University of Southern California Dissertations and Theses

Alzheimer's disease: risk factors, value, and alternative payment models

(USC Thesis Other)

Alzheimer's disease: risk factors, value, and alternative payment models

Pdf

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

ALZHEIMER’S DISEASE:

RISK FACTORS, VALUE, AND ALTERNATIVE PAYMENT MODELS

by

Jeffrey Yu

A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(HEALTH ECONOMICS)

December 2022

ii
Acknowledgements

I would like to express great appreciation for my advisor Dr. Darius Lakdawalla, who
offered tremendous support and guidance during my doctoral studies. I would also like to thank
Dr. Bryan Tysinger, Dr. Erin Trish, Dr. Jakub Hlávka, Dr. Dana Goldman, Dr. Peter Huckfeldt,
and Dr. Ann Harada who have all provided incredibly important research mentorship to me over
the years. Special thanks to my family and friends for their support through this endeavor.

iii
Table of Contents
Acknowledgements ......................................................................................................................... ii
List of Tables .................................................................................................................................. v
List of Figures ................................................................................................................................. v
Abstract ......................................................................................................................................... vii
Introduction ..................................................................................................................................... 1
Disease Overview ....................................................................................................................... 1
Diagnosis and Access ................................................................................................................. 2
Payment Models .......................................................................................................................... 4
Cardiovascular Risk Factors – Heart Valve Disease .................................................................. 7
Research Aims ............................................................................................................................ 9
Chapter 1: Association between Heart Valve Replacement Surgery and the Incidence of
Alzheimer’s Disease and Related Dementias (ADRDs) in Patients with Severe Aortic Stenosis:
Results from a Medicare Claims Analysis .................................................................................... 11
Introduction ............................................................................................................................... 11
Methods..................................................................................................................................... 13
Overview ............................................................................................................................... 13
Data Source and Study Population ....................................................................................... 13
Outcome Measure and Covariates ........................................................................................ 15
Statistical Analysis ................................................................................................................ 15
Results ....................................................................................................................................... 17
Patient Populations................................................................................................................ 17
Clinical Outcomes ................................................................................................................. 18
Discussion ................................................................................................................................. 20
Conclusions ............................................................................................................................... 23
Chapter 2: Impact of Delayed Diagnosis and Access on the Value of Alzheimer’s Disease-
Modifying Therapies: Implications by Socioeconomic Status ..................................................... 36
Introduction ............................................................................................................................... 36
Methods..................................................................................................................................... 38
Overview ............................................................................................................................... 38
Future Elderly Model ............................................................................................................ 38
Cohort Selection .................................................................................................................... 40
Cognitive and Functional Modeling ..................................................................................... 40
Treatment Effectiveness ........................................................................................................ 41
Delayed Diagnosis and Access Scenarios ............................................................................. 42
Results ....................................................................................................................................... 42
Overall Incident MCI Patients .............................................................................................. 42
Comparing by Baseline Educational Attainment .................................................................. 43
iv
Comparing by Age and Baseline Educational Attainment ................................................... 44
Impact on Total Costs ........................................................................................................... 46
Discussion ................................................................................................................................. 46
Conclusions ............................................................................................................................... 49
Chapter 3: Access to Disease-Modifying Alzheimer’s Therapies: Addressing Possible
Challenges Using Innovative Payment Models ............................................................................ 64
Introduction ............................................................................................................................... 64
Methods and Data ..................................................................................................................... 66
Future Elderly Model ............................................................................................................ 67
Treatment Scenarios .............................................................................................................. 68
Future Elderly Model Simulations ........................................................................................ 70
Payment Modeling ................................................................................................................ 70
Payment at Time of Treatment ............................................................................................. 71
Constant Installment Payments ............................................................................................. 71
Pay-for-Performance Installment Payments ......................................................................... 72
Estimating Net Value by Payer ............................................................................................. 72
Results ....................................................................................................................................... 73
Discussion ................................................................................................................................. 76
Limitations ............................................................................................................................ 78
Conclusions ............................................................................................................................... 79
Appendix ................................................................................................................................... 81
Conclusions ................................................................................................................................... 96
Concluding Remarks ................................................................................................................. 99
References ................................................................................................................................... 100

v
List of Tables

Table 1.1. Patient Selection – Identifying Patients with Severe Aortic Stenosis for Study
Cohorts .......................................................................................................................................... 24
Table 1.2. SAVR vs. Medically Managed – Propensity Score Matched Baseline Characteristics
Table ............................................................................................................................................. 25
Table 1.3. SAVR vs. Medically Managed – Propensity Score Matched Cox Proportional Hazard
Model Results for Time-to-ADRD Diagnosis .............................................................................. 27
Table 1.4. TAVR vs. Pre-TAVR Era – Baseline Demographic and Clinical Characteristics ...... 29
Table 1.5. TAVR vs. Pre-TAVR Era – Cox Proportional Hazard Results for Time-to-ADRD
Diagnosis....................................................................................................................................... 31
Appendix Table 1.1A. SAVR vs. Medically Managed – Propensity Score Matched Baseline
Characteristics Table for Sensitivity Scenario and Falsification Outcomes ................................. 32
Appendix Table 1.1B. SAVR vs. Medically Managed – Propensity Score Matched Baseline
Characteristics Table for Sensitivity Scenario and Falsification Outcomes ................................. 33
Appendix Table 1.2. SAVR vs. Medically Managed – Propensity Score Matched Cox
Proportional Hazard Model Results for Sensitivity Scenario and Falsification Outcomes .......... 34
Table 2.1. Overall MCI Sample – Gains from Timely Diagnosis and Access ............................. 50
Table 2.2. MCI Sample by Education – Gains from Timely Diagnosis and Access .................... 51
Appendix Table 2.1. Baseline Characteristics of the MCI Sample .............................................. 56
Appendix Table 2.2. MCI Sample by Age and Education – Gains from Timely Diagnosis and
Access ........................................................................................................................................... 57
Appendix Table 2.3. Gains in Social Value (Aggregate) from Timely Diagnosis and Access .... 60
Appendix Table 2.4. Forgone Gains in Social Value (Aggregate) from Delayed Diagnosis and
Access ........................................................................................................................................... 61
Appendix Table 2.5. QALYs Spent in Cognitive Health States When Using Different Health
Utility Weights .............................................................................................................................. 63
Table 3.1: Demographic Characteristics of the Simulated Population ......................................... 83
Appendix Table 3.1: Discounted QALY Benefit Under and Over 65, by Treatment Effect
Scenario and Cohort ...................................................................................................................... 88
Appendix Table 3.2: Discounted Medical Expenses, by Treatment Effect Scenario and Cohort 89
Appendix Table 3.3: Value-Based Cost of Therapy, by Treatment Effect Scenario and Cohort . 90
Appendix Table 3.4 Sensitivity Analysis (Durability of Treatment Effect) ................................. 91

vi
List of Figures

Figure 1.1. SAVR vs. Medically Managed – Propensity Score Matched Unadjusted Kaplan-
Meier Curve for Time-to-ADRD Diagnosis ................................................................................. 26
Figure 1.2. SAVR vs. Medically Managed – Propensity Score Matched Unadjusted Kaplan-
Meier Curves for Sensitivity Scenario and Falsification Outcomes ............................................. 28
Figure 1.3. TAVR vs. Pre-TAVR Era – Unadjusted Kaplan-Meier Curve for Time-to-ADRD
Diagnosis....................................................................................................................................... 30
Appendix Figure 1.1 Sensitivity Analysis – TAVR vs. Pre-TAVR Era Kaplan-Meier Curves for
Time-to-ADRD Diagnosis using Different Cut-Off Years ........................................................... 35
Figure 2.1. Overall MCI Sample – Forgone Gains from Delayed Diagnosis and Access ............ 52
Figure 2.2. Overall MCI Sample – % Forgone Gains from Delayed Diagnosis and Access ....... 53
Figure 2.3. MCI Sample by Education – Forgone Gains from Delayed Diagnosis and Access .. 54
Figure 2.4. MCI Sample by Education – % Forgone Gains from Delayed Diagnosis and
Access ........................................................................................................................................... 55
Appendix Figure 2.1. MCI Sample by Age and Education – Forgone Gains from Delayed
Diagnosis and Access ................................................................................................................... 58
Appendix Figure 2.2. MCI Sample by Age and Education – % Forgone Gains from Delayed
Diagnosis and Access ................................................................................................................... 59
Appendix Figure 2.3. HUI3 Health Utilities by Years since MCI Onset, by Treatment Effect ... 62
Figure 3.1: Notional Allocation of Treatment Costs by Payment Model ..................................... 84
Figure 3.2: Share of Discounted QALY Gain Under and Over 65 Years of Age, by Treatment
Scenario and Age Cohort .............................................................................................................. 85
Figure 3.3: Medical Expenditures Estimates and Value-Based Cost of Therapy, by Treatment
Scenario and Age Cohort .............................................................................................................. 86
Figure 3.4: Net Costs or Benefits Accrued to Private and Public Payer under Treatment
Scenarios, by Age Cohort and Payment Model ............................................................................ 87

vii
Abstract

Alzheimer’s disease (AD) is a degenerative disease that is characterized by loss of
cognitive function, functional impairment, and neuropsychological symptoms. It is a highly
debilitating disease that results in substantial tolls to patient quality-of-life, and is one of the top
leading causes of death in the US. Aside from aducanumab, which was approved by the U.S.
Food and Drug Administration (FDA) in June 2021, other potentially disease-modifying
therapies (DMTs) for AD are currently under development. These potential DMTs under
development target both early stages such as preclinical and prodromal, as well as mild and
moderate dementia, and have included patients as young as 50 years old in their pivotal trials.
This dissertation encompasses three aims. These include: (1) investigating the
relationship between heart valve disease and the development of Alzheimer’s disease and related
dementias (ADRDs), (2) measuring the impact of delayed mild cognitive impairment (MCI)
diagnosis and access on the clinical and cost benefits of DMTs – overall and by socioeconomic
status, and (3) studying whether traditional upfront payments might limit coverage of DMTs and
how alternative payment models could incentivize greater coverage.
We use Medicare administrative claims data to conduct survival analyses in a cohort of
severe aortic stenosis patients to understand whether aortic valve replacement surgeries – a
surgical treatment for heart valve disease – delay onset of ADRDs. We then utilize the Future
Elderly Model (FEM) microsimulation to understand the burden of MCI and how delays in MCI
diagnosis and DMT treatment may compress or exacerbate health disparities. Lastly, we use the
FEM to understand the net benefits accrued to private and public payers from covering AD
DMTs in patients less than 65 years of age, through upfront, constant installment, and
performance-based installment payment models.
1
Introduction
Disease Overview
Alzheimer’s disease (AD) is the most common type of dementia, and accounts for an
estimated 60% to 80% of cases.
1
It is a degenerative disease characterized by memory loss and
loss of cognitive function, functional impairment, and neuropsychological symptoms.
2
AD is
highly debilitating, and results in substantial tolls to patient quality-of-life.
3
In the absence of a
therapy, the number of persons with clinical AD in the United States (US) is projected to grow
from 6.07 million in 2020 to 13.85 million by 2060.
4
In addition, it is ranked as one of the top
leading causes of death in the US.
5

Although the causes of AD are not completely understood, researchers believe they
include a combination of genetic, environmental, and lifestyle factors. Risk factors include age,
gender, genetics (e.g., APOE4 status), and environmental reasons such as air pollution. Risk
factors such as high cholesterol, high blood pressure, poorly controlled type 2 diabetes mellitus,
obesity, lack of exercise, and smoking – which have been associated with heart disease – may
also increase the risk of AD.
6

AD results in a significant economic and social burden – with annual health care costs
between $28,501 and $42,074, and informal caregiving costs of between $13,188 and $29,229
(both in 2010 dollars)).
7,8
Given the aging of the U.S. population, it is estimated that the
economic burden of both formal and informal care is $159-215 billion annually (depending on
the method used to value the cost of informal care), with an expected more than doubling by
2040 to $379-511 billion per annum.
7
It has also been estimated that delaying AD onset by
1 year could have an economic value of $183,227 per patient, and a delay by 3 years a value of
2
$355,222 per patient (calculated as the difference between the value of additional AD-free years
minus the change in formal and informal costs).
8

Aside from Aduhelm (aducanumab), which was approved by the U.S. Food and Drug
Administration (FDA) in June 2021 using an accelerated approval pathway,
9
other potentially
disease-modifying therapies (DMTs) for AD are currently under development, with the majority
of drugs in trials (83.2%) targeting the underlying biology with the potential of disease
modification.
10
These potential DMTs under development target both early stages such as
preclinical and prodromal, as well as mild and moderate dementia, and have typically included
patients as young as 50 years old in their pivotal trials – although several enroll patients as young
as 18 years of age.
11

Diagnosis and Access
While the availability of effective DMT interventions may provide patients an
opportunity to enjoy longer and healthier life, the benefits of these DMTs may not be fully
realized due to issues revolving missed or delayed diagnosis. In the US, a sizable proportion of
patients that would meet the diagnostic criteria for Alzheimer’s disease and other dementias
(ADRDs) are not diagnosed by a physician.
12-15
Prior research has estimated that the rate of
missed or delayed diagnosis of dementia is considerable, and nears 51%.
13,16
This is problematic,
since delays in timely diagnosis can impede disease management, and in the DMT era, delays
may worsen access to potentially effective therapy. In AD, patients typically experience
cognitive symptoms for an average of three years before receiving a diagnosis.
17,18
As such,
methods to increase timely detection are crucial. Earlier detection at stages such as mild
cognitive impairment (MCI), is also critical, since earlier detection offers a period for better
disease management and intervention before significant neurodegeneration has begun.
18-20

3
Previous estimates of delayed diagnosis have been estimated in the absence of potentially
effective DMT therapies. The Alzheimer’s Association projected that if those on the AD
trajectory were diagnosed during the MCI stage, rather at the dementia stage, that would allow
US national savings of $231 billion by 2050.
20
As this research was conducted in the absence of
DMTs, which have since become available – the consequences of delayed diagnosis may be even
greater. In particular, the consequences for delayed MCI diagnosis are especially concerning,
since clinically any effects to slow disease progression require early detection.
21

