Changes in Medication Use After the Inflation Reduction Act: A Difference-in-Differences Analysis

Transcript

• 00:02Okay. Good afternoon, everyone. Let's
• 00:04get started. Welcome to the
• 00:06last
• 00:07research in progress, meeting for
• 00:09twenty twenty five. It's great
• 00:10to see everyone here and
• 00:13online.
• 00:15Okay.
• 00:17So
• 00:20Yeah.
• 00:21So I just press,
• 00:24Okay.
• 00:26Okay. The CME code for
• 00:28today's meeting is five five
• 00:29six five eight. Five five
• 00:31six five eight.
• 00:34Just a reminder of our
• 00:35re retreat schedule. We just,
• 00:37completed our research and scholarship
• 00:39retreat
• 00:40a couple weeks ago. Have
• 00:42your calendars marked for February
• 00:43sixth
• 00:44for the professional development retreat
• 00:46and then the education retreat
• 00:49in May. It was great
• 00:50seeing everyone at the research
• 00:51and scholarship retreat.
• 00:53Thanks again, to Carrie Gross
• 00:55and Jeanette Tetreault for organizing
• 00:57what was really a spectacular
• 00:58day
• 00:59of learning for everybody.
• 01:02This is our weekly reminder
• 01:03this time of year for
• 01:04the FDAQ process. So,
• 01:07we're in step one.
• 01:09I I've seen a number
• 01:10of these,
• 01:11completed FDA queue documents come
• 01:13in.
• 01:14Please,
• 01:15get those done. The deadline
• 01:17is on the slide there,
• 01:18February second. So you do
• 01:19have time.
• 01:21But it's just a matter
• 01:22of sitting down and doing
• 01:23it as you all know,
• 01:25And the rest of the
• 01:25steps are outlined
• 01:27on the slide.
• 01:29Thanks to all who attended
• 01:31our annual,
• 01:32section holiday party. It was
• 01:34a lot of fun, a
• 01:35lot of food. We actually
• 01:36ran out of food for
• 01:37the first time, but that's
• 01:39okay. That means,
• 01:40people were well fed. And,
• 01:42of course,
• 01:43we have our
• 01:45we had our ultrasounds there.
• 01:53It's that melody there.
• 01:57That's hard.
• 01:59So,
• 02:01thanks. The ultrasounds.
• 02:05Yes. These are this is
• 02:06a medical student acapella group.
• 02:09Now we have a lot
• 02:10of acapella groups here at
• 02:11Yale, including at the medical
• 02:12school. So these guys did
• 02:14a great job.
• 02:19Okay.
• 02:21So,
• 02:22for the next two weeks,
• 02:24due to, the Christmas and
• 02:26New Year's holidays, we will
• 02:27not have meetings.
• 02:29So happy holidays,
• 02:31to everyone, and, of course,
• 02:32a happy New Year.
• 02:35And then we'll resume our,
• 02:37meetings on January eighth.
• 02:39Kelsey Bryant will be speaking
• 02:41at our general medicine grand
• 02:42rounds at seven thirty.
• 02:44I read the hypertension guidelines,
• 02:46so you don't have to.
• 02:48That'll be good.
• 02:49And then Jen Miller,
• 02:51will be here. Jen is
• 02:52here somewhere.
• 02:53Yes. There she is. Hi,
• 02:55Jen.
• 02:56She will be speaking on
• 02:57transparency, access, and ethics in
• 02:59the pharmaceutical c sector,
• 03:01an update on conceptual development
• 03:03and performance
• 03:04tracking
• 03:04with the good pharma scorecard.
• 03:07So that's
• 03:08next,
• 03:09January eighth, at seven thirty
• 03:11and at noon. Be there.
• 03:15Here's our usual disclosure slides.
• 03:17And we have a special
• 03:19out of town guest today.
• 03:20I'll have Carrie Gross
• 03:22introduce our speaker. Carrie.
• 03:28Thank you. I I will
• 03:29not be providing the harmonies,
• 03:31but we'll be providing a,
• 03:33a welcome to our guest,
• 03:35Chris Kai. So,
• 03:37Chris comes to us, well,
• 03:39now comes to us from,
• 03:41Boston and the Brigham where
• 03:42he's currently a general medicine
• 03:44fellow.
• 03:45And, I'm
• 03:46excited for a number of
• 03:47reasons to have you here.
• 03:48First of all,
• 03:50just reading your work over
• 03:52the years and actually,
• 03:53within JAMA Internal Medicine, being
• 03:55able to work with you
• 03:56on some of your manuscript
• 03:57submissions and seeing the,
• 03:59terrific, innovative, and impactful work
• 04:01you've been doing,
• 04:04leaves me really,
• 04:06excited about meeting you and
• 04:08hearing your talk today.
• 04:10Those some of you may
• 04:11remember that the original,
• 04:13lecture was supposed to be
• 04:14six weeks ago, but,
• 04:16Chris's
• 04:17baby Ezra
• 04:19had other plans. It was
• 04:20delivered pretty much, like, exactly
• 04:22at the same time as
• 04:23the the new conference was
• 04:24scheduled for.
• 04:25So,
• 04:27and kudos to
• 04:28Ezra for making his voice
• 04:30heard.
• 04:32Chris began his education,
• 04:35in Virginia. He's been out
• 04:36at UCSF for medical school
• 04:38back here to the Brigham
• 04:39for residency and,
• 04:41on the East Coast, I
• 04:42should say, for residency and
• 04:44for fellowship.
• 04:45And we're really excited to,
• 04:47welcome you here to Yale
• 04:48to talk about changes in
• 04:50medication use after the Inflation
• 04:52Reduction Act,
• 04:54difference in differences analysis.
• 04:56Thank you.
• 05:01Okay.
• 05:04So thank you so much
• 05:05for a very warm introduction.
• 05:07I'm thrilled to be here,
• 05:09and, I welcome,
• 05:11any, you know, questions or
• 05:13feedback and interruptions. So, just
• 05:15really excited to chat with
• 05:16you all. So thank you
• 05:17for having me.
• 05:19So I have about thirty
• 05:21five to forty minutes of
• 05:22material today,
• 05:24and,
• 05:25breaking it down. We'll spend
• 05:26a little bit about my
• 05:27personal background,
• 05:29review some of the prior
• 05:31work I've done as a
• 05:32resident and a fellow,
• 05:33and then
• 05:35spend about twenty five minutes
• 05:36on,
• 05:37some papers on the Inflation
• 05:39Reduction Act, which, is a
• 05:41big piece of legislation that
• 05:43totally changed the way we
• 05:45pay for prescription drugs for
• 05:46people with Medicare,
• 05:48who represent about fifty million
• 05:50people,
• 05:51in the US.
• 05:53And then I'll end with
• 05:54the preview of what I
• 05:56hope to do as a
• 05:57junior faculty member, which focuses
• 05:59on on a k award
• 06:01application to the National
• 06:03Institute on Aging.
• 06:05I wanna thank everyone who
• 06:07made this visit possible.
• 06:09My son came a little
• 06:11early,
• 06:12on the day I was
• 06:13supposed to be here, and
• 06:14so he's very excited. He's
• 06:15I know he's exploring
• 06:16New Haven with my wife
• 06:18right now, and so, these
• 06:19are some,
• 06:21photos
• 06:22to you know,
• 06:24that I just had to
• 06:24share. You know?
• 06:28So, just for background,
• 06:31you know, I grew up
• 06:32as the son of two
• 06:33Chinese immigrants, and this really
• 06:35was tremendously influential.
• 06:38I got to see what
• 06:39it was like to be
• 06:40a new American,
• 06:42some of the wonderful things
• 06:44that come with that, and
• 06:45of course, some of,
• 06:47the challenges that come with
• 06:48that.
• 06:49I remember growing up seeing,
• 06:51my parents went through periods
• 06:53where they didn't have health
• 06:54insurance. And even as a
• 06:55young kid, you begin to
• 06:57wonder, Oh, wow, like what
• 06:58how is this affecting, their
• 07:00lives? And I remember one
• 07:01story I'll share.
• 07:03My mom had a jaw
• 07:04tumor where she really delayed
• 07:06care because she didn't have
• 07:07health insurance. And,
• 07:09from a young age that
• 07:10got me interested into what,
• 07:12you know, what was going
• 07:12on and and why,
• 07:14why why these larger systems,
• 07:17existed.
