.
18:45To me,
18:45my goal is that everybody's in any studies,
18:48ours is delighted by the experience.
18:50Will brag to their friends about
18:51how good it was and try to tell
18:53others they would do it again.
18:54And so that means that we constructed
18:55in a way that the any advances we make
18:57are ones that they can also feel good about.
19:00They can talk about it at the dinner table.
19:01They can they can recognize that we
19:03honor and respect their contribution
19:05that we we we guarantee that we're
19:07going to tell them what we learn and
19:08we're going to give them the credit
19:10that they deserve for taking the time to
19:11work with us to be able to do the work.
19:13So I I think the people with
19:14obesity is a prime group to be able
19:16to pull pull in and learn from.
19:18By the way,
19:19I want to say with humility that it's
19:20not just that you do this because it's a
19:22good strategy to keep people in studies.
19:24It's a smart strategy if you want to
19:26be a good researcher because there's
19:28wisdom that resides in people who live
19:30with the conditions and we'd be well,
19:32well served to to humbly learn from them
19:35when they've got things to tell us.
19:38I wholeheartedly agree. Well,
19:40thank you for that wonderful talk,
19:43Harlan, and we are going to move forward
19:46with our final speaker for the day.