Earlier diagnosis may offer meaningful economic savings at the patient level. At the
patient level, early diagnosis of cognitive impairment can reduce patient medical care costs by
almost 30%.
22
Early diagnosis allows for greater opportunities for treatment and management
that can reduce caregiver and nursing home burden.
23-25
Therefore, beyond clinical benefits
alone, early diagnosis offers opportunities to alleviate economic burden. However, given the
potential life expectancy and healthy life expectancy extension that DMTs are likely to offer, in
an area with no disease-course altering alternatives, these drugs are not expected to be cheap. Of
interest is studying how much additional cost savings can be achieved from early diagnosis.
While early diagnosis may not offset the full cost of these DMTs, it may help alleviate the total
costs inclusive of DMT therapy.
Delays in dementia diagnosis have been shown to vary by socioeconomic status (SES).
Lower SES has been associated with greater missed or delayed dementia diagnosis.
26-28
This may
explain why in lower SES groups there is a higher proportion of progressed cases at the time of
diagnosis.
26,28,29
In a study of 1,658 Americans with AD, having fewer years of education was
associated with later detection of disease and greater severity of disease at diagnosis.
26
Delayed
diagnosis in those with less SES may be due to access reasons (medical visit costs or lack of
4
health insurance),
30
constraints in medical resources,
31,32
less awareness/education of dementia,
30

and language barriers.
30,32,33
As such, due to social inequalities in dementia diagnostic
evaluation, lower SES individuals tend to have delays in timely diagnosis. Strategies should be
considered to target patients with lower SES for earlier detection and management. As it stands,
it is unclear whether the arrival of DMTs would exacerbate or compress health disparities
between socioeconomic groups. Since vulnerable populations have been shown to experience
greater delays in AD diagnosis,
26-29
it is worth studying whether delayed or missed diagnosis
may further exacerbate health disparities.
Payment Models
As DMTs for AD head to the US market, in addition to timely diagnosis and access as
discussed above, these treatments will also raise concerns over payment. Under current payment
models, DMTs could result in significant challenges to payers and patients since costs may
accrue sooner than the clinical benefits, especially if clinical benefits prove to be durable.
Although it is uncertain how effective new DMTs will be, even a modest improvement in clinical
outcomes will likely result in significant economic and societal benefit, due to the enormous
disease burden and the lack of treatment options. As such, payers will be considering the budget
impact of covering AD therapies, which may affect their coverage decisions. Given this, there is
an urgency around the economic question of how to pay for the treatment of AD.
The potential DMTs, which are currently undergoing development, target both
preclinical/prodromal and mild or moderate dementia patients, and have typically included patients
as young as 50 years old in pivotal clinical trials, although several trials may enroll patients as
young as 18 years old.
11
While these DMTs create new opportunities for the management of AD
5
they also result in challenges related to access and payment under current upfront payment
methods – especially for private insurers.
Under the status quo upfront payment method, which is fee-for-service payment, services
are paid in full on a per item basis.
34
This payment method does not consider the results of the
treatment (i.e., treatment effect or treatment durability). In certain cases, it can raise concerns
over free-riding. Private payers who pay under status quo payment may face challenges related
to benefit non-internalization, as they pay for the cost of therapies in patients months or years
before they become eligible for Medicare. It is unlikely that any individual payer and
manufacturer can resolve this market failure on their own without some government intervention
given the high likelihood of free-riding. As such, it is clear that in patients younger than 65 years
of age, that there is concern among private payers over free-riding under this status quo payment
method. From the patient perspective, this raises concern that under existing benefit designs and
routine payer therapeutic management techniques, access to life-saving medications may be
unaffordable given anticipated high out-of-pocket costs. Furthermore, payers and providers have
expressed concern over whether they could pay upfront for high-cost therapies that do not yet
have evidence to support their projected long-term benefit, and for which value has yet to be
clearly characterized. As such, while these therapies offer the potential for improved health,
quality of life, productivity and reduced costs, stakeholders are concerned.
Alternative payment models such as annuity payments (i.e., installment payments) and
performance-based payments may mitigate concerns revolving upfront payments. Under annuity
payments, also known as constant installments, a fixed amount of money is paid out to
manufacturers each period (i.e., month or year) over a given period of time or for perpetuity.
35,36

Through an annuity payment arrangement, therapies that are high-value and one-off are paid as if
6
they were recurring therapies, instead of a single payment when the therapy is administered.
Given this structure, annuity payments reduce the yearly budget impact for payers, and can offers
means to handle the uncertainty around the long-term effects of the therapy. For example,
depending on the contracting arrangement, annuity payments could be discontinued if the patient
does not experience the expected treatment durability. This payment arrangement reduces the
risk on the part of the payer and shifts it to the drug developer. Because of how annuities can
reduce long-term risk, high-cost therapies that might otherwise not be covered due to a high
budget impact in the present term, may be more likely to be covered.
Another type of alternative payment model is performance-based payments.
36
Under this
payment method, payments are typically adjusted according to whether pre-specified health
outcomes are achieved. In the payment structure, based on the performance of the therapy there
may be discounts on future payments or rebates from manufacturers to payers. While upfront
payments shift the risk onto the payer, in performance-based payments the risk is shared between
the manufacturer and payer, and manufacturers are rewarded for maintaining the health of the
patient over time. An annuity payment model could also tie in the performance of the therapy,
and serve as a hybrid model (i.e., performance-based installment payments).
The emergence of DMTs will both create new opportunities for the management of AD
and result in challenges related to access and payment. Alternative payment models such as
annuity payments and performance-based annuity payments may offer a means to incentivize
coverage of AD therapies, in a payment environment (i.e., upfront payment) that might otherwise
create barriers to access.

7

Cardiovascular Risk Factors – Heart Valve Disease
While timely diagnosis and access, as well as payment for AD DMTs, are important areas
of study since DMTs may offer a valuable clinical benefit, it remains worthwhile to study the
risk factors for AD, given how the etiology of AD is still unclear. Furthermore, because non-
therapeutic options may have the potential to modify disease progression. There is an expanding
body of literature implicating heart disease as a risk factor for dementia.
37
Positive associations
have been shown between ischemic heart disease and cognitive decline, as well as atherosclerotic
cardiovascular disease and cognitive decline. While this relationship with heart disease has been
shown, it is important however, to focus on more specific types of heart disease in order to
clarify the relative contribution of each entity to cognitive impairment.
With that said, while certain vascular diseases have been established as a risk factor, one
specific type, heart valve disease, has not been well studied.
38
Studies using autopsy data have
reported significant aortic and mitral valve disease in AD subjects compared with a control group
of persons without dementia.
39
Clinical studies have also shown the presence of brain infarcts to
be associated aortic valve calcification, which is supportive of the association between valve
disease and risk of stroke and cognitive decline.
40-42
One small study of Medicare patients found
that compared with controls, patients with AD (n = 18) were more likely to have aortic valve
thickening, aortic valve regurgitation, left ventricular wall motion abnormalities, left ventricular
hypertrophy, and reduced ejection fraction.
43

In addition, one study of 1,401 US veterans found aortic valve calcification (AVC) to be
associated with new diagnosis of cognitive impairment;
44
however, its findings conflicted with a
prior study which did not find aortic valve calcification to be associated with cognitive decline.
45

Given renewed interest around the conversation of AD pathology and etiology, it is very
8
important to understand role of vascular disease in AD development. One way to study the role
of heart valve disease in AD development is to study the relationship between heart valve
surgery and AD.
There are two key aortic heart valve surgeries available. Traditionally, surgical aortic
valve replacement (SAVR), which has been in use since 1960, has been the standard treatment
for severe aortic stenosis (AS).
46
During an open-heart SAVR procedure, a physician makes an
incision in the chest to access the heart, removes the diseased aortic valve, and replaces it with a
new valve. In late 2011, a minimally invasive alternative was approved – transcatheter aortic
valve replacement (TAVR), in which a catheter is inserted into the leg or chest and is guided to
the heart.
47
While there is a lack of research on the effects of SAVR on cognitive outcomes, there
exists a small literature on the cognitive effects of TAVR. A meta-analysis by Khan et al., 2018
looked at 18 studies involving patients with severe AS who underwent TAVR, and found no
overall change in cognitive performance at 3 or 6 months after treatment, or over the long term
(12 to 34 months).
48
In this meta-analysis, 15 of the 18 studies utilized a follow-up of 6 months
or less. Only three studies (n = 190) utilized a follow-up that ranged between 12 to 34 months,
and of these only one had a follow-up longer than 24 months. Overall, the studies in this meta-
analysis emphasized short-term horizons and single-arm observational study designs, and age-
matched controls were rarely included in the reviewed studies. A second meta-analysis by Gu et
al., 2020 reviewed 6 studies on TAVR patients, in which the longest study horizon was 2 years,
and found that a majority of patients did not experience cognitive impairment at any time within
2 years.
49
They did note possible cognitive improvement in patients with impaired baseline
cognition. A third meta-analysis by Oldham et al., 2018 evaluated 12 studies of heart valve
surgeries (aortic, mitral, and mixed; essentially, surgeries other than just TAVR), in which the
9
longest horizon was 6 months.
50
This meta-analysis concluded decline after surgery that was
later restored within 6 months.
The aforementioned meta-analyses were based on observational studies which collected
information on objective measures of cognition before and at different time points after TAVR.
These studies emphasized the perioperative and short-term cognitive outcomes from treatment.
The studies rarely included age-matched controls, and mostly consisted of small sample sizes. A
commentary published by Talbot-Hamon et al., 2017 underscores these concerns, and the need
for much longer-term studies with a wider range of controls.
51
Moreover, none of the prior
literature evaluates the association between cognition and traditional SAVR. As such, there
remains an unmet research need on the relationship between heart valve surgery and dementia
outcomes.
Research Aims
In response to the concerns that have been highlighted above, I ask several research
questions: (1) What is the relationship between heart valve disease and the development of
ADRDs? (2) What is the impact of delayed MCI diagnosis and access on clinical and cost
benefits in the DMT era, and does this impact vary by SES? (3) Would traditional upfront fee-
for-service payment limit coverage of DMTs in younger patients covered by private insurance,
and if so, would alternative payment models resolve this problem?
In Chapter One, I seek to add a novel contribution to the risk factor literature for AD.
This is important because while policy at the moment is focused on therapies coming out of the
pipeline, it is also important to understand the mechanisms at play that give rise to the disease. I
study a cohort of patients with severe AS and perform time-to-event analyses to understand the
relationship between heart valve disease and subsequent development of ADRDs.
10
In Chapter Two, I seek to understand the consequences of delayed diagnosis and access
in the DMT-era. To do so, I model the effects of delayed diagnosis and DMT treatment on
clinical and cost outcomes to quantify how barriers to access diminish the full benefits of DMTs.
I stratify by educational groups to understand how certain groups may be differentially affected,
since delayed diagnosis has been shown to be associated with lower SES.
26-29

In Chapter Three, I study the implications that DMTs may have for private and public
payers since status quo upfront payments cause a misalignment of the accrual of benefits and
costs in time for payers – especially private payers. I explicitly model two alternative payment
approaches which may reduce the risk of barriers to access in younger AD patients – who are
typically covered by private insurance.
As DMTs for AD head to the US market it is important to understand the mechanisms
behind the disease, the impact of delays in diagnosis and access – overall and by SES, and the
potential barriers to drug coverage.

11
Chapter 1: Association between Heart Valve Replacement Surgery and the Incidence of
Alzheimer’s Disease and Related Dementias (ADRDs) in Patients with Severe Aortic
Stenosis: Results from a Medicare Claims Analysis
Introduction
In terms of etiology, scientists do not fully understand the causes of dementia and
Alzheimer’s disease and related disorders (ADRD), but believe it to be a combination of age-
related changes to the brain, along with genetic, environmental, and lifestyle factors.
52
While
certain vascular diseases have been established as risk factors,
53
one specific type, heart valve
disease, has been suspected but understudied. Studies using autopsy data have reported
significant aortic and mitral valve disease in AD subjects, compared to non-demented control
groups.
39
Clinical studies have also shown the presence of brain infarcts to be associated aortic
valve calcification, which is supportive of the association between valve disease and the risk of
stroke and cognitive decline.
40-42
One small study of Medicare patients found that compared with
controls, patients with AD were more likely to have valve thickening, aortic valve regurgitation,
left ventricular wall motion abnormalities, and other symptoms of heart valve disease.
43

Given this suspected relationship between heart valve disease and dementia, it is of great
interest to study whether heart valve surgeries are associated with slower cognitive decline
and/or delayed onset of ADRDs. There are two key heart valve surgeries currently available.
Traditionally, surgical aortic valve replacement (SAVR), which has been in use since 1960, has
been the treatment of choice for severe aortic stenosis (AS).
46
During an open-heart SAVR
procedure, a physician makes an incision in the chest to access the heart, removes the diseased
aortic valve, and replaces it with a new valve. In late 2011, a minimally invasive alternative was
approved – transcatheter aortic valve replacement (TAVR), in which a catheter is inserted into
the leg or chest and is guided to the heart, to perform the valve replacement.
47
While there is a
12
lack of research on the effects of SAVR on cognitive outcomes, there exists a small literature on
the cognitive effects of TAVR. A meta-analysis by Khan et al., 2018 looked at 18 studies
involving patients with severe AS who underwent TAVR, and found no overall change in
cognitive performance at 3 or 6 months after treatment, or over the long term (12 to 34
months).
48
However, these studies emphasized a short-term horizon, and age-matched controls
were rarely included in the reviewed studies. A second meta-analysis by Gu et al., 2020 reviewed
6 studies on TAVR patients, in which the longest study horizon was 2 years, and found that a
majority of patients did not experience cognitive impairment at any time within 2 years.
49
They
did note possible cognitive improvement in patients with impaired baseline cognition. A third
meta-analysis by Oldham et al., 2018 evaluated 12 studies of heart valve surgeries (aortic, mitral,
and mixed; essentially, surgeries other than just TAVR), in which the longest horizon was 6
months.
50
This meta-analysis concluded that decline after surgery that was later restored within 6
months.
The aforementioned meta-analyses were based on observational studies which collected
information on objective measures of cognition before and at different time points after TAVR.
Overall, these studies emphasized the perioperative and short-term cognitive outcomes from
treatment, with a large majority of studies utilizing a horizon less than a year long – only a
handful of studies examined cognitive outcomes beyond one year. Studies rarely included age-
matched controls and consisted of small sample sizes. A commentary published by Talbot-
Hamon et al., 2017 underscored these concerns.
51
Moreover, none of the prior literature has
evaluated the association between traditional SAVR and cognitive outcomes.
Given the uncertainty revolving the long-term impact of heart valve replacement
surgeries on ADRD outcomes, we propose a survival analysis using Medicare data that studies
13
the association between heart valve replacement surgeries and time-to-ADRD diagnosis. This
analysis studies whether SAVR is associated with delayed onset of ADRD diagnosis, and studies
whether the introduction of TAVR was associated with delayed onset of ADRD diagnosis. This
analysis uses a longer time horizon than existing studies, incorporates age-matched controls, and
draws from a much larger dataset. It is our hope that this study helps shed light on the association
between heart valve surgeries and time-to-ADRD diagnosis in patients with severe AS.
Methods
Overview

Our study seeks to understand the association between aortic valve surgeries, such as
SAVR and TAVR, and the development of ADRDs in patients with severe AS. We use Medicare
administrative claims data to algorithmically identify patients with severe AS,
54
and use a
survival analysis framework – involving Cox proportional hazard models – that accounts for
right-censoring in the data to understand the relationship between aortic valve surgeries and
time-to-ADRD diagnosis. The University of Southern California Institutional Review Board
(IRB) deemed this study exempt from review.

Data Source and Study Population

Analyses were conducted on a cohort derived from the 20% Medicare administrative
claims data who were enrolled as fee-for-service beneficiaries. The data spanned from 2002
through 2016 and included inpatient, outpatient, skilled nursing facilities, home health agency,
and carrier claims, as well as the Medicare beneficiary summary file and chronic conditions files.
We identify all patients with severe AS from 2004 through 2013 based on a validated
Medicare claims-based algorithm by Clark et al., 2012.
54
This algorithm defines severe AS based
14
on an inpatient claim for heart failure or balloon aortic valvuloplasty (BAVP), as well as a claim
for AS (either in the inpatient, outpatient, skilled nursing facility, home health agency, or carrier
files) within 2 years. This algorithm has been previously validated through chart review using
echocardiographic and cardiac catheterization data to assess the severity of AS.
54
We use the
earliest of this event as our index, and follow patients forward in time to identify earliest date of
ADRD onset. We also require patients have at least 2 years of continuous Part A and Part B
enrollment prior to index, and exclude those with a diagnosis for ADRD prior to index. Using
this sample, we apply additional criteria, described below, to obtain our cohort for studying
SAVR versus MM, and our cohort for studying the TAVR versus pre-TAVR era (Table 1.1).

SAVR vs. MM

When comparing SAVR versus MM, we analyze patients with an index between 2004
and 2010, which are the calendar years before the availability of TAVR. TAVR was approved in
the fall of 2011. We exclude patients undergoing coronary artery bypass surgery (CABG) or
percutaneous coronary intervention (PCI) in the 6 months post-index. We also exclude patients
that died within 6 months post-index, as this is the period used to identify treatment exposure.
Patients with onset of ADRD prior to index are also excluded. Finally, patients who receive
SAVR within 6 months post-index are defined as SAVR patients, while those who do not are
defined as MM patients.

TAVR Era vs. Pre-TAVR Era

When comparing the TAVR era versus pre-TAVR era, we compare the full universe of
SAVR + MM patients in the 3 calendar years before the availability of TAVR (2008, 2009,
2010), versus the full universe of TAVR + SAVR + MM patients in the three calendar years
15
following (2011, 2012, 2013). In this cohort, we also exclude patients with onset of ADRD prior
to index.

Outcome Measure and Covariates

The date of earliest diagnosis for ADRD was provided by the Medicare chronic
conditions file. Our statistical models accounted for covariates such as age at index, male sex,
race, EuroSCORE (surgery-related mortality risk), as well as Elixhauser Index, and components
of the EuroSCORE [European System for Cardiac Operative Risk Evaluation] that did not
overlap with the components of the Elixhauser Index – in specific, active endocarditis, unstable
angina, extracardiac arteriopathy, cardiac surgery, and myocardial infarction.