• 07:19I did medical or I
• 07:20did undergrad at the University
• 07:22of Virginia, and I didn't
• 07:23really know what I wanted
• 07:24to study. I,
• 07:26you know, basically,
• 07:27did a major that really
• 07:28was a lot of anthropology
• 07:29classes.
• 07:30And I knew that I
• 07:32wanted to go into medicine,
• 07:33but wasn't sure how to
• 07:34bring it all together.
• 07:36And it wasn't until medical
• 07:37school, I, did a a
• 07:39special program at my medical
• 07:41school at UCSF,
• 07:42where you do your rotations
• 07:43at the county hospital. And,
• 07:46I began to say, okay,
• 07:47this is, like, very interesting.
• 07:49Then I got to see
• 07:50firsthand,
• 07:51how important health insurance is
• 07:53to,
• 07:54clinical outcomes. And that was
• 07:56really when I decided I'd
• 07:58like to go down this
• 07:59road of primary care and,
• 08:01health policy.
• 08:05At the time, Medicare for
• 08:06all was really at the
• 08:07national forefront of the policy
• 08:10discussion, and so I was
• 08:11motivated to try to understand
• 08:13a little bit the larger
• 08:14structural forces which,
• 08:16influenced patient care. And so,
• 08:19I spent a summer working
• 08:20in the House of Representatives
• 08:22and working on the Medicare
• 08:23for all bill. And my
• 08:24goal was to really bring
• 08:26an academic lens to a
• 08:28health care debate, which was
• 08:29really being politicized.
• 08:30And I, we we did
• 08:32studies which found
• 08:34the legislation
• 08:34could, reduce national health expenditures
• 08:37and increase resources for primary
• 08:39care. And,
• 08:40we we published those in
• 08:41journals and shared them with
• 08:42lawmakers. And this was the
• 08:44first time where I I
• 08:46thought, oh, wow. Like, if
• 08:47doing research, you can play
• 08:49a small part in the
• 08:50larger health policy debate. And
• 08:52that was very, exciting for
• 08:53me.
• 08:55I did a residency in
• 08:57internal medicine with my now
• 08:58wife, and this is one
• 08:59of the first pages that,
• 09:01we were able to send,
• 09:03to each other.
• 09:04And I think it's really
• 09:05appropriate that it's an FYI
• 09:06page. You know, you have
• 09:07to be appropriate on the
• 09:08level of triage for your
• 09:10pages. So,
• 09:11I'm glad that I, was
• 09:13able to do that, day
• 09:14one of intern year.
• 09:17I did,
• 09:18my, continuity clinic as a
• 09:20resident was at FQHC,
• 09:23and,
• 09:24federally qualified health center in
• 09:25Dorchester.
• 09:27And this was something that
• 09:29was really attractive to me
• 09:30when trying to figure out
• 09:31where I wanted to train.
• 09:32And,
• 09:33I just remember
• 09:34rotating, you know, in the
• 09:35coronary,
• 09:37academic medical center and then
• 09:38driving thirty minutes. And you
• 09:40meet patients who have never
• 09:41seen a physician and who
• 09:42have, you know, come into
• 09:44you with hemoptysis and are
• 09:46refugees from other countries. And
• 09:48it's a very humbling experience
• 09:49and really cemented my interest
• 09:51in primary care and,
• 09:53working in the social safety
• 09:55net.
• 09:57And so through these experiences,
• 09:59I've,
• 10:01zeroed in on what I
• 10:02hope is the motivating question
• 10:03for my work, which is
• 10:04how can we reform the
• 10:05way we pay for health
• 10:06care,
• 10:07in the service of promoting
• 10:09justice, efficiency,
• 10:10and patient outcomes?
• 10:12And I really see Medicare
• 10:14as the area that I
• 10:15would like to focus on
• 10:17the first few years as
• 10:17a junior faculty member.
• 10:20And though I've branched out
• 10:21and looked at other aspects
• 10:23of healthcare financing and payment
• 10:24reform, like private equity and,
• 10:27the generic drug marketplace.
• 10:30And in terms of methods,
• 10:31as a resident and a
• 10:33medical student, I really focused
• 10:34on descriptive methods, like time
• 10:37trends, systematic reviews,
• 10:39and as a fellow, have
• 10:40tried to gain more skills
• 10:41in quasi experimental methods,
• 10:44including pharmacoepidemiology
• 10:46and and, techniques from the
• 10:48econometrics,
• 10:49literature, like difference in differences.
• 10:53So hoping to spend a
• 10:54few minutes just talking about
• 10:55some of the work I've
• 10:56done and then,
• 10:57dive more deeply into a
• 10:59study that I think has
• 11:00a lot of clinical relevance,
• 11:01on the Inflation Reduction Act.
• 11:05So as a resident,
• 11:06I did a series of
• 11:08studies looking at racial disparities
• 11:09and, high value diabetes medications,
• 11:13finding that,
• 11:14unsurprisingly,
• 11:16minority patients are less likely
• 11:18to receive SGLT twos and
• 11:20GLP ones,
• 11:22which could be because of
• 11:23lower visit rates to, specialist,
• 11:26physicians who at the time
• 11:27were much more likely to
• 11:28prescribe these medications.
• 11:31Building on my prior interest
• 11:33in Medicare policy,
• 11:35did a series of studies
• 11:36with, colleagues and, looking at
• 11:39the receipt of supplemental benefits,
• 11:41dental, vision, and hearing in,
• 11:43Medicare Advantage,
• 11:45which is a very important
• 11:46policy debate because Medicare Advantage
• 11:49is paid more to provide
• 11:50these benefits. And we found
• 11:51that,
• 11:52in a series of studies
• 11:53that patients don't actually receive
• 11:55those benefits at a higher
• 11:56rate.
• 11:57And so this has implications
• 11:59for the way we pay
• 11:59for, Medicare Advantage.
• 12:03And as a fellow, really
• 12:04focused on prescription drug policy
• 12:06work.
• 12:07So,
• 12:08we looked at a program
• 12:10that would have capped out
• 12:11of pocket costs to two
• 12:12dollars for generic medications,
• 12:15and found that it would
• 12:16only save about ten dollars
• 12:17annually,
• 12:18for thirty eight percent of
• 12:19Medicare beneficiaries.
• 12:22And we use this to,
• 12:23we shared the results with
• 12:25CMS,
• 12:26Center for Medicare and Medicaid
• 12:27Services Administrators. And we tried
• 12:29to use this data to,
• 12:31generate more evidence based policy
• 12:33proposals and to make the
• 12:35policy more generous.
• 12:38And finally,
• 12:40looking at one one another
• 12:41analysis we did was to
• 12:42look at, the patent,
• 12:44marketplace,
• 12:45and we found that we
• 12:47looked at one case example
• 12:48of doxepin, which is a
• 12:50medication for sleep,
• 12:52a first line medication recommended
• 12:54by sleep societies.
• 12:56And we looked at a
• 12:56very interesting scenario where despite
• 12:58being a fifty year old
• 12:59medication,
• 13:01a dose of Doxepin was
• 13:03actually patented, which is a
• 13:04very unusual case, where a
• 13:06generic medication was patented because
• 13:08of a lower dose. And
• 13:11we showed that this patent
• 13:12likely resulted in much higher
• 13:13prices
• 13:14and,
• 13:16likely limited access to a
• 13:18first line medication
• 13:19for sleep
• 13:20and led to a lot
• 13:21of excess spending in addition
• 13:23to that.
• 13:26You know, one goal that
• 13:27I had as a fellow
• 13:28was to try to build
• 13:29skills in pharmacoepidemiology,
• 13:31and target trial emulation
• 13:33and propensity score weighting,
• 13:36and to try to get
• 13:37more at causal questions. And
• 13:38so
• 13:40to address that,
• 13:42we did a new user
• 13:43analysis.
• 13:44And the basic question for
• 13:46this study was looking at,
• 13:48does private equity, which are
• 13:49investment firms, which,
• 13:51have been purchasing hospitals, clinics,
• 13:53and nursing homes,
• 13:55do they contribute to a
• 13:56higher risk of bankruptcy or
• 13:58closure?
• 13:59And you may have seen
• 14:00in the news the closure
• 14:01of hospitals like Henneman Hospital
• 14:03or Stewart Healthcare in Boston
• 14:05and Philadelphia.