Statistical Analysis

SAVR vs. MM

In our SAVR vs. MM analysis, we compare time-to-ADRD diagnosis between severe AS
patients who undergo SAVR and severe AS patients who are MM, using Cox PH models. To
address confounding, we employ propensity score matching as SAVR surgery patients tend to be
healthier and younger than MM patients. We utilize a propensity score-based method to help
match our dataset’s treatment group to control group. To estimate the propensity score, we
predict the likelihood of treatment using the covariates described above via a logistic model. We
match treatment to non-treatment subjects based on 1-to-1 nearest neighbor approach with no
replacement, using a caliper of 0.0001. In addition to this main analysis, we also conduct a
sensitivity scenario where we subset our study sample for lower surgical risk patients using the
EuroSCORE. High surgical risk has been defined as having a EuroSCORE of 20% or higher, and
we explored a subset of patients with a EuroSCORE of 10% or lower.
54
Furthermore, we conduct
16
falsification tests which evaluate outcomes that SAVR is not expected to impact. Falsification
endpoints include time-to-diagnosis of tumor with or without metastasis, osteoporosis, diabetes,
and rheumatoid arthritis or osteoarthritis.

TAVR Era vs. Pre-TAVR Era

In our TAVR era vs. pre-TAVR era analysis, we study the association between TAVR’s
introduction and time-to-ADRD diagnosis in severe AS patients. By doing so, we aim to shed
light on whether TAVR’s arrival coincided with changes in the incidence of ADRD among
severe AS patients. Since TAVR was approved in fall 2011 for high-risk AS patients (i.e.,
systematically sicker) who are not suitable for SAVR, a direct comparison of TAVR and SAVR
may suffer from substantial confounding by unobserved variation in health status. Comparing
SAVR + MM patients in the pre-TAVR era, versus TAVR + SAVR + MM patients in the TAVR
era allows us to avoid this selection bias issue, since we will be looking at the universe of severe
AS patients at both times. Essentially, this study design exploits the introduction of TAVR as a
“natural experiment” identification strategy. This design will help eludicate the incremental
benefit of TAVR. Although this study design will help remove our key confounding problem, we
acknowledge that it will dilute the treatment effect of TAVR. A similar study design was
employed recently to study survival of severe AS patients in the TAVR versus pre-TAVR era.
55

To test the robustness of this analysis, we ran sensitivity analyses where we varied the cut-off
year from 2011 to 2010, 2009, 2008, 2007, and 2006 to investigate whether our treatment effects
were due to the arrival of TAVR. These alternative cut-offs were chosen because they were the
years leading up to the arrival of TAVR. This analysis was also motivated by a potential secular
effect – the incidence of dementia has declined every decade for the past thirty years in the US
56

– and we wished to examine whether our findings were driven by the arrival of TAVR or this
17
secular pattern. We also investigate this secular effect by including calendar year fixed effects in
our main Cox PH models.
Results
Patient Populations

We identified a total of 193,154 patients with severe, symptomatic AS in the Medicare
20% administrative claims data from 2004 through 2013 (Table 1.1). Of these, 185,505 (96.0%)
were continuously enrolled in Part A and B in the 2 years pre-index. In parallel, additional
criteria were applied to obtain our SAVR vs. MM cohort, and our TAVR era vs. pre-TAVR
cohort.

SAVR vs. MM

To obtain our SAVR vs. MM cohort, we required patients have an index date prior to
2011, no claim for CABG or PCI in the 6 months post-index, no death within 6 months post-
index, and no diagnosis for ADRD pre-index. This resulted in a remaining 68,384 (35.4%) of
patients, of which 8,460 (12.4%) underwent SAVR within 6 months post-index and thus were
eligible for propensity score matching. The propensity-score matched cohort comparing SAVR
and MM patients included 8,298 SAVR patients and 8,298 MM patients, with a mean age of 76.6
years, 53.9% male, and a mean Elixhauser Index of 40.0. Baseline characteristics were well-
balanced across the two treatment groups (Table 1.2).

TAVR Era vs. Pre-TAVR Era

To obtain our TAVR era vs. pre-TAVR era cohort, we required patients have an index
date from 2008 through 2013, which is the 3 calendar years before and after the approval of
18
TAVR, and have no diagnosis for ADRDs pre-index. This resulted in a remaining 79,745
(41.3%) of patients. Of these, 38,622 (48.3%) were identified in the TAVR era and 41,123
(51.6%) were identified in the pre-TAVR era, with a mean age of 78.7, 48.2% male, a mean
Elixhauser Index of 42.5, and a mean EuroScore of 21.6. These baseline characteristics were
balanced between the two treatment groups (Table 1.4). In the TAVR era, 12.8% and 6.5% of
severe AS patients underwent SAVR and TAVR, respectively, within 6 months post-index
(Table 1.4). In the pre-TAVR era, 12.6% and 0% underwent SAVR and TAVR, respectively,
within 6 months post-index (Table 1.4).

Clinical Outcomes

SAVR vs. MM

Cox PH models demonstrate that, compared to MM, SAVR is associated with a
significantly lower risk in the development of ADRDs (hazard ratio [HR], 0.818; p < 0.000)
(Figure 1.1 and Table 1.3). We notice that the benefits of SAVR bow outward – in such a way,
that SAVR offers a protective effect in earlier periods up until 9.23 years after index, at which
point the risk then converges with MM. As such, while median time-to-ADRD diagnosis was not
statistically significant (9.58 vs. 8.73 years; p = 0.060), it was near significant, since at the 48
th

percentile of time-to-ADRD diagnosis, time-to-ADRD diagnosis was significantly longer in
SAVR patients (9.23 vs. 8.28 years; p = 0.044). When looking at the 25
th
percentile of time-to-
ADRD diagnosis, time-to-ADRD diagnosis was also significantly longer in SAVR patients (4.88
vs. 3.45 years; p < 0.000). When we further adjust for covariates in the Cox PH model, SAVR
was associated with a lower risk compared to MM (HR, 0.771; p < 0.000) than when without
additional adjustment (Table 1.3).
19
We also examine a sensitivity scenario and a perform series of falsification tests. We
examine a sensitivity scenario where we subset to patients with a EuroSCORE of 10% or less,
which reflects patients with lower surgical risk (Figure 1.2A). In this subset, SAVR continues to
be associated with a significantly lower risk in the development of ADRDs (HR, 0.779; p <
0.000) (Appendix Table 1.2). We examined this sensitivity scenario since patients who undergo
SAVR tend to be healthier with less surgical risk. Falsification tests were also performed to
understand whether outcomes not causally effected by SAVR were affected (Figures 1.2B-
1.2E). In these falsification tests, multivariate Cox PH models show that SAVR is not
significantly associated with greater risk than MM, in the development of tumors with or without
metastasis (HR, 1.051; p = 0.363), osteoporosis (HR, 0.973; p = 0.660), diabetes (HR, 1.046; p =
0.399), or rheumatoid arthritis or osteoporosis (HR, 0.983; p = 0.722) (Appendix Table 1.2).

TAVR Era vs. Pre-TAVR Era
We also compare time to development of ADRDs in severe AS patients in the TAVR era
vs. pre-TAVR era (Figure 1.3). Cox PH models demonstrate that compared to the pre-TAVR
era, the TAVR era was associated with a modest reduced risk in the development of ADRD (HR,
0.945; p < 0.000) (Figure 1.3 & Table 1.5). Due to insufficient follow-up in the TAVR era,
differences in median survival could not be reported. In earlier years since index, we see that
there is a significant delay in time-to-ADRD diagnosis in the TAVR era, compared to the pre-
TAVR era. However, this reduction in hazard diminishes over time. As such, we observe that at
the 31
st
percentile of time-to-ADRD diagnosis, time-to-ADRD diagnosis was significant longer
in the TAVR era than pre-TAVR era (4.04 vs. 3.81 years; p = 0.011), although this is not the
case afterward. When we further adjust for covariates in the Cox PH model, the TAVR era was
20
associated with a lower risk compared to MM (HR, 0.927; p < 0.000) than when without
additional adjustment (Table 1.3).
When we run these same models but change the cut-off year from 2011 to 2010, 2009,
2008, 2007, and 2006, we find no significant difference in hazard rates between the post-period
and pre-period (Appendix Figure 1.1). This sensitivity check indicates a robustness in our
findings. Furthermore, the p-value is in fact minimized when we use the 2011 cut-off, which
suggests that 2011 may be playing an important role in delaying ADRD onset. When we look at
the calendar year fixed effects of our main Cox PH model (Table 1.5), we see that covariates are
not significant and do not consistently decline over time which is in contrast to the declining
incidence of dementia in the US (Table 1.5).
56
However, this may be because the population we
study has cardiovascular disease and the decline in dementia incidence over the past three
decades is in part attributed to preventive efforts and health interventions that may improve
cardiovascular health.
56

Discussion
In this study, we find evidence that aortic valve surgeries may help delay the onset of
ADRDs. While the median time-to-ADRD diagnosis may not be significantly different between
aortic valve replacement surgery and MM, as shown in our work, we find a protective effect in
the earlier years after surgery, which diminishes over time. Our work expands beyond prior
literature by using a much longer follow-up, greater sample size, and matched controls. It is also
the first study to investigate the association between aortic valve replacement surgeries and
ADRD using Medicare administrative claims data.
Prior research has focused on the perioperative or short-term effects of TAVR on
cognition, rather than the long-term effects. A recent meta-analysis of cognitive outcomes after
21
TAVR was conducted by Khan et al., 2018.
48
This meta-analysis of 18 studies consisted of 1,065
participants in total, with 15 of the studies utilizing a follow-up of 6 months or less. Many of
these studies were single-arm observational studies, and rarely utilized age-matched controls.
While studies in meta-analyses such as Khan et al., 2018 studied TAVR, we were unable
to identify studies in the literature that examined the association between SAVR and cognition.
Although TAVR usage has risen rapidly since its approval in late 2011, SAVR is still a common
procedure for severe AS and offers a similar surgical function by replacing the defective aortic
valve. Our TAVR era vs. pre-TAVR era analysis found that increased TAVR uptake, holding
SAVR uptake stable, delayed onset of ADRDs. This finding would be worthy of further
exploration using additional sources of data. And while prior work has compared TAVR and
SAVR in terms of overall survival,
55
future studies may also want to evaluate TAVR and SAVR
in terms of their long-term cognitive trajectory through a variety of other data types).
Given the rapidly aging US population, in which older adults 85+ represent the fastest-
growing segment of the population,
57
and the greater incidence of dementia that comes with
older age,
58
it is important to understand the mechanisms that give rise to these diseases for
better prevention and/or management. Although there is some evidence that patients with AD
experience greater heart valve disease and valve abnormalities, the impact of vascular diseases
such as heart valve disease on cognitive outcomes has largely been understudied.
39,43
Our
research indicates that there may be a link, and highlights the value of better understanding the
role of heart valve disease – as well as vascular disease in general – on cognitive outcomes.
Our work has several limitations. First, the analysis of this study is limited by the
availability of variables in Medicare claims data. While AS can be identified in claims data using
diagnosis and procedural codes, severity of AS is not readily identifiable. As such, we utilized a
22
validated Medicare claims-based algorithm to identify severe AS.
54
However, while this is a
limitation, a benefit of using the Medicare claims dataset is its greater volume of patients and its
more comprehensive longitudinal account of the patient experience. An EMR dataset may not
offer the longitudinal record needed to properly assess time-to-ADRD diagnosis, and a registry
dataset may be limited by sample size. This is the first observational study of its type to be
performed at this scale, and we hope it informs future work using other types of data resources.
Second, as we chose to use the ADRD onset date provided in the Medicare chronic conditions
file, we perform an aggregate analysis of the different types of ADRDs and do not assess by
subtypes, or severity level (i.e., mild cognitive impairment, mild AD, moderate AD). Subtype or
severity may be more identifiable using survey-centered data. Given challenges in identifying
earlier stage ADRDs in administrative claims data, we believe our findings likely represent a
disease state that is more developed, rather than an early disease state. Third, we acknowledge
that we cannot control for all aspects of patient health status. As such, we take actions to account
for confounding. Since patients undergoing SAVR are more likely to be healthier and younger,
we conduct propensity score matching in our SAVR vs. MM analysis. We understand that
propensity score matching cannot account for confounding on unobservables, and are aware that
this will remain a limitation; however, we perform a series of falsification tests to inspect the
internal validity of our framework. Furthermore, we perform a supportive analysis comparing the
TAVR vs. pre-TAVR era to further explore the association between aortic valve replacement
surgery and the reduced risk of ADRDs. This supportive analysis helps address selection biases
since it compares to the universe of severe AS patients in the TAVR and pre-TAVR era, in
which the key difference is the percent using an AVR. While this design provides a solution to
unobserved confounding, it dilutes the effect of TAVR, which is a limitation of this approach.
23
Conclusions
In this study, we find supportive evidence that aortic valve replacement surgeries may
delay the onset of ADRDs. In patients with severe AS, SAVR is associated with a protective
effect and reduces the risk of ADRDs in the earlier years post-index. However, over time this
reduction in risk diminishes and the median time-to-ADRD diagnosis is not significantly
different between the SAVR and MM treatment groups. We also observe a modest delay in the
onset of ADRDs in the TAVR era, compared to the pre-TAVR era. These findings suggest there
may be an association between heart valve disease and ADRDs, and encourage further
investigation using alternative data sources.

24
Table 1.1. Patient Selection – Identifying Patients with Severe Aortic Stenosis for Study
Cohorts
Main Inclusion/Exclusion Criteria N
Inpatient claim with principal diagnosis of HF or a claim for BAVP in 2004 through
2013, AND a claim for AS diagnosis 2 years pre-index. The earliest event for this serves
as the index date
N = 193,154 (100%)
Exclude if patient was not continuously enrolled in Part A & Part B in the 2 years pre-
index
N = 185,505 (96.0%)

Additional Inclusion/Exclusion Criteria for SAVR vs. Medically Managed Cohort N
Exclude if patient’s index date prior to 2011, which is the TAVR era N = 131,784 (68.2%)
Exclude if patient had a claim for CABG in the 6 months post-index N = 128,300 (66.4%)
Exclude if patient had a claim for PCI in the 6 months post-index N = 127,000 (65.8%)
Exclude if patient died in the 6 months post-index N = 84,229 (43.6%)
Exclude if patient had a diagnosis for ADRD pre-index N = 68,384 (35.4%)
Define patients not undergoing SAVR within 6 months post-index as MM subjects N = 59,924 (87.6%)
Define patients undergoing SAVR within 6 months post-index as SAVR subjects N = 8,460 (12.4%)

Additional Inclusion/Exclusion Criteria for TAVR vs. Pre-TAVR Era Cohort N
Exclude if patient’s index date year is not within 2008 through 2013, which is the 3
calendar years before and after the approval year of TAVR
N = 106,702 (55.2%)
Exclude if patient had a claim for ADRD diagnosis pre-index N = 79,745 (41.3%)
Define as pre-TAVR era patient if patient’s index date was before 2008 N = 41,123 (51.6%)
Define as TAVR era subject if patient’s index date was in 2008 or later N = 38,622 (48.4%)
Note: HF = heart failure. BAVP = balloon aortic valvuloplasty. AS = aortic stenosis. CABG = coronary artery
bypass surgery. PCI = percutaneous coronary intervention. SAVR = surgical aortic valve replacement. TAVR =
transcatheter valve replacement. MM = medically managed.

25
Table 1.2. SAVR vs. Medically Managed – Propensity Score Matched Baseline
Characteristics Table

Matched Sample
(N = 16,596)
Baseline Characteristic Unmatched/Matched
SAVR
(N = 8,298)
MM
(N = 8,298)
P-Value
Age at Index (Years) Unmatched 76.3 79.5 0.000

Matched 76.6 76.6 0.802
Male (%) Unmatched 53.6% 43.8% 0.000

Matched 53.0% 54.8% 0.095
Race (White) (%) Unmatched 90.8% 85.0% 0.000
Matched 90.8% 91.1% 0.542
Race (Black) (%) Unmatched 6.1% 10.6% 0.000

Matched 6.1% 6.0% 0.897
Race (Other) (%) Unmatched 3.0% 4.2% 0.000

Matched 3.0% 2.8% 0.426
Elixhauser Index Unmatched 39.6 43.9 0.000

Matched 39.7 40.2 0.207
Active Endocarditis (%) Unmatched 3.7% 1.4% 0.000
Matched 2.4% 1.6% 0.007
Unstable Angina (%) Unmatched 20.6% 21.3% 0.222
Matched 20.6% 20.3% 0.646
Extracardiac Arteriopathy (%) Unmatched 5.2% 7.4% 0.000
Matched 5.3% 5.7% 0.417
Cardiac Surgery (%) Unmatched 47.8% 31.2% 0.000
Matched 46.6% 48.0% 0.174
Myocardial Infarction (%) Unmatched 18.8% 21.4% 0.000
Matched 18.9% 19.3% 0.565
Note: SAVR = surgical aortic valve replacement. MM = medically managed.

26
Figure 1.1. SAVR vs. Medically Managed – Propensity Score Matched Unadjusted Kaplan-
Meier Curve for Time-to-ADRD Diagnosis

Note: SAVR = surgical aortic valve replacement. MM = medically managed. ADRD = Alzheimer’s disease and
related disorders.

27
Table 1.3. SAVR vs. Medically Managed – Propensity Score Matched Cox Proportional
Hazard Model Results for Time-to-ADRD Diagnosis

Model 1
Time-to-ADRD
(N = 16,596)
Model 2
Time-to-ADRD
(N = 16,596)
Predictor Hazard Ratio P-Value Hazard Ratio P-Value
SAVR (vs. MM) 0.818 0.000 0.771 0.000
Age at Index (Years) --- --- 1.067 0.000
Male --- --- 0.879 0.000
Race (Ref = White)

Black --- --- 1.221 0.003
Other --- --- 1.034 0.706
Elixhauser Index --- --- 1.011 0.000
Active Endocarditis --- --- 0.983 0.870
Unstable Angina --- --- 1.007 0.854
Extracardiac Arteriopathy --- --- 1.193 0.051
Cardiac Surgery --- --- 0.942 0.062
Myocardial Infarction --- --- 0.933 0.105
Index Year (Ref = 2004)
2005 --- --- 0.961 0.443
2006 --- --- 0.987 0.801
2007 --- --- 0.943 0.278
2008 --- --- 0.903 0.076
2009 --- --- 0.968 0.569
2010 --- --- 0.990 0.870
Note: SAVR = surgical aortic valve replacement. MM = medically managed. ADRD = Alzheimer’s disease and
related disorders.

28
Figure 1.2. SAVR vs. Medically Managed – Propensity Score Matched Unadjusted Kaplan-
Meier Curves for Sensitivity Scenario and Falsification Outcomes

Note: EuroSCORE = European System for Cardiac Operative Risk Evaluation. SAVR = surgical aortic valve
replacement. MM = medically managed.

0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1
Survival Probability
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Years since Index
95% CI 95% CI
Medically Managed SAVR
Time-to-Alzheimer's-Disease-or-Related-Dementias (<10% EuroSCORE)
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1
Survival Probability
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Years since Index
95% CI 95% CI
Medically Managed SAVR
Time-to-Tumor-or-Metastasis
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1
Survival Probability
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Years since Index
95% CI 95% CI
Medically Managed SAVR
Time-to-Osteoporosis
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1
Survival Probability
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Years since Index
95% CI 95% CI
Medically Managed SAVR
Time-to-Diabetes
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1
Survival Probability
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Years since Index
95% CI 95% CI
Medically Managed SAVR
Time-to-Rheumatoid-Arthritis-or-Osteoarthritis
A B
C D
E
29
Table 1.4. TAVR vs. Pre-TAVR Era – Baseline Demographic and Clinical Characteristics
Characteristic
Pre-TAVR Era
(N = 41,123)
TAVR Era
(N = 38,622)
P-Value
Age at Index (Years) 78.6 78.8 0.085
Male (%) 48.0% 48.4% 0.099
Race

0.663
White (%) 87.0% 86.9% ---
Black (%) 8.9% 8.8% ---
Other (%) 4.2% 4.3% ---
Elixhauser Index 42.3 42.8 0.079
EuroSCORE Score 21.4 21.9 0.064
Active Endocarditis (%) 1.8% 2.0% 0.089
Unstable Angina (%) 15.4% 14.9% 0.051
Extracardiac Arteriopathy (%) 6.7% 6.1% 0.071
Cardiac Surgery (%) 30.8% 30.2% 0.102
Myocardial Infarction (%) 22.1% 21.8% 0.390
AVR Surgery (%) 0.000
SAVR 12.6% 12.8% ---
TAVR 0.0% 6.5% ---
MM 87.4% 80.7% ---
Note: TAVR = transcatheter aortic valve replacement. SAVR = surgical aortic valve replacement. AVR = aortic
valve replacement. MM = medically managed. EuroSCORE = European System for Cardiac Operative Risk
Evaluation.

30
Figure 1.3. TAVR vs. Pre-TAVR Era – Unadjusted Kaplan-Meier Curve for Time-to-
ADRD Diagnosis

Note: TAVR = transcatheter aortic valve replacement. ADRD = Alzheimer’s disease and related disorders.

31
Table 1.5. TAVR vs. Pre-TAVR Era – Cox Proportional Hazard Results for Time-to-
ADRD Diagnosis

Model 1
Time-to-ADRD
(N = 79,745)
Model 2
Time-to-ADRD
(N = 79,745)
Predictor Hazard Ratio P-Value Hazard Ratio P-Value
TAVR Era (vs. Pre-TAVR Era) 0.945 0.000 0.927 0.000
Age at Index (Years) --- --- 1.061 0.000
Male --- --- 0.937 0.000
Race (Ref = White)

Black --- --- 1.276 0.000
Other --- --- 1.036 0.401
Elixhauser Index --- --- 1.010 0.000
Active Endocarditis --- --- 1.027 0.699
Unstable Angina --- --- 0.991 0.706
Extracardiac Arteriopathy --- --- 1.245 0.058
Cardiac Surgery --- --- 0.917 0.070
Myocardial Infarction --- --- 1.009 0.970
Index Year (Ref = 2008)
2009 --- --- 1.011 0.681
2010 --- --- 1.006 0.812
2011 --- --- 1.049 0.153
2012 --- --- 1.005 0.895
2013 --- --- 1.060 0.335
Note: TAVR = transcatheter aortic valve replacement. ADRD = Alzheimer’s disease and related disorders.

32
Appendix Table 1.1A. SAVR vs. Medically Managed – Propensity Score Matched Baseline
Characteristics Table for Sensitivity Scenario and Falsification Outcomes

Model 1
Time-to-ADRD
Among EuroSCORE < 0.10
(N = 8,432)
Model 2
Time-to-Tumor-or-
Metastasis
(N = 14,746)
Model 3
Time-to-Osteoporosis
(N = 15,342)
Baseline
Characteristic
Unmatched/
Matched
SAVR MM
P-
Value
SAVR MM
P-
Value
SAVR MM
P-
Value
Age at Index
(Years)
Unmatched 74.0 76.5 0.000 76.2 80.3 0.000 76.0 79.5 0.000

Matched 74.4 74.5 0.705 76.6 76.6 0.952 76.3 76.4 0.704
Male Unmatched 56.8% 47.1% 0.000 50.7% 39.0% 0.000 60.1% 50.4% 0.000

Matched 56.1% 55.8% 0.824 50.0% 50.9% 0.386 59.7% 61.0% 0.184
Race (White) Unmatched 89.2% 82.5% 0.000 90.5% 84.6% 0.000 90.3% 83.3% 0.000
Matched 89.0% 89.4% 0.656 90.5% 91.3% 0.175 90.1% 90.5% 0.589
Race (Black) Unmatched 7.2% 13.0% 0.000 6.4% 10.8% 0.000 6.8% 12.3% 0.000

Matched 7.4% 7.4% 1.000 6.4% 5.9% 0.368 6.9% 6.3% 0.229
Race (Other) Unmatched 3.5% 4.3% 0.038 3.0% 4.4% 0.000 2.9% 4.2% 0.000

Matched 3.5% 3.1% 0.441 3.0% 2.7% 0.363 2.9% 3.2% 0.455
Elixhauser
Index
Unmatched 34.1 39.7 0.000 38.1 42.8 0.000 40.7 45.7 0.000

Matched 34.4 33.8 0.343 38.2 38.7 0.335 40.8 40.9 0.886
Active
Endocarditis
Unmatched 1.3% 0.4% 0.000 3.8% 1.2% 0.000 4.1% 1.4% 0.000
Matched 0.8% 0.6% 0.316 2.7% 2.3% 0.268 2.9% 2.2% 0.037
Unstable
Angina
Unmatched 9.8% 9.7% 0.873 20.9% 20.8% 0.860 21.1% 21.3% 0.701
Matched 9.9% 9.6% 0.709 21.0% 22.2% 0.180 21.2% 21.2% 0.979
Extracardiac
Arteriopathy
Unmatched 1.7% 2.9% 0.003 5.8% 8.6% 0.000 5.9% 8.6% 0.000
Matched 1.8% 1.6% 0.519 5.9% 6.3% 0.416 6.0% 6.4% 0.452
Cardiac
Surgery
Unmatched 22.4% 12.2% 0.000 48.4% 29.3% 0.000 49.4% 31.8% 0.000
Matched 21.1% 20.3% 0.433 47.4% 48.3% 0.386 48.4% 49.3% 0.394
Myocardial
Infarction
Unmatched 9.1% 9.3% 0.687 19.0% 21.7% 0.000 19.9% 22.4% 0.000
Matched 9.0% 8.5% 0.494 19.2% 19.5% 0.742 20.0% 19.5% 0.611
Note: EuroSCORE = European System for Cardiac Operative Risk Evaluation. SAVR = surgical aortic valve
replacement. MM = medically managed. ADRD = Alzheimer’s disease and related disorders.

33
Appendix Table 1.1B. SAVR vs. Medically Managed – Propensity Score Matched Baseline
Characteristics Table for Sensitivity Scenario and Falsification Outcomes

Model 4
Time-to-Diabetes
(N = 8,404)
Model 5
Time-to-Rheumatoid-Arthritis-or-
Osteoarthritis
(N = 7,958)
Baseline
Characteristic
Unmatched/
Matched
SAVR MM P-Value SAVR MM P-Value
Age at Index (Years) Unmatched 77.3 82.5 0.000 75.0 78.6 0.000

Matched 78.2 78.4 0.451 75.6 76.0 0.111
Male Unmatched 53.5% 41.0% 0.000 58.9% 52.0% 0.000

Matched 52.5% 53.9% 0.346 58.4% 60.5% 0.138
Race (White) Unmatched 92.7% 89.3% 0.000 90.5% 84.1% 0.000
Matched 92.9% 92.7% 0.741 90.3% 91.7% 0.081
Race (Black) Unmatched 5.1% 7.8% 0.000 6.6% 11.3% 0.000

Matched 5.0% 5.5% 0.443 6.7% 5.7% 0.155
Race (Other) Unmatched 2.2% 2.7% 0.138 2.9% 4.4% 0.000

Matched 2.0% 1.8% 0.532 2.9% 2.5% 0.334
Elixhauser Index Unmatched 33.2 37.3 0.000 38.6 43.3 0.000

Matched 33.3 33.3 0.965 38.7 38.5 0.733
Active Endocarditis Unmatched 3.6% 1.2% 0.000 4.3% 1.4% 0.000
Matched 2.4% 1.7% 0.087 2.3% 2.0% 0.494
Unstable Angina Unmatched 17.0% 17.0% 0.960 20.2% 19.5% 0.365
Matched 17.1% 17.1% 1.000 19.8% 20.4% 0.618
Extracardiac
Arteriopathy
Unmatched 5.2% 7.0% 0.000 5.1% 8.1% 0.000
Matched 5.4% 5.2% 0.751 5.2% 5.1% 0.897
Cardiac Surgery Unmatched 44.8% 25.2% 0.000 47.1% 32.2% 0.000
Matched 43.0% 45.1% 0.136 45.5% 46.4% 0.509
Myocardial
Infarction
Unmatched 16.7% 18.3% 0.040 20.8% 22.6% 0.036
Matched 16.8% 17.1% 0.761 20.9% 20.9% 0.972
Note: EuroSCORE = European System for Cardiac Operative Risk Evaluation. SAVR = surgical aortic valve
replacement. MM = medically managed. ADRD = Alzheimer’s disease and related disorders.

34
Appendix Table 1.2. SAVR vs. Medically Managed – Propensity Score Matched Cox
Proportional Hazard Model Results for Sensitivity Scenario and Falsification Outcomes

Model 1
Time-to-ADRD
Among
EuroSCORE
< 0.10
(N = 8,432)
Model 2
Time-to-
Tumor-or-
Metastasis
(N = 14,746)
Model 3
Time-to-
Osteoporosis)
(N = 15,342)
Model 4
Time-to-Diabetes
(N = 8,404)
Model 5
Time-to-
Rheumatoid-
Arthritis-or-
Osteoarthritis
(N = 7,958)
Predictor
Hazard
Ratio
P-
Value
Hazard
Ratio
P-
Value
Hazard
Ratio
P-
Value
Hazard
Ratio
P-
Value
Hazard
Ratio
P-
Value
SAVR
(vs. MM)
0.779 0.000 1.051 0.363 0.973 0.660 1.046 0.399 0.983 0.722
Age at Index
(Years)
1.070 0.000 1.008 0.003 1.026 0.000 0.989 0.000 1.014 0.000
Male 0.860 0.000 1.438 0.000 0.322 0.000 1.137 0.004 0.792 0.000
Race
(Ref = White)

Black 1.229 0.008 1.034 0.721 0.600 0.000 1.352 0.001 1.095 0.257
Other 0.769 0.019 0.715 0.019 1.299 0.050 1.277 0.049 0.817 0.098
Elixhauser
Index
1.012 0.000 1.006 0.000 1.005 0.000 1.005 0.000 1.005 0.000
Active
Endocarditis
0.989 0.954 1.06 0.712 1.093 0.590 0.816 0.207 0.827 0.183
Unstable
Angina
0.980 0.713 1.042 0.477 0.967 0.624 1.215 0.001 1.119 0.026
Extracardiac
Arteriopathy
0.976 0.832 0.880 0.198 0.964 0.749 0.933 0.494 1.109 0.242
Cardiac
Surgery
0.956 0.236 0.995 0.918 0.982 0.732 0.929 0.125 0.990 0.810
Myocardial
Infarction
0.980 0.739 0.915 0.150 0.879 0.072 1.093 0.142 0.890 0.031
Index Year
(Ref = 2004)

2005 1.020 0.748 1.013 0.855 0.927 0.400 1.013 0.866 0.834 0.007
2006 1.029 0.637 0.875 0.084 0.985 0.864 1.063 0.421 0.886 0.071
2007 0.990 0.869 0.840 0.027 1.006 0.944 1.005 0.947 0.897 0.115
2008 0.973 0.685 0.789 0.004 0.982 0.850 1.027 0.744 0.902 0.147
2009 1.026 0.716 0.787 0.006 1.008 0.932 0.990 0.907 0.976 0.760
2010 0.973 0.707 0.720 0.000 0.972 0.771 0.904 0.250 0.985 0.844
Note: EuroSCORE = European System for Cardiac Operative Risk Evaluation. SAVR = surgical aortic valve
replacement. MM = medically managed. ADRD = Alzheimer’s disease and related disorders.