• 14:07And if you talk to
• 14:07people at PE firms, they
• 14:09say, well, we really we
• 14:10buy hospitals that are going
• 14:12to close anyway. We focus
• 14:13on distressed assets.
• 14:15And to try to address
• 14:16that potential confounding, we did
• 14:18a new user analysis where
• 14:20we looked at healthy hospitals,
• 14:22clinics, and nursing homes, which
• 14:24had good finances. And we
• 14:25compared them to
• 14:27other hospitals, clinics, and nursing
• 14:28homes that,
• 14:30had very similar, finances, but
• 14:31were purchased by other,
• 14:34for profit firms that were
• 14:36for profit, but not private
• 14:37equity.
• 14:38And we, addressed potential confounding
• 14:41using propensity score weighting.
• 14:44And using this this new
• 14:45user active comparator framework, we
• 14:47found that the private equity
• 14:49acquired
• 14:50healthcare organizations were about three
• 14:52times higher,
• 14:53more likely to experience,
• 14:55the prespecified
• 14:56composite outcome of, bankruptcy or
• 14:58closure.
• 14:59So we hope that this
• 15:00contributes to literature and, ongoing,
• 15:03congressional legislation, which would regulate,
• 15:06these investment firms, which,
• 15:08could potentially be contributing to
• 15:10a higher risk of bankruptcy
• 15:11or closure.
• 15:14So a little bit about
• 15:15my background and and now
• 15:17hoping to focus on, what
• 15:19I think is gonna be
• 15:20the the focus of my
• 15:22early years as a FACT
• 15:23team member, and that's Medicare
• 15:24policy and specifically
• 15:26looking at the Inflation Reduction
• 15:28Act.
• 15:30So as many of you
• 15:31may know, this is a
• 15:33the Inflation Reduction Act is
• 15:34a really big piece of
• 15:35legislation. It was passed in
• 15:37twenty twenty two.
• 15:38It touched many parts of
• 15:40the economy.
• 15:41It took a lot of
• 15:42political will to get across
• 15:43the finish line. And so,
• 15:46important to understand what it
• 15:47did it did not do,
• 15:49and also to,
• 15:51make sure we get it
• 15:51right, because it's probably the
• 15:53largest piece of health care
• 15:54legislation since the Affordable Care
• 15:56Act.
• 15:57And so we'll be focusing
• 15:59on Medicare Part D, which
• 16:00is the prescription drug benefit
• 16:02for, that the Inflation Reduction
• 16:04Act, reformed.
• 16:06And just for background, Medicare
• 16:08Part D, it's a voluntary
• 16:09prescription drug benefit. And so
• 16:11when you become eligible for
• 16:12Medicare, you can buy
• 16:14insurance to cover your prescription
• 16:16drugs, and fifty million people
• 16:18in the country
• 16:19have some sort of Medicare
• 16:21part d.
• 16:22This is a public benefit.
• 16:24We all pay for it
• 16:25with our tax dollars, and
• 16:26it is administered through private
• 16:28plans.
• 16:29And there are two types
• 16:30of private plans. You can
• 16:31get Medicare Advantage, which provides
• 16:33wraparound care, inpatient,
• 16:35outpatient,
• 16:37and prescription drug, or you
• 16:39can get traditional Medicare and
• 16:41supplement it. So a little
• 16:42complicated, but they're all private
• 16:44plans,
• 16:45and beneficiaries
• 16:46shop for them through,
• 16:48a regional marketplace.
• 16:50And so that is the
• 16:52background. And what was the
• 16:53problem the legislation was aiming
• 16:55to address, which, well, that
• 16:56is Medicare patients historically have
• 16:58paid very high out of
• 17:00pocket costs if they took
• 17:01expensive drugs.
• 17:02And so it's a very
• 17:03complicated benefit historically.
• 17:06There was the deductible,
• 17:08there was initial coverage, and
• 17:10then somehow it got worse
• 17:11after a while. You may
• 17:12have heard of the donut
• 17:13hole, and then there was
• 17:14the catastrophic phase.
• 17:16The bottom line is that
• 17:18historically,
• 17:19costs were not capped. And
• 17:20so you could pay ten,
• 17:22fifteen, twenty thousand dollars out
• 17:23of pocket if you were
• 17:25unfortunate enough to, you know,
• 17:26be diagnosed with cancer or
• 17:28if you had rheumatoid arthritis,
• 17:29if you were taking an
• 17:30expensive medication.
• 17:33And you paid five percent
• 17:34indefinitely,
• 17:35even in the most generous
• 17:36phase, the catastrophic phase.
• 17:38And so that's a lot
• 17:40of wonky information, a very
• 17:42complicated benefit. And I think
• 17:43it's really helpful to illustrate,
• 17:45this with some patients who
• 17:46you, likely have seen. So,
• 17:49we'll take Mr. A on
• 17:50the left. He's a seven
• 17:51year old man with hypertension,
• 17:53atrial fibrillation, hyperlipidemia.
• 17:55And then, Mrs. B, same
• 17:58conditions, but she has chronic
• 17:59lymphocytic leukemia.
• 18:01And here are the medications
• 18:03they take. These are likely
• 18:04gonna look familiar to you.
• 18:06And then the monthly costs.
• 18:08And you'll see right away,
• 18:09for mister a, the majority
• 18:11of the monthly price, the
• 18:13prices,
• 18:14are absorbed by apixaban or
• 18:16Eliquis brand name.
• 18:18And then for missus b,
• 18:19she's taking the same medications,
• 18:21but now she's taking imbrutinib,
• 18:22which is a very costly
• 18:24medication effective, but very costly,
• 18:26for her CLL.
• 18:28And so if those are
• 18:30the prices and if you
• 18:31just do the algebra that
• 18:32we saw on the last
• 18:33slide, the complicated benefit, this
• 18:35is how much they would
• 18:35pay per year. So Mr.
• 18:37A is paying about six
• 18:38hundred per year and then,
• 18:39oh, wow, Mrs. Buchi's paying
• 18:41about fifteen thousand.
• 18:42And that's because you take
• 18:43the price of Imbrutinib
• 18:45and you just run it
• 18:46through the four phases.
• 18:47And because it's a two
• 18:49hundred thousand dollars a year
• 18:50medication, even if you pay
• 18:51a small amount, dollars five
• 18:52percent, it just starts to
• 18:53really add up. And so
• 18:55this is really the problem
• 18:56the law zeroed in on.
• 18:57We have these patients
• 18:58like Mrs. B who are
• 19:00paying untenable amounts for their
• 19:01medications.
• 19:03And so what did the
• 19:04law do? It imposed an
• 19:06out of pocket cap. And
• 19:07so in twenty twenty four,
• 19:09that cap was thirty three
• 19:10hundred dollars, and then it
• 19:11was lowered to two thousand
• 19:12dollars in twenty twenty five.
• 19:14And an important feature of
• 19:16the law was that these
• 19:17private plans were now tasked
• 19:19with paying a large proportion
• 19:21of that. And so now
• 19:22these private plans, you go
• 19:24on a marketplace, they have
• 19:25to pay. It's about sixty
• 19:26percent of all costs above
• 19:28the cap. And so we're
• 19:30asking these private plans, like,
• 19:31to in a in a
• 19:32noble fashion to to pay
• 19:33more to cover patients like
• 19:35missus Speed.
• 19:36And so this is a
• 19:38brand new benefit. Twenty twenty
• 19:39five is the first time
• 19:40this two thousand dollar out
• 19:42of pocket cap was imposed.
• 19:43And so,
• 19:45some studies that my colleagues
• 19:46and I have have attempted
• 19:47to do was to assess
• 19:49insurer responses. And so we
• 19:51know that these are private
• 19:52entities. We know that they
• 19:53are likely
• 19:54to respond in a fairly
• 19:56predictable fashion.
• 19:57Number one, could they leave
• 19:58the marketplace?
• 20:00Could they raise deductibles or
• 20:01the amount that you have
• 20:03to pay as a patient
• 20:04before the plan kicks in?
• 20:06Could they modify a formulary?
• 20:07Could they make it more,
• 20:08you know, more difficult or
• 20:09more costly to access these
• 20:11medications?
• 20:12Or premiums, the the gym
• 20:13membership, the monthly peep payment
• 20:15you have.
• 20:16With the asterisk there that
• 20:18the law really prevented that
• 20:19from happening, they subsidized the
• 20:21premiums.
• 20:22So we didn't think that
• 20:23was really going to happen.
• 20:25And so that's objective number
• 20:26one of my fellowship, a
• 20:28series of studies looking at
• 20:29what did the insurers do
• 20:30when they had to provide
• 20:32much more generous coverage.
• 20:34We used publicly available data,
• 20:37from the Medi Center for
• 20:38Medicare and Medicaid Services, and
• 20:39we looked at all, of
• 20:41these private plans.
• 20:42And consistent with those hypothesized
• 20:45insurer responses, we calculated,
• 20:47the,
• 20:48the enrollee weighted and inflation
• 20:50adjusted,
• 20:51percent of people who were
• 20:52affected by their insurer totally
• 20:54leaving,
• 20:55out of pocket costs for
• 20:56nine top selling medications.
• 20:59We looked at whether a
• 21:00medication was paid by a
• 21:01co payment, which is typically
• 21:02a flat amount, thirty to
• 21:04forty dollars, or coinsurance,
• 21:05which is typically a percent
• 21:07of a drug's cost,
• 21:08with the idea that coinsurance
• 21:10is typically much more because
• 21:11it's, you know, thirty percent
• 21:12of a thousand or five
• 21:14hundred. So it just ends
• 21:15up being much more.
• 21:16And then we looked at
• 21:17deductibles and premiums.
• 21:20We looked at different subgroups
• 21:21as well, and this was
• 21:22a partially descriptive analysis.
• 21:26And so on the x
• 21:27axis here, you have year.
• 21:29On the y axis, you
• 21:30have the percent of people
• 21:32who were affected by their
• 21:33insurance plan just totally leaving,
• 21:35the marketplace.
• 21:37And, you know, bef
• 21:39a priority that could be
• 21:40a really bad thing. Right?
• 21:41It leads to disruptions in
• 21:43coverage.
• 21:44This is a voluntary benefit,
• 21:45so people might not even
• 21:46know. They may, you know,
• 21:48basically, in January, they may
• 21:49be like, oh, I don't
• 21:50have health insurance anymore for
• 21:51my medications.
• 21:53And you can see that
• 21:54essentially over time, consistent with
• 21:56the implementation of the law,
• 21:58more and more people are
• 21:59affected by their insurer just
• 22:00leaving. And so by twenty
• 22:02twenty four, the most recent,
• 22:03year of data we have
• 22:04available,
• 22:06eight percent of people,
• 22:07are just
• 22:08in that year,
• 22:10they had no prescription drug,
• 22:12insurance, that their insurer just
• 22:13totally left. And so unless
• 22:15they took the effort to
• 22:17go to open enrollment and
• 22:18to find another insurer, they
• 22:20they no longer had, prescription
• 22:22drug, coverage.
• 22:24This, unfortunately, was consistent too
• 22:26in different, subgroups of insurance
• 22:28plans.
• 22:30We looked at deductibles. And
• 22:32I I think on the
• 22:33x axis here, you see
• 22:34year, and on the y
• 22:35axis, the inflation adjusted deductible.
• 22:37And I think the key
• 22:38thing to highlight is there's
• 22:39a huge jump for Medicare
• 22:40Advantage plans, which were charging
• 22:42about sixty six dollars on
• 22:44average in twenty twenty four,
• 22:45which was trending down. And
• 22:47then, there's a large increase
• 22:48to twenty twenty five when
• 22:49the two thousand dollar cap
• 22:51was implemented,
• 22:52about a hundred and sixty
• 22:53dollar increase.
• 22:55In the stand alone plans,
• 22:56a more gentle increase that
• 22:58really predated when the law
• 22:59was implemented.
• 23:02Here are nine top selling
• 23:03medications, and we we assess
• 23:05what the monthly out of
• 23:06pocket cost would be.
• 23:08And you can see these
• 23:09are high value medications. They're
• 23:11used to treat diabetes,
• 23:13atrial fibrillation.
• 23:16They they
• 23:17CKD, and they have very
• 23:19high clinical value.
• 23:21And you can see here
• 23:22on,
• 23:23the x axis is year
• 23:25again, and the y axis,
• 23:26the monthly costs. And unfortunately,
• 23:28a similar trend where the
• 23:29costs are increasing in twenty
• 23:31twenty five in both in
• 23:33both kinds of plans by
• 23:34about thirty three percent.
• 23:38This is basically explaining what
• 23:40caused,
• 23:41these costs to go up.
• 23:42This is the percent of
• 23:43people who paid coinsurance, which
• 23:45is that percent of a
• 23:46drug's cost.
• 23:47And and, essentially, that is
• 23:49what's explaining it. People are
• 23:50shifting from these thirty dollar
• 23:52co payments per month for
• 23:53brand name medications to a
• 23:55coinsurance, which could be hundreds
• 23:56of dollars per month.
• 24:00And just to say it
• 24:00briefly, we saw that the
• 24:02premiums decreased as expected, and
• 24:04they were subsidized by by
• 24:06the legislation.
• 24:08So a lot of data
• 24:10there, and I think illustrative
• 24:11to take our two example
• 24:12patients and run them through
• 24:14those numbers and, and see
• 24:16what the net effect is.
• 24:18As I also think it's
• 24:19helpful to know that the
• 24:21government estimates only about twenty
• 24:22percent of people are gonna
• 24:23hit the cap. And so
• 24:24most people are like Mr.
• 24:26A, They're taking maybe one
• 24:27brand name medication and maybe
• 24:29a few generic medications. But,
• 24:32but all about twenty percent
• 24:33of people are, like, gonna
• 24:34be like, missus b.
• 24:36And when you do the
• 24:37math, this is sort of
• 24:38the net effect of the
• 24:39law in twenty twenty five.
• 24:41Both patients,
• 24:42experience
• 24:43increases in deductibles.
• 24:45But the key is that
• 24:46patients like mister a are
• 24:48not hitting the cap, whereas
• 24:50patients like missus b are.
• 24:52And so at the end
• 24:53of the day, we find
• 24:55that if you if based
• 24:56on these averages,
• 24:58if you're a patient like
• 24:59mister a, you're not hitting
• 25:00the cap, you may pay
• 25:01actually more under this legislation,
• 25:03whereas patients like missus b
• 25:05are paying less.
• 25:06Yeah.
• 25:08Thank you, Chris. So twenty
• 25:10percent of beneficiary
• 25:12are
• 25:13on the visa.
• 25:16And generally speaking, the purpose
• 25:18of insurance
• 25:19is that
• 25:21the way I put it,
• 25:22but the healthy people are
• 25:23a subscriber
• 25:24of the of the.
• 25:26So if we institute a
• 25:27new shield, which is,
• 25:30we're saying that
• 25:47does the out of pocket
• 25:48cost from the eighty percent
• 25:50of the fifty percent, would
• 25:51that totally,
• 25:53make up for the reduction
• 25:56in patient responsibility?
• 25:59Like, Like, is this
• 26:00all just a complete unintended
• 26:03consequences for it now? The
• 26:04beneficiary
• 26:05are just completely
• 26:07paying the full amount of
• 26:08additional,
• 26:10additional coverage, or, conversely, is
• 26:13this, like, more general tax
• 26:14dollars?
• 26:17Yeah. I think it's a
• 26:18really good question.
• 26:21You know, I think you're
• 26:22totally right. That's, like, the
• 26:24social purpose of insurance to
• 26:26spread risk. And so, you
• 26:28know, in some ways, you
• 26:29could say, like,
• 26:30this is sort of why
• 26:32we have insurance, and, like,
• 26:33maybe this is not a
• 26:34bad thing.
• 26:35I think that
• 26:37there could be an argument
• 26:38that, like, the sum is
• 26:39is less than their parts
• 26:41because what we see is,
• 26:43and it's not shown here,
• 26:44but because of a lot
• 26:46of plans are leaving the
• 26:47marketplace,
• 26:48we have a follow-up study
• 26:49that shows that
• 26:50the lack of competition leads
• 26:52to way less generous plans.
• 26:54And so,
• 26:55that is sort of what
• 26:56I mean by the sum
• 26:57is not just you know,
• 26:58like, I think the amount
• 26:59of money that is going
• 27:00to the ultimately trickling down
• 27:02to beneficiaries could be less
• 27:03because competition is is is
• 27:05less is one way to
• 27:06address that.