35
Appendix Figure 1.1 Sensitivity Analysis – TAVR vs. Pre-TAVR Era Kaplan-Meier Curves
for Time-to-ADRD Diagnosis using Different Cut-Off Years

0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1
Survival Probability
0 1 2 3 4 5 6 7 8 9 10 11 12 13
Years since Index
95% CI 95% CI
Pre-TAVR Era TAVR Era
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1
Survival Probability
0 1 2 3 4 5 6 7 8 9 10 11 12
Years since Index
95% CI 95% CI
Pre-TAVR Era TAVR Era
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1
Survival Probability
0 1 2 3 4 5 6 7 8 9 10 11
Years since Index
95% CI 95% CI
Pre-TAVR Era TAVR Era
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1
Survival Probability
0 1 2 3 4 5 6 7 8 9 10
Years since Index
95% CI 95% CI
Pre-TAVR Era TAVR Era
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1
Survival Probability
0 1 2 3 4 5 6 7 8 9 10
Years since Index
95% CI 95% CI
Pre-TAVR Era TAVR Era
Using 2007 as a Cut-Off Year
Using 2008 as a Cut-Off Year Using 2009 as a Cut-Off Year
p = 0.074
p = 0.062 p = 0.056
Using 2006 as a Cut-Off Year
p = 0.078
Using 2010 as a Cut-Off Year
p = 0.052
Using 2011 as a Cut-Off Year
p < 0.000
36
Chapter 2: Impact of Delayed Diagnosis and Access on the Value of Alzheimer’s Disease-
Modifying Therapies: Implications by Socioeconomic Status
Introduction
A substantial share of patients who meet the diagnostic criteria for Alzheimer’s disease
and related dementias (ADRDs) remain undiagnosed.
12-15
Prior research estimates that the rate of
missed dementia diagnoses remains considerable and nears 51%.
13,16
This is problematic, since
delays in timely diagnosis can impede disease management, and in the disease-modifying
therapy (DMT) era, delays may also worsen access to potentially effective therapy. In
Alzheimer’s disease (AD), patients experience cognitive symptoms for an average of three years
before receiving a diagnosis.
17,18
Delayed detection at earlier stages such as mild cognitive
impairment (MCI) is even more concerning, as early disease offer a crucial period for
intervention before significant neurodegeneration has begun.
18-20

Along with the recent approval of the first DMT – aducanumab
9
– which targets
prodromal and early stages of AD, a number of other DMT candidates, such as lecanemab
(NCT03887455), donanemab (NCT04437511) and solanezumab (NCT02008357), and
gantenerumab (NCT05256134), are in the clinical pipeline. While controversy persists around
aducanumab,
59,60
the prospect of disease-modification remains, whether from aducanumab itself
or a DMT still in the research pipeline. The healthcare system that cares for patients would do
well to prepare for the possibility of disease-modifying treatment of AD.
Delayed diagnosis has been shown to vary by socioeconomic status (SES). Prior work
has found lower SES to be associated with a greater likelihood of missed or delayed dementia
diagnosis.
26-28
At the time of diagnosis, lower SES groups also tend to have a greater proportion
of cases that have already progressed.
26,28,29
In one study of 1,658 Americans with AD, fewer
years of education were associated with later detection and also greater severity of disease at
37
diagnosis.
26
Lower SES patients may experience greater delays in diagnosis due to less access to
medical care (due to cost of medical visits or lack of health insurance),
30
greater constraints on
medical resources available,
31,32
less awareness/education of dementia,
30
and language
barriers.
30,32,33
As such, due to social inequalities, lower SES individuals experience greater
delays in timely diagnosis and the adverse consequences that result. In addition to greater missed
or delayed diagnosis, patients with lower SES also experience increased incidence of AD or
dementia than patients with greater SES.
61-63

In terms of cost savings, the Alzheimer’s Association projects that if those on the AD
trajectory were diagnosed earlier during the MCI stage, rather than at the dementia stage, cost
savings from better disease management alone would amount to $231 billion in 2050.
20
As this
research was conducted in the absence of DMTs – the cost benefits from early diagnosis may be
even greater in a disease-modifying era. At the individual level, one study of routine screening
for cognitive impairment found that early diagnosis of cognitive impairment can reduce
healthcare costs by almost 30% within one year, by eliminating unnecessary tests and
treatments.
22
Early diagnosis has also been shown to reduce caregiver and nursing home
burden.
23-25
As such, beyond clinical benefits, early diagnosis may offer opportunities to help
alleviate the cost burden on the US healthcare system. Of interest is exploring whether these
opportunities may be concentrated in lower SES individuals.
In our work, we measure the burden of MCI, as well as model the impact of timely and
delayed diagnosis on the clinical benefits of DMT treatments. We stratify by baseline
educational attainment to investigate possible disparities in the burden of MCI, and to identify
whether timely or delayed diagnosis and access would compress or exacerbate health disparities
38
by baseline educational attainment. In performing these analyses, we differentiate by select
treatment scenarios.
Methods
Overview

We sought to measure the burden of MCI – overall and by baseline educational
attainment – and to estimate the impact of delayed diagnosis on the benefits of treating MCI,
with a specific regard to health disparities. We employed the Future Elderly Model (FEM), a
microsimulation which uses nationally representative data from the Health and Retirement
Survey (HRS), to project health and economic outcomes for older Americans aged 51+ with
incident MCI under three delayed diagnosis and treatment access scenarios.

Future Elderly Model

The FEM microsimulation projects health and economic outcomes for Americans aged
51+ using longitudinal data from the HRS, which is a representative, panel survey of Americans
aged 51+. This microsimulation also uses other surveys such as the National Health Interview
Survey (NHIS), the Medical Expenditure Panel Survey (MEPS), and the Medicare Current
Beneficiary Survey (MCBS). The FEM has previously been used to explore related policy
questions, such as the benefits of preventing disease,
64-69
the costs of chronic disease,
70

alternative payment models,
71
as well as disparities in life expectancy and their policy
implications.
72-74

The model comprises of three core modules. The initial cohort module sets the cohort
that is being studied. The transition module estimates transition probabilities using the 1998 to
2018 biennial waves of the HRS survey. This module contains multivariate models for cognition
39
and functional status, chronic conditions, disability status, mortality, and caregiving. These
transition models are estimated as first-order Markovian limited-dependent variable models, and
are a function of patient-level characteristics such as age, education, marital status, chronic
conditions, and smoking history. Probabilities from these models are applied to simulate the
future clinical path of HRS respondents, which we refer to as “simulants.” Essentially, they
determine how patients shift across health states. The summary module takes projections of
simulant-level outcomes and summarizes them into policy outcomes such as discounted life-
years (LYs), quality-adjusted life-years (QALYs), and total costs (medical costs plus caregiver
costs). The estimates for social value are then calculated by multiplying QALY gains by
$150,000 per QALY gained, and then totaled across patients.
75
All outcomes are discounted by
an annual 3%.
Simulation begins with an initial population of Americans aged 51+ with incident MCI
from the HRS dataset. The FEM cycles every 2 years, and simulants that remain alive after each
2-year period continue cycling forward for the remainder of the lifetime model. The
microsimulation is stochastic, and transition probabilities are predicted for simulants based on
their time-invariant and prior time-varying characteristics. Each simulant undergoes 100 Monte
Carlo replications, which are then averaged to calculate our study outcomes. The online
technical Appendix delineates the parametric structure and estimation of the FEM, including all
key inputs and outputs, including how they were measured. A recent study conducted validation
exercises on the FEM, and indicated robust prediction of both quantity and quality of life.
76
The
online technical Appendix provides additional detail on FEM cross-validation, external
validation, and external corroboration.
40

Cohort Selection

The cohort we studied consists of individuals age 51+ from the HRS with incident MCI.
They were selected by identifying individuals with a TICS-27 (Telephone Interview for
Cognitive Status – 27 Points) score reflecting MCI, as well as a TICS-27 score reflecting
cognitively normal status in their prior wave. As a confirmatory check, we also required that
these individuals score a TICS-27 reflecting MCI or dementia in the subsequent wave, a criterion
used in prior research.
77
If an individual died prior to this confirmatory wave, we kept them in
our sample. On the TICS-27 scale, cognitively normal ranged from 12 to 27, MCI from 7-11, and
dementia from 0-6, based on staging from Crimmins et al.
78
Overall, cohort members had
developed incident MCI in years spanning 2012 through 2016, and served as the basis of our
modeling efforts. When calculating model results for this cohort, we averaged across all cohort
members for patient-level outcomes.

Cognitive and Functional Modeling

In this FEM study, cognition is modeled using TICS-27, and is estimated through a first-
order Markov model based on observed transitions in the TICS-27 measure in the HRS. The
estimation sample used is weighted by the likelihood of having a Clinical Dementia Rating
(CDR) score of 0.5 to reflect patients included in DMT clinical trials.
71,79
In the microsimulation,
TICS-27 can impact other transition models, such as mortality, Activities of Daily Living
(ADLs), Instrumental Activities of Daily Living (IADLs), and nursing home entry. Therefore,
when we intervene on TICS-27, these downstream outcomes are impacted as well. We model
functional limitations based on an ordered probit for the number of functional limitations (0, 1, 2,
3 or more) – which include activities such as walking, dressing, bathing, eating, getting in or out
41
of bed, and using the toilet. The predictors in this probit model include previous functional
limitations, demographics, chronic conditions, risk factors, as well as TICS-27.

Treatment Effectiveness

In our analysis, we implement three treatment effects of DMTs. Cognitive benefit is
modeled as a decrease in expected cognitive decline using TICS-27, and functional benefit is
modeled as a decrease in the expected worsening in ADLs. The treatment effects are:
§ Treatment Effect 1: 20% cognitive effect and 40% functional effect – standing in for a
future technology that approximates the benefit observed in the best-case interpretation
of Aduhelm (whose EMERGE trial reported an 18% reduction in decline on the
cognitive MMSE measure and a 40% reduction in decline on the functional ADCS-
ADL-MCI measure) – this joint treatment effect has been previously modeled
71
;
§ Treatment Effect 2: 40% cognitive and 40% functional effect – calibrated at twice the
cognitive treatment effectiveness of Treatment Effect I, to reflect a possibly higher
treatment effect for future treatment innovation that exceeds our best-case;
§ Treatment Effect 3: 100% cognitive and 100% functional effect – to reflect the impact of
a potentially curative treatment and understand the upper bound of potential health
disparities. This treatment effect was applied to measure the burden of MCI – overall and
by baseline educational attainment.
When implementing these treatment effects, we assume these slower rates of decline are
persistent. However, once a simulant reaches a TICS-27 score of 6.4 or lower, which is reflective
of moderate dementia or worse based on a previously published MMSE-to-TICS-27 crosswalk,
80

we discontinue the treatment effect and allow the simulant to follow their natural path of
progression. To implement joint treatment effects, we introduce an additional calibration term to
42
the ordered probit model for functional limitations, and calibrate that parameter to impose our
joint cognitive and functional treatment effect.

Delayed Diagnosis and Access Scenarios

To estimate the effect sizes in our analysis, we first consider an untreated base case in
which we project the remaining LYs, QALYs, and total costs (medical costs plus caregiver costs)
for those in our patient population using the FEM. This base case – known as the no-treatment
scenario – reflects status quo trends in health and is used to understand the impact of
counterfactual delay scenarios that we study.
The counterfactual scenarios involve diagnosing and treating patients 0, 2, and 4 years
after the start of MCI, and are known as the immediate access, 2-year delayed access, and 4-year
delayed access scenarios, respectively.
For each of these different counterfactual scenarios, we compare the remaining lifetime
benefits and costs of diagnosed and treated patients to those of the base case patients – whose
outcomes reflect status quo aging, disability, and death – to measure the effect of varying delays
in diagnosis and treatment access.
Results
Overall Incident MCI Patients

In the FEM microsimulation, incident MCI patients were diagnosed with MCI and treated
using one of the three treatment effect assumptions. Under Treatment Effect I (20% cognitive
and 40% functional), patients receiving immediate access experienced 0.130 QALY gains, under
Treatment Effect II (40% cognitive and 40% functional) experienced 0.217 QALY gains, and
under Treatment Effect III (100% cognitive and 100% functional) experienced 0.975 QALY
43
gains (Table 2.1). Given that QALY gains exceed LY gains, we find that immediate access
presents an opportunity for compression of morbidity under these treatments (Table 2.1).
When a 4-year delay in access was imposed on these incident MCI patients, forgone
QALYs were 0.045 (34%) under Treatment Effect I, 0.070 (32%) under Treatment Effect II, and
0.296 (30%) under Treatment Effect III (Figures 2.1 & 2.2). Respectively, this 4-year delay
corresponded to forgone social value (aggregate) of $6.59 billion, $10.28 billion, and $43.56
billion (Appendix Table 2.4). Since forgone QALYs are greater than forgone LYs, we find that
delaying access causes an expansion of morbidity (Figure 2.1 & 2.2).

Comparing by Baseline Educational Attainment

When comparing QALY gains from the immediate access scenario across educational
groups, we find a gradient – QALY gains tend to be greater among those with lower baseline
educational attainment. Under Treatment Effect II, patients with less than high school education
receiving immediate access experienced 0.229 QALY gains, compared to 0.216 QALY gains in
those with some college or more (Table 2.2). And under Treatment Effect III, patients with less
than high school education receiving immediate access experienced 1.062 QALY gains,
compared to 0.957 QALY gains in those with some college or more (Table 2.2). This finding
under Treatment Effect III measures the lifetime burden of MCI and indicates that there is an
educational disparity in the lifetime burden.
Losses from delayed access also present an education gradient – forgone QALYs are
greater in patients with less baseline educational attainment. If we impose a 4-year delay,
forgone QALYs in patients with less than high school education are 0.087 (38%) under
Treatment Effect II, compared to 0.060 (28%) in those with some college or more (Figure 2.3 &
2.4). And under Treatment Effect III, the same delay results in forgone QALYs of 0.388 (37%) in
44
patients with less than high school education, compared to 0.245 (26%) in those with some
college or more (Figure 3 & 4). We also observe that losses in QALY gains are greater than
losses in LY gains in the less educated; as such, delays will cause a greater expansion of
comorbidity in the less educated (Figure 2.3 & 2.4). As such, coupled with greater burden of
MCI in patients with less baseline educational attainment, delays in diagnosis and access will
further exacerbate the health disparity that lower SES patients experience due to MCI.
The forgone social value (aggregate) due to delay is also greater in the less educated, and
is greatest in patients with high school education (Appendix Table 2.4), in large part because
they represent the largest share of the incident MCI population (Appendix Table 2.1) and
experience greater losses in QALY gain than the most educated. These two characteristics offset
the smaller losses in the high school educated population compared to those with less than high
school education (Figure 2.3).

Comparing by Age and Baseline Educational Attainment

When comparing QALY gains under the immediate access scenario by age and baseline
education, we find that an education gradient persists in those under age 65 – QALY gains tend
to be greater among those with a lower baseline educational attainment. Under Treatment Effect
II, in the under age 65, we find immediate access to increase QALYs by 0.356 in those with less
than high school education, and by 0.319 in those with some college or more (Appendix Table
2.2). And under Treatment Effect III, in the under age 65, QALYs increased by 1.734 in those
with less than high school education, and by 1.447 in those with some college or more
(Appendix Table 2.2). This further exemplifies the greater burden from MCI that patients with
lower SES experience, compared to those with greater SES.
45
When a delay in access is imposed, forgone QALYs are generally greater in those under
age 65 than over age 65, and among both age groups studied the loss is greater in the less
educated (Appendix Figure 2.1). As such, in patients with either younger or older onset of MCI,
those with lower SES experience both greater burden from MCI and greater loss from delayed
diagnosis and access. When we look at percentage terms, the forgone QALY from delay is
greater in those over age 65 (Appendix Figure 2.2) – this is due to their lower QALY gains
from the immediate access scenario (Appendix Table 2.2), coupled with their similar (albeit
moderately smaller) forgone QALYs from delay (Appendix Figure 2.1). For example, in
patients with less than high school education, under Treatment Effect II and III, the percent
forgone QALYs from a 4-year delay is about 47% in the over age 65, but about 27% in the under
age 65 (Appendix Figure 2.2). Interestingly, we observe that in patients with some college or
more, the forgone QALYs from delay are greater among the over age 65 than under age 65
(Appendix Figure 2.1) – which may be due to faster decline after age 65 in the more educated.
In terms of forgone social value (aggregate) due to delay, the loss is greater in the over
age 65 than under age 65, and greater for those with high school education (than other
educational groups), in part because these segments represent larger shares of the incident MCI
population (Appendix Figure 2.1 & Appendix Table 2.1). The difference in forgone social
value (aggregate) between the age groups studied is sizeable – among those with high school
education, the loss due to a 4-year delay is 198% greater in the over age 65 than under age 65,
under Treatment Effect II, while the same metric under Treatment Effect III is 276.67%
(Appendix Table 2.4).