• 27:08And then the other piece
• 27:09of it is that,
• 27:10you know, that eighty percent
• 27:11number is based on estimates,
• 27:13historical data. But one worry
• 27:15I have is I see
• 27:16patients who are like Mr.
• 27:18A, but they,
• 27:20or I'm sorry, are like
• 27:21Mrs. B, but she is
• 27:22not there. They're not hitting,
• 27:23they're not, they don't have
• 27:24two thousand dollars to pay.
• 27:26So they may eventually hit
• 27:27the cap if, if they
• 27:28just fill their medications, but
• 27:30they see a higher deductible
• 27:31and they go to the
• 27:32pharmacy and they're like, why,
• 27:34woah, my medications cost way
• 27:35more. And you want to
• 27:37say to them, I know,
• 27:38but if, but in March,
• 27:39you're gonna hit the cap.
• 27:40And so you have to
• 27:41pay through, unfortunately.
• 27:43But I think we're gonna
• 27:44see a scenario where because
• 27:45the deductibles went up, that
• 27:47discourages people from filling their
• 27:49medications. And that ultimately may
• 27:51mean that fewer people than
• 27:52expected
• 27:53benefit from the law is
• 27:54one worry I have. I
• 27:55don't know if that is
• 27:56gonna bear out, but that,
• 27:57I think, could be a
• 27:58potential consequence.
• 27:59You're gonna measure that?
• 28:01I think you really just
• 28:02have to go back. And
• 28:03so the government that eighty
• 28:05percent number is from, like,
• 28:06I think, twenty nineteen data.
• 28:08Just thinking that I think
• 28:09is a very important Yeah.
• 28:11That do you have any
• 28:12way of detecting?
• 28:14Yeah. That's a good question.
• 28:15You know, one I'll show
• 28:17you some data that suggests
• 28:19that, in January twenty twenty
• 28:20five, fills go down a
• 28:22little bit, but then they
• 28:22quickly recover. So that
• 28:25we'll have to really wait
• 28:25for the claims, the person
• 28:27level data. But in population
• 28:28level data, we see that
• 28:30the fills actually
• 28:31decrease, which is concerning.
• 28:36Fine. I'll I'll first, two
• 28:38questions. First,
• 28:39please repeat questions from the
• 28:40audience so that Oh, yes.
• 28:42Okay. Great.
• 28:44Microphone's on, so we should
• 28:45be good.
• 28:46And second,
• 28:47I thought
• 28:48is for well, David Rosenthal.
• 28:50Do you wanna just jump
• 28:51on?
• 28:52Yeah. Sure. Thank you so
• 28:53much. I just, ibrutinib is
• 28:54one of the drugs that's
• 28:55gonna, I think, adjust, correct,
• 28:56in twenty twenty six. I'm
• 28:57just curious how that's gonna
• 28:59change these mister a, mister
• 29:01b out of pocket costs
• 29:02or your your modeling.
• 29:04You're totally right. So, yeah,
• 29:05it it is it's one
• 29:07of the medications, as you
• 29:08point out, that is subject
• 29:09to price negotiation. So
• 29:11a whole separate arm of
• 29:13the legislation empowered Medicare to
• 29:15negotiate prices. And so
• 29:18I think that it hopefully,
• 29:21on the one hand, it
• 29:22it won't really make a
• 29:22difference for missus b. She's
• 29:23gonna pay two thousand dollars
• 29:25anyway, but I think the
• 29:26theory is that, okay, the
• 29:28medication costs less, the insurers
• 29:29pay less, and then eventually,
• 29:31that will help other patients.
• 29:34Only a select group of
• 29:35medic of medications are,
• 29:37eligible for negotiation.
• 29:38So I think a bigger
• 29:40picture takeaway from this data
• 29:41is that we really should
• 29:43be empowering Medicare to negotiate
• 29:44all drugs. Right? And that
• 29:46that is the way to
• 29:47have your cake and eat
• 29:48it too. Right? We're we
• 29:49know of we're trying to
• 29:50get people to take more
• 29:51medications. And if you don't
• 29:53want cost to go up,
• 29:54you should allow the government,
• 29:55like any other payer, to
• 29:57negotiate prices.
• 30:01Just to support that point,
• 30:03the VA has been able
• 30:04to negotiate prices for many
• 30:05years and has effectively halved
• 30:07the cost. Yeah. So I
• 30:09totally agree. That's a very
• 30:10good point. Yeah. We have
• 30:12a very nice, you know,
• 30:13health system to look to
• 30:14for for guidance there and
• 30:16the, you know, the reasons
• 30:17for
• 30:18those stipulations.
• 30:20Restricting negotiation are are not
• 30:22really evidence based or the
• 30:24result of lobbying.
• 30:26So I wanna make sure
• 30:27I get enough time for
• 30:29quest, for discussion. So,
• 30:31I'm gonna skip ahead to
• 30:33the second study I was
• 30:34going to present, which is
• 30:35to look at the effects
• 30:36of the law of medication
• 30:37use, which I think is
• 30:39actually a little bit more
• 30:39interesting. So,
• 30:41big picture, this is a
• 30:42difference in differences analysis. We
• 30:44have a
• 30:45control group, and then we
• 30:47have a treatment group. We
• 30:48see what happens to both
• 30:49after an intervention, and then
• 30:51we compare those differences.
• 30:53We used, a database called
• 30:55Symphony Metis, which is a
• 30:56retail all pair pharmacy claims
• 30:58database.
• 31:00And it has data at
• 31:01the drug level. So it's
• 31:02population data. There's no individual
• 31:05level data,
• 31:06but it provides very timely
• 31:08analyses. We have data from
• 31:10actually, like, last week in
• 31:11there.
• 31:12We linked
• 31:13that prescription fill data to
• 31:15pricing data,
• 31:16and we included about three
• 31:18thousand medications.
• 31:19We excluded medications which had
• 31:21no fills in twenty twenty
• 31:22two, the first quarter of
• 31:24our study period, and also
• 31:26vaccine and insulin formulations, which
• 31:28underwent separate reforms during the
• 31:30period.
• 31:31And our key outcome was
• 31:32prescription fills per capita.
• 31:34And we calculated this as
• 31:36a weighted change from baseline.
• 31:38And so we were basically
• 31:40assessing medication use.
• 31:42We hypothesized
• 31:43that the higher cost drugs
• 31:45would increase more. And so
• 31:46we divided medications into different
• 31:47buckets based on their monthly
• 31:49price.
• 31:50And so these are sort
• 31:51of arbitrary cutoffs, but we
• 31:52tried to make cutoffs that
• 31:53were based on out of
• 31:54pocket costs. And so, for
• 31:56example,
• 31:57if a medication costs more
• 31:58than eight hundred and thirty
• 31:59dollars per month, patients have
• 32:00to pay typically more out
• 32:02of pocket for that because
• 32:03the medication is called a
• 32:04specialty medication.
• 32:07And we did a difference
• 32:08of difference.
• 32:09We chose commercial as our
• 32:11control group. The the Inflation
• 32:13Reduction Act did not really
• 32:14affect the commercial population, so
• 32:16we hypothesized that use would
• 32:17be,
• 32:18fairly stable there, and we
• 32:20compared that to use in
• 32:21the Medicare population.
• 32:23We had some seasonal,
• 32:25and drug level fixed effects.
• 32:27We evaluated the parallel trends
• 32:28assumption,
• 32:30just using a simple regression.
• 32:32And then we had a
• 32:33more conservative value for
• 32:35statistical significance given that we
• 32:36did four tests,
• 32:38consistent with our four different,
• 32:40monthly pricing groups.
• 32:43We did various subgroup analyses
• 32:45looking at the most common
• 32:46medications, different diseases.
• 32:48And we knew the law
• 32:49was implemented in January twenty
• 32:50twenty four, and so that
• 32:52is when deductibles reset. And
• 32:53so that's a potential confounder.
• 32:55And so we had a
• 32:56falsification analysis where we we
• 32:59pretended that the law was
• 33:00implemented in twenty twenty three
• 33:01and reran all of our
• 33:02analyses to see if, okay,
• 33:04was it really the the
• 33:05year the law was in
• 33:06lit, or was it the
• 33:07deductibles
• 33:11so some some data here,
• 33:13this is just the table
• 33:14of medications at the drug
• 33:15level and set. And you
• 33:17you'll note that the low
• 33:18cost drugs, less than a
• 33:19hundred dollars per month,
• 33:20are the most common. About
• 33:22half of the cohort is
• 33:23there.