46
Impact on Total Costs

In incident MCI patients, immediate access under Treatment Effect II and III lead to total
cost savings per patient of $12,300 and $47,500, respectively (Table 1). In these patients, a 4-
year delay under Treatment Effect II and III results in forgone total cost savings of $4,200 (34%)
and $16,500 (34%), respectively (Figure 2.1 & 2.2). Total cost savings per patient from
immediate access are greater among the less educated. This indicates that patients with SES
accumulate greater economic burden than their greater SES counterparts. Across education
groups, they vary from $11,100-$14,600 under Treatment Effect II, and $43,000-$58,200 under
Treatment Effect III (Table 2). Forgone total cost savings from delay are greatest in patients with
less than high school education – a 4-year delay under Treatment Effect II and III cuts total cost
savings by $5,600 (39%) and $23,100 (40%), respectively (Figure 2.3 & 2.4).
When inspecting results by age, total cost savings per patient from immediate access are
greatest in patients under 65 with less than high school education – at the upper end, we find
$21,300 in total cost savings under Treatment Effect II, and $86,300 in total cost savings under
Treatment Effect III (Appendix Table 2.2). However, forgone total cost savings from delay are
greatest in patients over 65 with less than high school education –we find that in this subgroup a
4-year delay under Treatment Effect II cuts total cost savings by $6,100 (47%), while under
Treatment Effect III it cuts total cost savings by $23,800 (46%) (Appendix Figure 2.2 & 2.3).
Discussion
The availability of effective DMT interventions may provide patients with cognitive
impairment due to AD an opportunity to enjoy longer and healthier life. Based on our analysis,
patients with lower baseline educational attainment face greater burden of MCI. As such, timely
diagnosis and treatment will be most effective in those with less educational attainment and
47
possibly compress existing disparities. This is due in part to the earlier age of MCI incidence in
the less educated, and their poorer baseline cognitive and functional status (Appendix Table
2.1). While these emerging therapies may be an equalizing force, their ability to reduce health
disparities is sensitive to the risk of delays in access. Compared to the more educated, the less
educated face steeper losses in QALY gains due to delays – as such, the same exact 2-year delay
is more consequential in the less educated. Even more concerning, the less educated are more
likely to experience missed or delayed diagnosis than the more educated.
26
Together, (1) greater
loss from delayed diagnosis in the less educated, and (2) greater likelihood of delayed diagnosis
in the less educated, will further exacerbate the existing educational disparities that we have
identified in our burden of disease estimates.
There are a number of reasons why diagnosis may be delayed. Notable reasons include
brief encounters between patients and physicians, time burden associated with cognitive testing
and counseling, the lack of routine cognitive testing, reluctance of patients and caregivers to
report signs or seek diagnosis due to stigma, and lack of diagnostic resources.
13,81,82
Delays in
diagnosis for patients with less education, in particular, may be driven by barriers to healthcare
access such as cost or insurance coverage,
30
cultural or language barriers,
31,32
and limited
awareness of dementia as a disease (rather than a natural consequence of aging).
30
Even though
Medicare has included cognitive screening in its annual wellness visits, only 23% of fee-for-
service beneficiaries actually receive a cognitive assessment at their wellness visit each year.
Notably, this rate is 30% in Medicare Advantage patients – who tend to be higher SES.
83
As
such, better strategies need to be developed to increase earlier detection, particularly in lower
SES individuals.
48
More timely detection is increasingly relevant in light of the rapidly aging US population,
in which older adults 85+ represent the fastest-growing segment of the population.
57
Coupled
with the fact that dementia incidence increases substantially at older age,
58
the consequences of
delays will pose an even greater cost. We estimate that the aggregate social value benefit from a
hypothetical DMT – under Treatment Effect II – will fall by $5.18 billion (16%) due to a 2-year
delay in diagnosis and access, and $10.28 billion (32%) due to a 4-year delay. These losses from
delay are considerable, since on average it costs $985 million alone to bring a new drug to
market.
84
We find that this forgone social value benefit will be more concentrated in the less
educated, since they experience greater losses from delay and represent a majority of incident
MCI cases. Given this, strategies should improve earlier detection in lower SES patients, to help
minimize losses in social value while compressing health disparities.
In confirmatory checks, we verified that our model shifts simulant time spent in sicker
cognitive health states to healthier cognitive health states, and that the QALYs spent in health
states is sensitive to the health utilities used. (Appendix Table 2.5). Our confirmatory checks
also showed that patients receiving intervention experienced healthier utilities over time in the
years after MCI onset, than their status quo counterparts (Appendix Figure 2.3). Furthermore, in
a review of the literature, we identified a number of modeling studies studying aducanumab’s
treatment benefit,
85-87
and found that our estimates fell within the range from these studies.
Our work has several limitations. First, we choose stylized treatment effects. Since many
DMTs are still in the drug pipeline and long-term data on Aduhelm is still uncertain, it will take
years to understand the true effects of DMTs. Under these stylized treatment effects, we also
assume effects persist over the remainder of the patient’s lifetime unless they discontinue
treatment due to moderate/severe dementia. Second, we assume a uniform treatment effect for
49
patients with equivalent TICS-27 at baseline, which may not be the case if patients with lower
SES have a different biology due to less cognitive reserve. Third, the lack of biomarker testing in
our patient selection restricts us to identifying patients with MCI due to any cause. But only
patients with amyloid pathology are expected to benefit from AD DMT’s in the pipeline, while
non-amyloid patients would likely experience a marginal effect. However, prior research
indicates that a majority of patients with MCI have amyloid deposits and are thus prodromal
AD.
88
Fourth, we assume that treatment promptly occurs following timely or delayed diagnosis.
In real-world practice there are likely to be additional delays related to treatment access due to
therapy costs, prior authorization procedures, noninsurance, and so forth. Since DMT access will
likely favor higher SES individuals, the education gradient we found could be further augmented
by decreased access. Lastly, we value social welfare gain using a single threshold per QALY
gained, which may have its limitations. For example, in the literature it has been argued that a
single QALY threshold may not properly account for differences in responses to treatment
between patients.
89
However, this concern could be remedied by simply scaling estimates
through adjustment of QALY thresholds.
Conclusions
Our analyses indicate that patients with lower SES face greater burden from MCI. DMT
treatment for AD – in concert with timely diagnosis and access – favors the less educated and
may offer compression of health disparities between educational groups. We find that delays in
early diagnosis and access to AD DMTs may result in lost opportunities for millions of patients –
particularly those with lower SES. Mechanisms to improve diagnosis should be prioritized since
delays will pose a risk that may diminish or even erase the equalizing force of new therapies.

50
Table 2.1. Overall MCI Sample – Gains from Timely Diagnosis and Access

Treatment:
Cognitive 20% /
Functional 40%
Treatment:
Cognitive 40% /
Functional 40%
Treatment:
Cognitive 100% /
Functional 100%
Outcome (Discounted)

Incremental LYs 0.069 0.101 0.391
Incremental QALYs 0.130 0.217 0.975
Incremental Total Costs -$5,669 -$12,348 -$47,950
Note: MCI = mild cognitive impairment; LYs = life-years; QALYs = quality-adjusted life-years. Total costs include
both medical costs and caregiver costs.

51
Table 2.2. MCI Sample by Education – Gains from Timely Diagnosis and Access
Treatment:
Cognitive 20% / Functional 40%
Outcome (Discounted) Less than High School High School
Some College or
More
Incremental LYs 0.068 0.068 0.070
Incremental QALYs 0.133 0.128 0.128
Incremental Total Costs -$6,442 -$5,464 -$5,305
Treatment:
Cognitive 40% / Functional 40%
Outcome (Discounted) Less than High School High School
Some College or
More
Incremental LYs 0.107 0.097 0.102
Incremental QALYs 0.229 0.209 0.216
Incremental Total Costs -$14,567 -$11,943 -$11,093
Treatment:
Cognitive 100% / Functional 100%
Outcome (Discounted) Less than High School High School
Some College or
More
Incremental LYs 0.447 0.368 0.375
Incremental QALYs 1.062 0.933 0.957
Incremental Total Costs -$58,228 -$45,303 -$43,009
Note: MCI = mild cognitive impairment; LYs = life-years; QALYs = quality-adjusted life-years. Total costs include
both medical costs and caregiver costs.

52
Figure 2.1. Overall MCI Sample – Forgone Gains from Delayed Diagnosis and Access

Note: MCI = mild cognitive impairment; LYs = life-years; QALYs = quality-adjusted life-years. Total costs include
both medical costs and caregiver costs. Based on discounted outcomes.

0.000
0.020
0.040
0.060
0.080
0.100
0.120
0.140
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
Forgone Gains in LYs
Delay 2 Years Delay 4 Years
0.000
0.050
0.100
0.150
0.200
0.250
0.300
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
Forgone Gains in QALYs
Delay 2 Years Delay 4 Years
$0
$2,000
$4,000
$6,000
$8,000
$10,000
$12,000
$14,000
$16,000
$18,000
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
Forgone Total Cost Savings
Delay 2 Years Delay 4 Years
53
Figure 2.2. Overall MCI Sample – % Forgone Gains from Delayed Diagnosis and Access

Note: MCI = mild cognitive impairment; LYs = life-years; QALYs = quality-adjusted life-years. Total costs include
both medical costs and caregiver costs. Based on discounted outcomes.

0%
5%
10%
15%
20%
25%
30%
35%
40%
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
% Forgone Gains in LYs
Delay 2 Years Delay 4 Years
0%
5%
10%
15%
20%
25%
30%
35%
40%
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
% Forgone Gains in QALYs
Delay 2 Years Delay 4 Years
0%
5%
10%
15%
20%
25%
30%
35%
40%
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
% Forgone Total Cost Savings
Delay 2 Years Delay 4 Years
54
Figure 2.3. MCI Sample by Education – Forgone Gains from Delayed Diagnosis and Access

Note: MCI = mild cognitive impairment; LYs = life-years; QALYs = quality-adjusted life-years. Total costs include
both medical costs and caregiver costs. Based on discounted outcomes.

0.000
0.020
0.040
0.060
0.080
0.100
0.120
0.140
0.160
0.180
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
Forgone Gains in LYs
Less than High School High School Some College or More
0.000
0.050
0.100
0.150
0.200
0.250
0.300
0.350
0.400
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
Forgone Gains in QALYs
Less than High School High School Some College or More
$0
$5,000
$10,000
$15,000
$20,000
$25,000
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
Forgone Total Cost Savings
Less than High School High School Some College or More
55
Figure 2.4. MCI Sample by Education – % Forgone Gains from Delayed Diagnosis and
Access

Note: MCI = mild cognitive impairment; LYs = life-years; QALYs = quality-adjusted life-years. Total costs include
both medical costs and caregiver costs. Based on discounted outcomes.

0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
% Forgone Gains in LYs
Less than High School High School Some College or More
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
% Forgone Gains in QALYs
Less than High School High School Some College or More
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Delay 2
Years
Delay 4
Years
Cognitive 20% /
Functional 40%
Cognitive 40% /
Functional 40%
Cognitive 100% /
Functional 100%
% Forgone Total Cost Savings
Less than High School High School Some College or More
56
Appendix Table 2.1. Baseline Characteristics of the MCI Sample

All
Less than
High School
High School
Some
College or
More
Under 65 Over 65
Less than
High School
High School
Some
College or
More
Less than
High School
High School
Some College
or More
N 1,494 444 571 479 135 152 126 306 419 353
Weighted N 5,887,931 1,567,071 2,285,766 2,035,094 472,614 589,546 522,265 1,077,123 1,696,220 1,512,829
Mean Age
(years)
74.3 72.9 74.5 75.1 59.4 59.6 60.3 78.4 79.7 80.1
Male (%) 46% 40% 44% 52% 38% 55% 57% 42% 40% 51%
Race (%)

White 70% 51% 77% 76% 21% 53% 49% 64% 85% 86%
Black 17% 22% 14% 16% 35% 28% 35% 16% 9% 10%
Hispanic 10% 24% 5% 5% 34% 12% 10% 19% 3% 3%
Educational
Attainment (%)

Less than High
School
30% 100% 0% 0% 100% 0% 0% 100% 0% 0%
High School 38% 0% 100% 0% 0% 100% 0% 0% 100% 0%
Some College
or More
32% 0% 0% 100% 0% 0% 100% 0% 0% 100%
Mean TICS-27
Score
9.70 9.34 9.75 9.91 9.59 9.75 9.77 9.29 9.76 9.96
Number of
ADL's (%)

1 71% 66% 74% 72% 65% 74% 71% 66% 74% 72%
2 11% 10% 12% 10% 10% 4% 8% 10% 15% 10%
3 or more 18% 24% 14% 18% 25% 22% 21% 24% 12% 17%
Number of
IADL's (%)

1 82% 77% 87% 82% 76% 87% 88% 77% 86% 80%
2 11% 15% 7% 12% 16% 5% 8% 14% 8% 13%
3 or more 7% 8% 6% 6% 7% 7% 4% 8% 6% 7%
Mean Remaining
LYs (Discounted)
10.3 10.0 10.3 10.6 14.5 15.2 15.1 8.1 8.6 9.0
Mean Remaining
QALYs
(Discounted)
6.2 5.4 6.3 6.6 8.0 9.7 9.6 4.4 5.1 5.6
Note: MCI = mild cognitive impairment; TICS-27 = Telephone Interview for Cognitive Status – 27 Points; ADL =
Activities of Daily Living; IADL = Instrumental Activities of Daily Living; LYs = life-years; QALYs = quality-
adjusted life-years. Mean remaining LYs and remaining QALYs are base case outcomes, and are not baseline
characteristics.

57
Appendix Table 2.2. MCI Sample by Age and Education – Gains from Timely Diagnosis and
Access
Treatment:
Cognitive 20% / Functional 40%

Under 65 Over 65
Outcome
(Discounted)
Less than
High School
High School
Some College
or More
Less than
High School
High School
Some College
or More
Incremental LYs 0.100 0.096 0.102 0.055 0.059 0.059
Incremental QALYs 0.201 0.192 0.188 0.105 0.105 0.110
Incremental Total
Costs
-$9,579 -$8,502 -$7,450 -5,154 -$4,408 -$4,565
Treatment:
Cognitive 40% / Functional 40%

Under 65 Over 65
Outcome
(Discounted)
Less than
High School
High School
Some College
or More
Less than
High School
High School
Some College
or More
Incremental LYs 0.167 0.144 0.149 0.082 0.081 0.085
Incremental QALYs 0.356 0.320 0.319 0.177 0.171 0.180
Incremental Total
Costs
-$21,182 -$17,018 -$15,412 -$11,854 -$10,179 -$9,602
Treatment:
Cognitive 100% / Functional 100%
Under 65 Over 65
Outcome
(Discounted)
Less than
High School
High School
Some College
or More
Less than
High School
High School
Some College
or More
Incremental LYs 0.731 0.581 0.556 0.329 0.294 0.312
Incremental QALYs 1.734 1.518 1.477 0.782 0.729 0.777
Incremental Total
Costs
-$86,251 -$67,454 -$61,602 -$46,704 -$37,604 -$36,589
Note: MCI = mild cognitive impairment; LYs = life-years; QALYs = quality-adjusted life-years. Total costs include
both medical costs and caregiver costs.

58
Appendix Figure 2.1. MCI Sample by Age and Education – Forgone Gains from Delayed
Diagnosis and Access

Note: MCI = mild cognitive impairment; LYs = life-years; QALYs = quality-adjusted life-years. Total costs include both medical costs and
caregiver costs. Based on discounted outcomes.

0.0000
0.0200
0.0400
0.0600
0.0800
0.1000
0.1200
0.1400
0.1600
0.1800
0.2000
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Under 65 Over 65 Under 65 Over 65 Under 65 Over 65
Forgone Gains in LYs
Delay 2 Years Delay 4 Years
0.000
0.050
0.100
0.150
0.200
0.250
0.300
0.350
0.400
0.450
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Under 65 Over 65 Under 65 Over 65 Under 65 Over 65
Forgone Gains in QALYs
Delay 2 Years Delay 4 Years
$0
$4,000
$8,000
$12,000
$16,000
$20,000
$24,000
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Under 65 Over 65 Under 65 Over 65 Under 65 Over 65
Forgone Total Cost Savings
Delay 2 Years Delay 4 Years
Cognitive 20% / Functional 40%
Cognitive 20% / Functional 40%
Cognitive 20% / Functional 40%
Cognitive 40% / Functional 40% Cognitive 100% / Functional 100%
Cognitive 40% / Functional 40%
Cognitive 100% / Functional 100%
Cognitive 40% / Functional 40% Cognitive 100% / Functional 100%
Cognitive 100% / Functional 100%
59
Appendix Figure 2.2. MCI Sample by Age and Education – % Forgone Gains from Delayed
Diagnosis and Access

Note: MCI = mild cognitive impairment; LYs = life-years; QALYs = quality-adjusted life-years. Total costs include both medical
costs and caregiver costs. Based on discounted outcomes.

0%
10%
20%
30%
40%
50%
60%
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Under 65 Over 65 Under 65 Over 65 Under 65 Over 65
% Forgone Gains in LYs
Delay 2 Years Delay 4 Years
0%
10%
20%
30%
40%
50%
60%
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Under 65 Over 65 Under 65 Over 65 Under 65 Over 65
% Forgone Gains in QALYs
Delay 2 Years Delay 4 Years
0%
10%
20%
30%
40%
50%
60%
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Less than
High School
High School Some
College or
More
Under 65 Over 65 Under 65 Over 65 Under 65 Over 65
% Forgone Total Cost Savings
Delay 2 Years Delay 4 Years
Cognitive 20% / Functional 40%
Cognitive 20% / Functional 40%
Cognitive 20% / Functional 40%
Cognitive 40% / Functional 40%
Cognitive 40% / Functional 40%
Cognitive 40% / Functional 40% Cognitive 100% / Functional 100%
Cognitive 100% / Functional 100%
Cognitive 100% / Functional 100%
Cognitive 100% / Functional 100%
60
Appendix Table 2.3. Gains in Social Value (Aggregate) from Timely Diagnosis and Access
Outcome (Discounted)
Treatment:
Cognitive 20% /
Functional 40%
Treatment:
Cognitive 40% /
Functional 40%
Treatment:
Cognitive 100% /
Functional 100%
Overall $19.13 Billion $31.93 Billion $143.59 Billion
By Education

Less than High School $5.22 Billion $8.99 Billion $41.60 Billion
High School $7.30 Billion $11.97 Billion $53.30 Billion
Some College or More $6.61 Billion $10.97 Billion $48.68 Billion
By Age

Under 65

Less than High School $2.38 Billion $4.20 Billion $20.49 Billion
High School $2.83 Billion $4.72 Billion $2.37 Billion
Some College or More $2.45 Billion $4.17 Billion $19.29 Billion
Over 65

Less than High School $2.82 Billion $4.76 Billion $21.06 Billion
High School $4.47 Billion $7.25 Billion $30.93 Billion
Some College or More $4.16 Billion $6.80 Billion $29.40 Billion

61
Appendix Table 2.4. Forgone Gains in Social Value (Aggregate) from Delayed Diagnosis and
Access
Outcome (Discounted)
Two-Year Delay Four-Year Delay
Treatment:
Cognitive 20% /
Functional 40%
Treatment:
Cognitive 40% /
Functional 40%
Treatment:
Cognitive 100% /
Functional 100%
Treatment:
Cognitive 20%
/ Functional
40%
Treatment:
Cognitive 40% /
Functional 40%
Treatment:
Cognitive 100%
/ Functional
100%
Overall
$3.44 Billion
(18%)
$5.18 Billion
(16%)
$21.50 Billion
(15%)
$6.59 Billion
(34%)
$10.28 Billion
(32%)
$43.56 Billion
(30%)
Education

Less than High School
$1.11 Billion
(21%)
$1.82 Billion
(20%)
$8.06 Billion
(19%)
$2.01 Billion
(38%)
$3.42 Billion
(38%)
$15.21 Billion
(37%)
High School
$1.34 Billion
(18%)
$1.98 Billion
(17%)
$7.56 Billion
(15%)
$2.54 Billion
(35%)
$3.83 Billion
(32%)
$15.87 Billion
(30%)
Some College or More
$0.99 Billion
(15%)
$1.38 Billion
(13%)
$5.68 Billion
(12%)
$2.04 Billion
(31%)
$3.03 Billion
(28%)
$12.48 Billion
(26%)
By Age

Under 65

Less than High School
$0.35 Billion
(15%)
$0.59 Billion
(14%)
$2.68 Billion
(13%)
$0.64 Billion
(27%)
$1.14 Billion
(27%)
$5.38 Billion
(26%)
High School
$0.35 Billion
(12%)
$0.51 Billion
(11%)
$1.83 Billion
(8%)
$0.70 Billion
(25%)
$0.96 Billion
(20%)
$4.03 Billion
(18%)
Some College or More
$0.22 Billion
(9%)
$0.30 Billion
(7%)
$1.44 Billion
(7%)
$0.52 Billion
(21%)
$0.75 Billion
(18%)
$3.31 Billion
(17%)
Over 65

Less than High School
$0.74 Billion
(26%)
$1.21 Billion
(25%)
$5.34 Billion
(25%)
$1.35 Billion
(48%)
$2.26 Billion
(47%)
$9.78 Billion
(46%)
High School
$0.98 Billion
(22%)
$1.46 Billion
(20%)
$5.93 Billion
(19%)
$1.84 Billion
(41%)
$2.87 Billion
(40%)
$11.84 Billion
(38%)
Some College or More
$0.77 Billion
(19%)
$1.09 Billion
(16%)
$4.24 Billion
(14%)
$1.52 Billion
(36%)
$2.28 Billion
(34%)
$9.17 Billion
(31%)

62
Appendix Figure 2.3. HUI3 Health Utilities by Years since MCI Onset, by Treatment Effect

Age 60-65 Years Old at Baseline
Base Case Scenario 100% Cog / 100% Func Scenario

Age 70-75 Years Old at Baseline
Base Case Scenario 100% Cog / 100% Func Scenario

Age 80-85 Years Old at Baseline
Base Case Scenario 100% Cog / 100% Func Scenario

63
Appendix Table 2.5. QALYs Spent in Cognitive Health States When Using Different
Health Utility Weights

FEM Results:
LYs by Health State
FEM Results:
QALYs by Health State
Cognitive
Health
State
Baseline Intervention Baseline

(Using
FEM’s
Predicted
HUI3
Weights)
Intervention

(Using
FEM’s
Predicted
HUI3
Weights)
Baseline

(Using
Neumann
1999
3

HUI2
Weights)
Intervention

(Using
Neumann
1999
3

HUI2
Weights)
Baseline

(Using
Hessman
2016
90

EQ-5D
Proxy-
Rated
Weights)
Intervention

(Using
Hessman
2016
90
EQ-
5D Proxy-
Rated
Weights)
Normal
0.67
(4.4%)
3.77
(23.6%)
0.45 2.48 0.60 3.40 0.60 3.40
MCI
7.96
(52.2%)
10.02
(62.6%)
4.43 5.76 5.81 7.73 5.97 7.51
Mild
Dementia
5.59
(36.7%)
2.13
(13.3%)
2.52 1.07 3.80 1.44 3.41 1.30
Moderate
Dementia
0.92
(6.0%)
0.07
(0.4%)
0.29 0.02 0.50 0.04 0.38 0.03
Severe
Dementia
0.10
(0.7%)
0.0
(0.0%)
0.02 0.00 0.04 0.00 0.02 0.00
Note: QALY = quality-adjusted life-year. FEM = Future Elderly Model; LYs = life-years; MCI = mild cognitive
impairment; HUI2 = Health Utilities Index Mark 2; HUI3 = Health Utilities Index Mark 3; EQ-5D = EuroQol-5
Dimension.

64
Chapter 3: Access to Disease-Modifying Alzheimer’s Therapies: Addressing Possible
Challenges Using Innovative Payment Models
Note: This Chapter has been accepted for publication (currently in press).
Introduction
Disease-modifying therapies (DMTs) for Alzheimer’s disease (AD) are headed to the
U.S. market. In 2021, the first possible DMT for the disease, aducanumab (Aduhelm), received
accelerated approval by the FDA
9
targeting the prodromal and early dementia stages of AD, and
two other drug candidates (Eli Lilly’s donanemab and Eisai’s lecanemab) have initiated rolling
submissions.
91,92
Under current payment models, these could pose significant challenges to
payers and patients since costs may accrue sooner than benefits, especially if benefits prove
durable. In 2020, the number of people estimated to suffer from clinical AD was 6.07 million
which, in the absence of a therapy, is expected to grow to 13.85 million by 2060.
4
Although it is
yet to be determined how effective new DMTs will be at reducing prevalence and incidence of
the disease over a longer term, even a modest improvement in clinical outcomes is likely to
result in significant economic and societal benefits due to the enormous burden of the disease
and the limitations of symptomatic treatments. Moreover, other AD DMTs remain in
development and are nearing approval decisions of their own. These scientific developments
have increased the urgency around the economic question of how to pay for the treatment of AD.
In the US, AD has been show to result in significant economic and social burden (with
annual health care costs of between $35,736
7
and $52,755
8
(in 2021 dollars) and informal
caregiving costs of between $16,536
7
and $36,649
8
(also in 2021 dollars)). Given the aging of
the U.S. population, it was estimated that the economic burden of both formal and informal care
was $159-215 billion in 2010 (depending on the method used to value the cost of informal care),
and expected to more than double by 2040 to $379-511 billion
7,8
estimate that the delay of AD
65
onset by 1 year could have a cumulative economic value of $183,227 per patient and a delay of
AD onset by 3 years a cumulative value of $355,222 per patient (this was calculated as the
difference between the value of additional AD-free years minus the change in formal and
informal costs). The real-world, long-term impact of emerging therapies is not well understood
as trials do not commonly track patient outcomes over a longer time horizon.
Aside from Aduhelm (aducanumab), which was approved by the U.S. Food and Drug
Administration (FDA) in June 2021 using an accelerated approval pathway,
9
other potentially
disease-modifying therapies for AD are currently under development, with the majority of drugs
in trials (82.5%) targeting the underlying biology with the potential of disease modification.
93

These potential DMTs under development target both preclinical and prodromal (also described
as ‘mild cognitive impairment (MCI) due to AD’) and mild or moderate dementia patients. They
have typically targeted patients as young as 50 years old in pivotal trials (with exceptions such as
a Phase 2/3 trial of gantenerumab and solanezumab which may enroll patients as young as 18
years).
11
These DMTs are expected to be administered in monthly infusions until treatment
discontinuation (e.g. when a patient experiences severe adverse events or progresses in their
disease). Different primary endpoints are used in trials, with Phase 3 DMT candidates active in
January 2022 most commonly utilizing outcome measures in the cognitive and functional
domains, such as the Clinical Dementia Rating-Sum of Boxes (CDR-SB) (NCT03486938,
NCT03444870, NCT03443973, NCT03887455, NCT03605667), Alzheimer Disease Assessment
Scale-Cognition (ADAS-Cog) (NCT03790709), ADAS-Cog 11 (NCT03823404, NCT04520412,
NCT04187547, NCT03605667, NCT03446001), ADAS-Cog 12 (NCT04669028), ADAS-Cog 13
(NCT02051608), Alzheimer's Disease Cooperative Study ADL Scale (ADCS-ADL)
(NCT03823404, NCT03790709, NCT02051608, NCT03446001), Alzheimer's Disease
66
Cooperative Study - Clinical Global Impression of Change (ADCS-CGIC) (NCT04520412,
NCT04669028), Preclinical Alzheimer Cognitive Composite 5 (PACC5) (NCT04468659), Free
and Cued Selective Reminding Test (FCSRT) (NCT04098666), Alzheimer's Disease Cooperative
Study-Preclinical Alzheimer Cognitive Composite (ADCS-PACC) (NCT02008357,
NCT02913664), and NIH Toolbox (NIH-TB) Cognition Battery (NCT02913664).
The emergence of such DMTs will create both new opportunities for the management of
AD and MCI due to AD as well as result in challenges related to access and payment. In our
work, we model the differential effect AD DMTs may have by therapeutic effect size and age
and study the implications new therapies may have for private and public payers in the U.S. We
explicitly model two alternative payment approaches which may reduce the risk of barriers to
access in younger patients who are commonly covered by private insurance – constant and
performance-based installment (deferred) payments. Deferred payment has previously been
shown to result in incentives to treat patients sooner, which may have non-trivial clinical
benefits.
94

Although only one therapy has so far been approved, this work may inform potential
future considerations in the design of payment models for disease-modifying AD therapies. It is
not our goal to focus entirely on aducanumab, in light of the uncertainty about its real-world
effectiveness. Rather, we propose hypothetical effectiveness scenarios anchored in part on the
best-case clinical trial results for aducanumab under the view that future therapies may
eventually meet or exceed these.
Methods and Data
We use a microsimulation, the Future Elderly Model, which draws on nationally
representative data from the Health and Retirement Study (HRS) and other sources, to predict
67
clinical and economic outcomes of interest in a population of adults between 51 and 85 years
under 4 different treatment effectiveness scenarios. We develop a payment model to estimate the
net benefit/loss which may be attributable to AD DMTs by age, payer type and payment model,
as discussed below. Our results provide a health care sector perspective of treatment benefits and
costs, given the focus on direct costs and benefits that private and public payers will accrue.
95,96

Future Elderly Model

The Future Elderly Model (FEM) was first applied to estimate the benefits of medical
innovation in the U.S. elderly in 2005.
97
The FEM is a simulation model that projects person-
level health and economic outcomes using longitudinal data from the HRS and other surveys,
including the National Health Interview Survey (NHIS), the Medical Expenditure Panel Survey
(MEPS), and the Medicare Current Beneficiary Survey (MCBS). The FEM consists of
multivariate models of cognitive and functional status, disease conditions, disability, mortality,
nursing home entrance, caregiving, and others, based on patient-level characteristics, such as a
history of smoking, age, marital status, and education. Earlier work has shown that functional
decline follows cognitive decline, and is largely explained by it
98,99
– the FEM itself follows this
path dependency and allows for functional decline to be causally linked to cognitive decline in
all adults.
The FEM predicts economic and clinical outcomes of interest at the patient level in 2-
year waves and takes into account a range of individual-level characteristics. Cognitive function
in the FEM is estimated as a first-order Markov model based on observed transitions in the
Telephone Interview for Cognitive Status (TICS-27) score in the HRS sample. We weight our
estimation sample by the probability of having a Clinical Dementia Rating score of 0.5 to mimic
those included in clinical trials for AD DMTs. TICS-27 also impacts other transition models
68
within the microsimulation: mortality, activities of daily living, instrumental activities of daily
living, and nursing home entry. Consequently, an intervention on TICS-27 will impact these
downstream outcomes. Functional limitations are modeled with an ordered probit for the number
of functional limitations (0, 1, 2, 3 or more) and include walking, dressing, bathing, eating,
getting in/out of bed, and using the toilet. This model is a function of previous functional
limitations, demographics, chronic diseases, risk factors, and TICS-27.
Applying the transition model for TICS-27 yields an updated cognitive score based on
the projected characteristics of the individual. In the intervention scenarios this cognitive score is
adjusted to be consistent with the hypothetical intervention. For example, consider an individual
who declined from a TICS-27 score of 10 to a TICS-27 score of 7 between two periods, a change
of 3 points. In the 20 percent cognitive treatment scenario this 3-point decline would be reduced
to a 2.4-point decline, yielding a score of 7.6.
Direct interventions on functional limitations are handled slightly differently. Functional
limitations are modeled with an ordered probit model. We introduce an additional calibration
term to the ordered probit model and then calibrate that parameter to yield a particular outcome,
such as a two-year change in the proportion of patients with any limitations.
More information about the FEM is provided in the FEM’s Technical Documentation.
100

Treatment Scenarios

We model four treatment effect scenarios of AD DMTs in a population of Americans
over the age of 50 years with mild cognitive impairment in FEM. These scenarios allow us to
model the expected benefit of hypothetical DMTs at different levels of effectiveness in the
cognitive and functional domains. We model cognitive benefit as a reduction in expected
69
cognitive decline using TICS-27 and functional effect as a reduction in the expected
development of difficulties with Activities of Daily Living (ADLs). The scenarios are:
• Scenario 1: 20% cognitive effect and 40% functional effect – approximately referencing
the benefit observed in the EMERGE trial of Aduhelm (which reported a 22% reduction
in decline on the cognitive and functional composite measure CDR-SB, an 18% reduction
in decline on the cognitive measure MMSE, a 27% reduction in decline on the cognitive
ADAS-Cog 13 measure, and a 40% reduction in the decline on functional ADCS-ADL-
MCI measure
1
);
• Scenario 2: 10% cognitive and 20% functional effect – calibrated at half of Scenario 1
(as a more pessimistic treatment effectiveness outcome);
• Scenario 3: 40% cognitive effect and 20% functional effect (as a sensitivity analysis
testing the inverse relationship between cognitive and functional benefit relative to
Scenario 1);
• Scenario 4: 20% cognitive effect and 10% functional effect (half of Scenario 3).
Our treatment scenarios assume the slower rates of decline are permanent. In sensitivity
analyses, we investigate the consequences of shorter-lived effects. For each scenario, we
compare the lifetime benefits and costs to matched controls – identical patients who are
simulated to age naturally in the absence of an AD DMT – and report benefits and costs before
and after the age of 65 years individually. Additional information about TICS-27 and ADLs, as
well as the specification of the treatment effect under each scenario, is provided in the Appendix.