• 33:24And then by the time
• 33:24you get to the very
• 33:25high cost drugs,
• 33:27it's only a hundred and
• 33:27fifty medications, but they're responsible
• 33:29for a large proportion of
• 33:30spending.
• 33:32And to add some color,
• 33:33you see these are the
• 33:34most common medications, atorvastatin,
• 33:37apixaban,
• 33:39Biktarvy. And I like to
• 33:40use the generic name, but
• 33:41the brand name is, like,
• 33:42so catchy when there are
• 33:43three drugs, so I just
• 33:45kinda cave.
• 33:46And then, Ofev, which is
• 33:48a medication for, interstitial,
• 33:50lung disease.
• 33:53We evaluated what's called the
• 33:54parallel trends assumption. We wanna
• 33:56make sure that the two
• 33:57groups are
• 33:59moving in parallel prior to
• 34:00the law of being implemented.
• 34:03And so that is ideal,
• 34:04right, if one is going
• 34:05up and one's going down.
• 34:05And how can you really
• 34:06tell if if it's the
• 34:08if it's the policy?
• 34:09And this is the statistical
• 34:11test we did, which suggests
• 34:13that, controlled for seasonality,
• 34:15the two groups were essentially
• 34:16moving
• 34:17in in parallel.
• 34:19I think it's more intuitive
• 34:20just to look at it
• 34:21visually. And so here's,
• 34:23here's the the preliminary results
• 34:25based on that data.
• 34:27A lot in this slide,
• 34:28I'll walk you through it.
• 34:30The blue is commercial, which
• 34:31is our control group. And
• 34:32so
• 34:33and then the red is
• 34:34our treatment group, Medicare.
• 34:36And so what you have
• 34:37on the x axis is
• 34:39time, and then on the
• 34:40y axis, a change from
• 34:41baseline.
• 34:43So everything starts at zero
• 34:44because that's the baseline. And
• 34:45then over time,
• 34:47you see growth in use
• 34:48or, you know, decreases in
• 34:50use. And this is all
• 34:51before the policy was implemented.
• 34:53So
• 34:54what you would hope for
• 34:55from a research perspective is
• 34:56that the two groups, if
• 34:57they don't overlap, at least
• 34:58they, they are moving in
• 34:59parallel.
• 35:00And you can see qualitatively,
• 35:02the big caveat is the
• 35:04confidence intervals are very wide,
• 35:05but qualitatively, they they match
• 35:07the statistical test that you
• 35:08saw on the prior slide.
• 35:09The two groups are moving
• 35:11similar.
• 35:14Sorry to interrupt.
• 35:15How do you account for
• 35:16the I I know it's
• 35:17marginal increases in numeric numbers
• 35:20of eligible Medicare adults over
• 35:22time as the population is
• 35:24aging slightly just in terms
• 35:25of thinking about
• 35:27change?
• 35:28Yeah. So that's a good
• 35:29question. And,
• 35:31what I should have said
• 35:32is that this was
• 35:34a a prescription fills per
• 35:36capita.
• 35:37Yeah. So we, yeah, we
• 35:38added,
• 35:40that,
• 35:41denominator.
• 35:43And I should've I should've
• 35:44said that on the prior
• 35:45slide.
• 35:46Okay. So that's all. This
• 35:48is all before the law,
• 35:48and they are if I've
• 35:50you know, my argument that
• 35:51they're very similar, and then
• 35:52this is after the law
• 35:53was implemented.
• 35:55And so I'll walk you
• 35:56through and then just quote,
• 35:58like what the picture suggests,
• 35:59and then we'll add some
• 36:00numbers to this. And one
• 36:01is here's the low cost
• 36:03medications. I would argue they
• 36:04are very similar. There's not
• 36:06really a change.
• 36:07Medium, there is some divergence,
• 36:09but it's hard to tell.
• 36:10The common denominators are wide.
• 36:12High, maybe a little bit
• 36:13more. And then very high,
• 36:14you see a clear divergence
• 36:16where the very high cost
• 36:18medication use is soaring in
• 36:19the Medicare population where it's
• 36:21maybe going down or the
• 36:23same in commercial.
• 36:25And so this is adding
• 36:26some numbers to that.
• 36:28What you see here is
• 36:29the difference in difference in
• 36:30the post intervention period. That's
• 36:32the summary metric here. And
• 36:33then the dots just represent
• 36:35that over time. And so
• 36:37overall, the low cost medications,
• 36:40there was a one percent
• 36:41point estimate, but it wasn't
• 36:42significant.
• 36:43Medium,
• 36:45four percent, but not significant.
• 36:47High,
• 36:48six percent, but not significant.
• 36:50And then the very high,
• 36:52consistent with our last graph,
• 36:53you see a a sixteen
• 36:55percent difference in difference,
• 36:57which is is statistically significant.
• 37:02So,
• 37:03that's a good question.
• 37:05You know, what we see
• 37:06is that, you know, use
• 37:06is going up in both
• 37:08groups. And so
• 37:10Exactly.
• 37:11Start covering.
• 37:12Right. So I
• 37:15we we should definitely do
• 37:16a, like, a sensitivity
• 37:17there.
• 37:19You know, it's weighted by
• 37:21it's weighted by prescriptions.
• 37:24And so
• 37:27because,
• 37:28Medicare
• 37:29hadn't covered it for weight
• 37:30loss, and so the very
• 37:32expensive versions that we talked
• 37:34That's true. Yeah.
• 37:35Probably not very much simple.
• 37:38Yeah.
• 37:39Yeah. So I I should
• 37:40go back and do that.
• 37:45Yeah. My my guess is
• 37:46that, you know, despite being
• 37:48very costly, they represent,
• 37:50like, less than one percent,
• 37:51I would guess, of total
• 37:53prescriptions. So my guess is
• 37:54that the numbers would be
• 37:55very similar, but maybe that
• 37:56would be, like, a separate
• 37:57analysis you could just do
• 37:58and and show, like,
• 38:00what percent of the cost
• 38:01is being attributed to that.
• 38:03But thank you for pointing
• 38:04that. Okay. I should write
• 38:05this down to
• 38:08One more question. Just so
• 38:09your summary estimates are all
• 38:11five quarters averaged, but you're
• 38:13clearly seeing
• 38:14it's not like you're not
• 38:15just reporting on your fifth
• 38:16quarter.
• 38:17Right?
• 38:19Yeah. So I think that's
• 38:20a really good point too
• 38:21that, you know, clearly,
• 38:23the trend is not stable
• 38:24over time.
• 38:26And, I mean, so one
• 38:27way you could think about
• 38:28it is you could look
• 38:28at the most recent quarter
• 38:29as the most illustrative. And
• 38:31so,
• 38:32thank you for pointing that
• 38:33out. And so when you
• 38:35look at that, it's like,
• 38:36okay. Now it's the if
• 38:37you just go straight to
• 38:38the very high cost group,
• 38:39it's a thirty five percent
• 38:40increase, which is I think
• 38:42that's, like, pretty significant, and
• 38:43you see the clear dose
• 38:44response too.
• 38:47And this is based on
• 38:47June twenty twenty five data,
• 38:49and so we could update
• 38:50this. And I wonder what
• 38:51where that number would be.
• 38:52Maybe it'd be, like, forty
• 38:53percent, fifty percent. We don't
• 38:55know.
• 38:57And so
• 38:58It would change. Not relative.
• 39:00Right? That's a thirty five
• 39:01percentage point difference between the
• 39:03two.
• 39:04Right. An absolute thirty five
• 39:06percent. Yeah. Which I think
• 39:07is quite large.
• 39:10And so,
• 39:12we'll get into the implications
• 39:13of that, which I think
• 39:14are interesting. So this is,
• 39:16like, our false test. We
• 39:17wanted to see, was it
• 39:18really the law, or was
• 39:19it just twenty twenty three?
• 39:20And I think the takeaway
• 39:21is visually, you see not
• 39:23much is going on in
• 39:24twenty twenty three. And when
• 39:25you repeat all the analyses,
• 39:27there's a lot of noise,
• 39:28but the point estimates essentially
• 39:30are are small or are
• 39:31they, decrease over time.
• 39:35And so I'll bring it
• 39:36back to the patients we
• 39:37saw at the very beginning.
• 39:38Here are the medication, some
• 39:40very common medications those patients
• 39:41were taking as,
• 39:43amlodipine
• 39:44use.
• 39:45We found, you can see
• 39:46here's the individual level difference
• 39:48in difference. We saw a
• 39:49three percent increase.
• 39:51Apixaban,
• 39:52about a five percent.
• 39:54I'll put aside Biktarvy for
• 39:56a moment and return to
• 39:57it. But Ibrutinib use went
• 39:59about eighteen percent. And so,
• 40:01that is the classic, you
• 40:02know, this is exactly what
• 40:03we thought when we had
• 40:04our two example patients, mister
• 40:06a and missus b.
• 40:07At a population level, you
• 40:09see a huge increase in
• 40:10Ibrutinib use,
• 40:12largely because patients like missus
• 40:13Bierp, they're probably hitting the
• 40:14cap, and they're not having
• 40:15to pay more for it.
• 40:18And so what do we
• 40:18make about the negative result
• 40:20for the HIV medications?
• 40:22You know, a little disappointing.
• 40:23We want people to take
• 40:24their Biktarvy. This is a
• 40:25very high value medication.
• 40:28And indeed, it's actually true
• 40:29for all antiretroviral
• 40:30medications. When you look at
• 40:32the class effect, there is
• 40:33no increase. I mean, it's
• 40:34the only class of medications
• 40:35we don't observe an increase.
• 40:38Here's the raw data for
• 40:40for the, HIV medications,
• 40:45and my interpretation of this
• 40:46is that
• 40:48maybe in this you know,
• 40:49we have Ryan White programs.
• 40:50We have Medicaid.
• 40:53This is a major buffer
• 40:54here. Yeah. Yeah. I totally
• 40:56agree. And so I want
• 40:57this is almost like a
• 40:58negative control where it's like,
• 40:59okay. If I had been
• 41:01smart enough to, you know,
• 41:02prespecify this, maybe this would
• 41:04have been a nice negative
• 41:05control. But,
• 41:06you know, post hoc, that's
• 41:08how I interpret that data.
• 41:10I I
• 41:11there are multiple sources of
• 41:13of support for HIV meds
• 41:15that aren't included in the
• 41:16data. So Yeah. Yeah.
• 41:19So,
• 41:20that's a really good point.
• 41:21And so I think big
• 41:22picture, you know, there are
• 41:23a lot of limitations to
• 41:24this population level analysis, and
• 41:26I would love to get
• 41:26your feedback.
• 41:28I'm hoping to follow this
• 41:28up with a patient level
• 41:30analysis for, my k award
• 41:32application.
• 41:34And, you know, there these
• 41:35two populations are very different.
• 41:36Medicare and commercial
• 41:38is is a very big
• 41:40limitation.
• 41:41My hope is that we
• 41:42did drug level analyses. And
• 41:44so you're just looking at
• 41:45people taking napixaban, that that
• 41:46is gonna be more similar
• 41:47than the two populations
• 41:49at large.
• 41:52So in conclusion, the inflation
• 41:53reduction act was associated with,
• 41:56an increase in medication use,
• 41:57particularly for high cost drugs.
• 41:59This could represent improved access.
• 42:01However,
• 42:02not all patients are affected
• 42:04equally, and we know that
• 42:05insurers may respond again.
• 42:07And this could represent both
• 42:09high and low value use.
• 42:10And so in a follow-up
• 42:11analysis,
• 42:12we look at the in
• 42:13the oncology space, and we
• 42:14see that low value care
• 42:15actually increases more.
• 42:17Medications that don't prolong life,
• 42:19that don't improve quality of
• 42:21life increase,
• 42:22about threefold, whereas the other
• 42:24medications,
• 42:25twenty percent ish. So,
• 42:28I think that is potentially
• 42:29concerning. You know, we want
• 42:30patients be taking the best
• 42:31medications, and, ultimately, everyone is
• 42:33paying for that. And so
• 42:34I think that could be
• 42:35an unintended consequence where low
• 42:37value care increases
• 42:38after this cap, and then
• 42:40that those costs are shifted
• 42:42around.
• 42:43So So I think there
• 42:44definitely are there's a need
• 42:45for additional studies, and,
• 42:48I think a clear implication
• 42:49of this data is that
• 42:50there's a need for,
• 42:52expanded price negotiation.
• 42:55Yeah. I have a a
• 42:55question. I mean, so, obviously,
• 42:56you have an interesting background.
• 42:58Were you, like, been on
• 42:58the Hill or sort of
• 42:59around around the Hill before.
• 43:01What's your sense of, like,
• 43:02when the IRA was written
• 43:03that this was like, there
• 43:04was modeling done ahead of
• 43:05time or they anticipated some
• 43:07of these changes might happen?
• 43:08Like, how much is this
• 43:09like, oh, man. We didn't
• 43:09know what was gonna happen,
• 43:10or there's actually kinda what
• 43:11we expected was gonna happen
• 43:13when we implemented,
• 43:15these policy changes.
• 43:16Yeah. It's, so the question
• 43:18was, yeah, how much of
• 43:18this is was planned and,
• 43:20you know, you know, could
• 43:21could we have prevented this,
• 43:23if I summed that up
• 43:23right? And I I think
• 43:24it's a really good point.
• 43:25You know, some of it
• 43:26seems kinda predictable.
• 43:28You know,
• 43:29the so the law lawmakers
• 43:31did plan for the premium
• 43:32spikes so that there are
• 43:33subsidies to address that,
• 43:36probably because premiums are very
• 43:37politically salient. You know, peep
• 43:39patients experience them every month.
• 43:43The, you know, the CBO,
• 43:45the Congressional Budget Office found
• 43:46that the law would save
• 43:47money, actually, in total, driven
• 43:50by the negotiated prices.
• 43:52I don't think they thought
• 43:54that the use would go
• 43:55up this much, and so
• 43:56I think it's important to
• 43:57update that. And,
• 43:58they just put out a
• 43:59call for data, and we're
• 44:00we're gonna meet with the
• 44:01CBO to kinda see, like,
• 44:03compare our models and see
• 44:04what if anything, like, if
• 44:06there were any differences there.
• 44:09And so that is sort
• 44:10of what I I think.
• 44:11I think that people thought
• 44:12the premiums would go up,
• 44:13and they tried to address
• 44:14it, but it's, like, kinda
• 44:15whack a mole situation, unfortunately.
• 44:18Do you have any thoughts
• 44:20about future work in the
• 44:21sense of, like, given these
• 44:22results,
• 44:23thinking about either inappropriate versus
• 44:25appropriate medication prescribing and or
• 44:27are,
• 44:29prescribers, clinicians substituting
• 44:31higher cost meds now that
• 44:32they know their patients can
• 44:34afford them? And so they're
• 44:35less likely
• 44:36to prescribe
• 44:37therapeutic alternatives as in generics
• 44:39or whatever and and just
• 44:40like, oh, now you can
• 44:41have the Cadillac. Everyone gets
• 44:42a Cadillac.
• 44:44Yeah. Yeah. And so I
• 44:45think that's really a good
• 44:47question.
• 44:48I mean, a couple things.
• 44:49Like, one is, like, there's
• 44:51there's now a clear incentive
• 44:53if you are if you
• 44:53have cancer and you're taking
• 44:55an infusion and there is
• 44:56an oral medication.
• 44:58So infusions are not subject
• 44:59to this cap, the part
• 45:01their part b. So
• 45:03I that's, like, something that
• 45:04you could look at if
• 45:05is there a substitution between
• 45:06infusions and to to part
• 45:08d? Like, there's one. And
• 45:10then as you pointed out
• 45:11within class,
• 45:13I didn't put put it
• 45:14on on
• 45:15in my slides, but I
• 45:16I I believe, like, brand
• 45:17name proton pump inhibitor use,
• 45:19I saw, like, goes way
• 45:20up, which is obviously low
• 45:22value when you have generic
• 45:23alternatives that are clinically interchangeable.
• 45:25So there's probably a lot
• 45:26you could do there.
• 45:29You know, other examples,
• 45:30I think in our dataset,
• 45:32you know, antipsychotic use goes
• 45:34up for long acting injectables,
• 45:35which are very expensive. And
• 45:36so I think that's an
• 45:38open question of whether that
• 45:39is a good or bad
• 45:40thing. It could be good
• 45:41if you're switching from the
• 45:42orals to the LAIs, and,
• 45:44like, that seems great. Like,
• 45:45now patients who cannot afford
• 45:47that are are getting that.
• 45:48It could be bad if
• 45:49it just enables people to
• 45:51get antipsychotics
• 45:52that are, of course, associated
• 45:53with mortality. And so I
• 45:55there's, like, a lot you
• 45:55could you could do
• 45:57there,
• 45:58and I'd love to get
• 45:59your feedback on, like, you
• 46:01know, are any of these
• 46:02interesting or what I should
• 46:03do next.
• 46:04So build building off of
• 46:05that of Joe's question and
• 46:07and your answer, so concomitant
• 46:09with this sort of timeline
• 46:10in the last few years,
• 46:12These real time prescription benefits
• 46:14are now being integrated into
• 46:15EHRs and are about to
• 46:16be mandated to be,
• 46:19included. And and I wonder
• 46:21how you build in that,
• 46:23the impact of those on
• 46:24prescribing, especially when it comes
• 46:26to within class alternate
• 46:27alternatives. Because that's what they're
• 46:28meant to do. Right? They're
• 46:29meant to trigger clinicians
• 46:32to choose lower cost options,
• 46:35or lower cost pharmacies. But,
• 46:38anyway, I I think it's
• 46:39important to Yeah. Incorporate that
• 46:41or think about that as
• 46:42we I think that'd be
• 46:43a really interesting study,
• 46:45and
• 46:46I guess you would need
• 46:47EHR data to do it.
• 46:50And,
• 46:53I mean, I think on
• 46:54a population level, it's it's
• 46:55worth asking, like, should some
• 46:57medications not be subject to
• 46:58the cap if they're like
• 47:00like, maybe, like, brand name
• 47:01proton pump inhibitors, we don't
• 47:02cap costs for that. And
• 47:03then,
• 47:04you know, maybe it leads
• 47:05to a week delay when
• 47:07you, you know, maybe you
• 47:08you it's frustrating as a
• 47:09clinician and as a patient.
• 47:10You don't get your proton
• 47:11pump inhibitor for a week
• 47:13because you clicked the you
• 47:14know, you clicked it and,
• 47:15oh, it's not covered and
• 47:16you have to go through
• 47:17that. But ultimately, relatively low
• 47:19stakes for most patients, and
• 47:21it is sort of something
• 47:22that is clearly not indicated.
• 47:24So stuff like that, I
• 47:25think, could be reasonable.
• 47:27But, you know, it's it's
• 47:28tough to
• 47:30change laws and to implement
• 47:31with that amount of nuance.
• 47:33So I don't know. I
• 47:34I think that's a really
• 47:35good question. I think those
• 47:36tools are probably gonna be
• 47:37very helpful.
• 47:39I mean, I think in
• 47:40a related point is to
• 47:41what extent have these changes
• 47:44exacerbated the problem of polypharmacy?
• 47:47Yeah.
• 47:48Low value use anyway.
• 47:51It's sort of a double
• 47:52whammy. Yes. Yeah. Totally. So
• 47:53you could look at potentially
• 47:55inappropriate medication use and
• 47:57yeah.
• 48:00This is really interesting. I
• 48:01was curious, though, in in
• 48:02looking at, like, some of
• 48:03these plans and especially the
• 48:04commercial plans, maybe more so
• 48:05than Medicare or even met
• 48:06MA, like, looking at utilization
• 48:08managements or techniques that they've
• 48:10been employing to actually dissuade
• 48:11you low value care. Because
• 48:12you'd imagine where the brand
• 48:14name, like, program pump inhibitors,
• 48:16they would just say, we're
• 48:17just gonna move it to,
• 48:17like, a different tier, right,
• 48:19and, like, put in fire
• 48:20off and put in these
• 48:21sorts of barriers.
• 48:23So are you seeing in
• 48:23response to this
• 48:25changes from how formulas are
• 48:27being structured
• 48:29as a result
• 48:30of, like, the caps and
• 48:31whatnot?
• 48:32Yeah.
• 48:33We didn't assess that.
• 48:36Other colleagues have looked at,
• 48:38changes in prior auth for
• 48:40medications and have, for whatever
• 48:42reason, they didn't really change.
• 48:44So I don't I was
• 48:44actually surprised by that.
• 48:46The cohort of medications that,
• 48:48this is work by, like,
• 48:49Stacy Dusetzina
• 48:50that she's led,
• 48:52were, like, top selling brand
• 48:54name medications. So,
• 48:56maybe prior auth change for
• 48:58other medications, but,
• 48:59we didn't see that. And
• 49:01then, the formulary tier point
• 49:02you make is is good
• 49:03is a good one.
• 49:05We,
• 49:06let me see if I
• 49:07have this.
• 49:09We looked at
• 49:11different,
• 49:17these are different tiers for
• 49:19different medications and,
• 49:21oh, sorry.
• 49:22This is the next slide.
• 49:23Slide. So this is on
• 49:24the x axis. You have
• 49:25year, and then the y
• 49:26axis, the the tier.
• 49:28And none of these medications
• 49:30really changed their tier before
• 49:31and after the law. So,
• 49:33you can see the lot
• 49:34it basically, the lines are
• 49:35basically just, like, straight lines.
• 49:37So,
• 49:38maybe we should do that
• 49:39for more medications, but at
• 49:40least on first pass, they
• 49:41didn't really change.
• 49:48Okay. I'll take silence as
• 49:50approval.
• 49:54Okay. I'll just,
• 49:55outline my future,
• 49:57plans, which is really to
• 49:58focus on a k award
• 50:00application.
• 50:01I'm hoping to apply for
• 50:02a Beeson award,
• 50:04this October.
• 50:06And then,
• 50:07so,
• 50:08here are the illustrative faculty
• 50:10I would like to work
• 50:11with,
• 50:12and my training and research
• 50:14aims.
• 50:16I really wanna focus on
• 50:17getting better at causal inference,
• 50:18and so training aim one
• 50:20would be,
• 50:21looking at methods and, or
• 50:23sorry, matching,
• 50:25and then two is to
• 50:26deepen my knowledge of geriatric,
• 50:28specific outcomes,
• 50:30and then my,
• 50:32actual research aims. I think
• 50:33I you know, all of
• 50:35this is interesting population level
• 50:36data, but really important to
• 50:38validate and expand upon it
• 50:39in the Part d claims
• 50:40data. And so you can
• 50:42see the research aims are
• 50:43to, you know, assess changes
• 50:45in use, assess changes in
• 50:46outcomes, and then, assess, treatment
• 50:49heterogeneity.
• 50:51And then what I'm hoping
• 50:52to do, in a few
• 50:53years, if the NIA still
• 50:55exists, would be to apply
• 50:56for an r o one.
• 50:59And then here's sort of
• 51:01my timeline. I'm hoping to
• 51:02apply for that in,
• 51:04this October, which I know
• 51:05is aggressive,
• 51:06but starting to work on
• 51:07that now.
• 51:09And then,
• 51:10I just submitted a grant
• 51:11with with colleagues that would
• 51:12be portable to to look
• 51:14at
• 51:15parts of the Inflation Reduction
• 51:16Act in partnership with, Express
• 51:18Scripts, which is a PBM,
• 51:19a pharmacy benefit manager.
• 51:21There is a program that
• 51:22would smooth out of pocket
• 51:23costs over the calendar year
• 51:25that could address some of
• 51:26this.
• 51:27So that's sort of the
• 51:28timeline of what I'm hoping
• 51:29to, to do in the
• 51:31next couple of years.
• 51:33Okay.
• 51:35At the risk of overstuffing
• 51:36this time period with too
• 51:38many more slides, I'll pause
• 51:39there. And and thank you
• 51:41all for for your feedback.
• 51:42It's truly an,
• 51:43obviously a privilege to present,
• 51:45and, you know, your engagement
• 51:47means a lot. So I'll
• 51:48I'll go back and make
• 51:49those changes,
• 51:51to to this to this
• 51:52paper. And so I really
• 51:53appreciate all your time.