1
MMSE, ADAS-Cog 13 and ADCS-ADL-MCI were secondary endpoints in the EMERGE trial.
70

Future Elderly Model Simulations

Simulated individuals enter the model based on their cognitive status. HRS respondents
with a TICS-27 score of seven or higher, based on,
78
who are likely to have a CDR score of 0.5,
are selected. Individuals are excluded if they have a recent heart condition, recent stroke, or
congestive heart failure. HRS does not allow for patient selection based on amyloid status, a
hallmark of Alzheimer’s pathology which is among common inclusion criteria in AD clinical
trials. In a large U.S. population study, 55.3% of MCI patients and 70.1% of dementia patients
had positive amyloid Positron Emission Tomography (PET) results.
88
Select demographics
characteristics of patients in our simulation are shown in Table 3.1.

Payment Modeling

We develop an economic model that translates gains due to treatment under each scenario
and for each five-year age cohort. The model assigns a value of $150,000 to each quality-
adjusted life year (QALY) gained,
75
which is considered as the social welfare gain due to
treatment. FEM estimates the effect of each of the modeled interventions on direct health care
costs (which include costs associated with nursing home stays). The model estimates the net
value accrued due to therapy under and over 65 years of age as the net balance of the social
welfare gain due to treatment, the difference in health care costs, and the cost of therapy. The
cost of therapy is assigned differently in each of the payment models studied, namely status quo
(payment upon treatment), constant installment payment and pay-for-performance installments
payments (see Figure 3.1 for a notional description of each of these payment models). We do not
account for possible termination of effect and apply both installment approaches for the patients’
remaining lifetimes (this could be adjusted in the real world without the loss of generalizability).
71
We study both constant and pay-for-performance installment options given the unique age
profile of patients with cognitive impairment caused by Alzheimer’s disease, which is
hypothesized to push benefits further into the future while costs are incurred early on (this
assumption is specific to the most advanced disease-modifying drug candidates in the pipeline,
many of which are biologic therapies).
All future costs and benefits are discounted at 3% annually. We notionally assume all
patients under 65 years are covered by a commercial plan (private payer) and patients over 65
years are covered by Medicare (public payer), although real-world insurance coverage under and
over 65 years is more complex. Our modeling also abstracts from the challenges of a fragmented
payment landscape in the U.S.

Payment at Time of Treatment

First, we model payment that occurs at the time of treatment (status quo). The cost of
treatment is calculated as a fraction (20%) of social welfare gain (quantified based on the
estimated QALY gain valued at $150,000 per QALY gained), drawing on historical estimates in
multiple disease areas in which biomedical innovators appropriated less than 25% of the surplus
produced by their innovations.
101,102

Constant Installment Payments

Second, we model constant installment payments. We first estimate the mean, discounted
life expectancy for each treated age group and distribute the cost of therapy, calculated as 20% of
the value of the QALY gain, which is based on a $150,000 per QALY gained threshold. We
allocate the cost of treatment to equal annual payments based on discounted years before (when
72
applicable) and over 65 years of age. This payment model does not assume any down payment
aside from the first annual installment payment.

Pay-for-Performance Installment Payments

Third, we model performance-based installment payments. We first estimate the
discounted QALY gain under (when applicable) and over 65 years of age, and distribute the cost
of therapy, calculated as 20% of the value of the QALY gain at a $150,000 per QALY gained
threshold, based on the share of the QALY gain before and after the age of 65 years. This
payment model is designed to match payment with the accrual of benefits over time.

Estimating Net Value by Payer

For each of these payment models, the net value accrued to the private payer (under 65
years of age) or public payer (over 65 years of age) is calculated as a difference between the
discounted cost of therapy (accrued as described above based on the payment model), net
difference in medical expenses (increase or decrease), and the welfare gain in QALY terms,
valued at $150,000 per QALY gained. For instance, a therapy which results in a 0.10 QALY gain
(of which a 0.05 QALY gain is observed before the age of 65 years) would have a value-based
price of $3,000 under status quo based on a $150,000 per QALY gained threshold and the
assumption that 20% share of societal gain is recovered by the manufacturer. In our model of
status quo payment, it would be covered in full by private insurance for patients under 65 years
of age. If such therapy reduces medical expenditures before 65 years of age by $5,000 and
increases medical expenses over 65 by $5,000, the net benefit accrued to private payers would be
calculated as follows: 0.05*$150,000 (social value) + $5,000 (medical expenditure savings) –
$3,000 (cost of therapy) = $9,500. For public payers, the net benefit accrued under status quo
73
(upfront payment) would be calculated as follows: 0.05*$150,000 (social value) – $5,000
(additional medical expenditures) = $2,500. In Appendix Table 3.4, we show how the net benefit
results would change if the durability of treatment was limited to 2, 4, 6, 8, or 10 years,
respectively.
Results
Using the Future Elderly Model, patients treated under each of the four treatment
scenarios realize discounted QALY gains that follow different patterns based on the treatment
effect assumed, as shown in Figure 3.2. Patients are expected to realize the highest discounted
QALY gains in our baseline Scenario 1 (20% cognitive and 40% functional effect), with the
highest clinical benefit in patients aged 56-60 years (0.55 QALYs gained), and the smallest
clinical benefit in patients aged 81-85 years (0.30 QALYs gained), both assuming a persistent
treatment effect. Reducing the treatment effect in half (Scenario 2) results in a reduction of
QALY gains by between 47-50% depending on the age group, with the highest clinical benefit in
patients aged 56-60 years (0.29 QALYs gained) and the lowest clinical benefit in the oldest
patients aged 81-85 years (0.15 QALYs gained).
In Scenarios 3 and 4, clinical benefit follows a similar trajectory; however, QALY gains
relative to Scenarios 1 and 2, respectively, are somewhat smaller. For example, the highest
reported QALY gain in Scenario 3 is in patients 56-60 years who expected to report 0.49 QALYs
gained, or approximately 10% less than the same cohort in Scenario 1. The difference between
Scenarios 2 and 4 is even a little bit bigger, with patients aged 56-60 years expected to report
0.25 QALYs gained in Scenario 4, or about 14% less than the same cohort in Scenario 2. The
relatively smaller clinical benefit in Scenarios 3 and 4 relative to Scenarios 1 and 2 indicates that
74
halting functional decline has a greater effect on improving quality of life than halting cognitive
decline.
As also shown in Figure 3.2, patients younger than 65 years at time of treatment stand to
accrue most of their QALY gains after the age of 65 years in all treatment scenarios. This
confirms the hypothesis that under the status quo payment model, which pays for therapy at time
of treatment, an incentive misalignment between payment and benefit accrual may occur. The
underlying data for Figure 3.2 are shown in the Appendix Table 3.1.
In Figure 3.3, we show the implications of each treatment scenario on medical
expenditure savings and show the value-based cost of therapy (based on the average QALY gain
and the value-based cost of treatment estimation described above). In all four scenarios, patients
are expected to achieve medical expenditure savings (even when they live longer), reaching up to
$7,020 for patients aged 66-70 years in Scenario 1 and $8,017 for patients of the same age in
Scenario 3. These results indicate that cognitive benefit may have a relatively larger impact on
medical expenditure savings than equivalent functional benefit. The relatively smallest medical
expenditure savings are observed for patients treated at younger age (however, this may be
affected by the timing of cost savings and the effects of discounting in our analysis). Across all
treatment scenarios, patients aged 61-75 typically achieve the highest savings of medical
expenditures, suggesting those patients may be most attractive to payers from a reimbursement
perspective. In Figure 3.3, we also show the value-based estimate of the discounted, lifetime
cost of therapy for each age group and treatment scenario. Given the QALY gains reported
above, we observe relatively large differences between age groups. In Scenario 1, the value-
based cost of therapy would be expected to range from $8,871 for patients aged 81-85 years to
$16,465 for patients aged 56-60 years. In other scenarios, the value-based cost of therapy would
75
also be highest for patients 56-60 years old ($3,673 in Scenario 2, $14,827 in Scenario 3, and
$7,406 in Scenario 4). Patients treated at older age would be expected to have a lower value-
based cost of therapy (accounting only for QALY gains and medical expenditure savings).
Should the manufacturers be able to appropriate more than 20% of the surplus produced by their
therapy, the value-based cost of therapy may increase. The underlying data for Figure 3.3 are
shown in the Appendix Tables 3.2 and 3.3.
In Figure 3.4, we compare the net benefit or loss accrued to private and public payers by
scenario, patient age cohort, and payment model. In some age cohorts, we find a net negative
impact on private payers from providing access to an AD DMT, largely driven by the proximity
to Medicare eligibility age, which prevents sufficient benefit accrual to outweigh the costs of
treatment.
Under all four scenarios in the status quo, payers are expected to accrue net negative
benefits (i.e. losses) for patients in the 61-65 cohort, estimated to be -$9,708 in Scenario 1, -$280
in Scenario 2, -$10,348 in Scenario 3, and -$5,049 in Scenario 4.
Net losses are completely avoided under both constant and pay-for-performance payment
models. Pay for performance installments would increase the financial incentive to private
payers relative to constant installments. For example, payers’ net benefit for patients aged 61-65
years old in Scenario 1 would be $4,704 under pay-for-performance installments compared with
$4,442 under constant installments.
As we show in sensitivity analyses (Appendix Table 3.4), there is limited need for
innovative payment contracts if treatment durability is limited to 2 years only (with patients
covered by private payers not expected to accrue a significant net loss due to therapy). However,
76
for treatment durability of 4 years and longer, temporal misalignment of costs and benefit
remains, and innovation in payment can help address it.
Discussion
Emerging DMTs in Alzheimer’s disease have the potential to improve people’s lives.
Yet, ensuring access under existing payment arrangements may prove to be a challenge,
particularly for patients under 65 years covered by commercial plans. Even if priced at a fraction
of overall welfare gain, such therapies may result in net losses accrued to private payers,
particularly for patients close to the Medicare eligibility age. With potential DMTs requiring
payment at time of treatment (e.g. for amyloid-based biologic therapies used for a defined period
of time), private payers may opt to delay the treatment of such patients. Alternatively, they may
increase premiums or put such drugs on high co-pay formulary tiers, which could result in
patients shouldering a larger share of the cost of treatment. Either of these approaches may result
in suboptimal outcomes as fewer patients receive timely access to these emerging treatments.
We find that both constant and pay-for-performance installment payment models offer
possible solutions. By spreading the cost of therapy over the reminder of a patient’s lifetime,
private payers would no longer face the disincentive to provide care as their share of the
treatment cost would no longer make the net benefit of treatment negative. Pay-for-performance
installments may serve as the first-best solutions from an actuarial perspective; however, their
implementation costs may outweigh their benefits. Constant installments result in net positive
benefits to private payers for all age cohorts studied. Treatment durability of 2 years or less
would obviate the need for innovative payment contracts, but durability of 4 years or longer
creates misaligned incentives.
77
Should installment-based payment models be considered for implementation,
assumptions about the long-term benefits of therapy may need to be made based on population
models and simulations, such as the FEM, because clinical trials typically provide only a few
years of relevant outcome data. The uncertainty about benefit accrual poses challenges to
payment design.
103
In addition, public and private payers would need to cooperate and agree on a
mechanism to transfer patients’ payment obligations between one another when they switch
plans or become eligible for Medicare.
104

Other approaches not explicitly modeled here may be considered. For instance, a move to
a single payer model or a transfer from public payers to commercial plans in order to incentivize
the treatment of younger patients may offset net negative benefits they would otherwise incur.
Such transfer may take different forms. Alternatively, patients not yet eligible for Medicare may
receive a subsidy to pay for their AD treatment. Yet another potential solution is an Alzheimer’s
risk pool which could be set up to provide access to both early screening, diagnosis and
treatment of Alzheimer’s disease. Screening and early diagnosis may include cognitive and
biomarker-based tests, requiring the reimbursement of primary care providers to administer
them.
105
An Alzheimer’s-specific risk pool would provide both public and private payers with
the option to make an actuarially fair contribution for patients over a certain age. In return, such
an insurance facility could cover the cost of screening, diagnosis, and treatment for anyone with
cognitive impairment due to Alzheimer’s disease, which would create the incentives to avoid
delay in care and possibly also negotiate optimal drug prices with manufacturers (based on
lifetime treatment value and relevant elements of value). A risk pool may also more optimally
react to new insights about long-term efficacy of emerging AD DMTs, and have more flexible
contracts with manufacturers (while reducing bureaucratic burden to many different payers
78
across the nation). Although theoretically this solution may have many benefits, regulatory
changes and buy-in from both public and private payers would be required for its success (which
could later extend to other disease areas with unique treatment paradigms, such as gene and cell
therapies). Our findings do indicate, however, that efficiency gains can be made when the
temporal misalignment of costs and benefits is addressed. We leave the study of other payment
models to future research.
While the implementation of innovation in payment may face many challenges, the
economic impact of AD DMTs may result in substantive demand for new solutions.
106

Moreover, AD DMTs may encourage the conduct of health technology assessment (HTA) that
would require the consensus of many stakeholders across the nation.
107

Limitations

Our work has several limitations. First, we use a stylized representation of the treatment
effect under four distinct scenarios that is assumed to persist over the remainder of a patient’s
lifetime. Given typical clinical trial duration, it will take many years before conclusions about
lifetime benefits in QALY terms will be more precisely understood, resulting in uncertainty
about the specification of treatment effect in our model and other similar research. The results in
the main text should be viewed as upper bounds on the expected gains from innovative payment,
because we assume that treatment permanently slows the rate of cognitive and/or functional
decline. In sensitivity analyses, we show that gains are smaller when treatment effects are
shorter-lived, but that gains persist so long as treatment effects last four years or more.
Additionally, while we select patients who most closely resemble eligible populations in AD
DMT clinical trials, the lack of biomarker testing in our selection restricts our model to patients
with cognitive impairment due to any cause. It is possible that patients with amyloid pathology
79
will observe greater or smaller benefits than expected, for instance if the treatment effect relative
to a control group increases over time, or if patients with prodromal and mild AD experience a
faster decline than our data indicate. Third, we quantify social welfare gain using a single
threshold per QALY gained, which has its limitations. Some have argued, for example, that it
does not account properly for differences between different patients’ response to a medication, or
its externalities, and that it may result in discrimination against persons with disabilities and
chronic illnesses.
108
Also, our estimates of treatment benefit (which translate to QALY gains) are
based on TICS-27, CDR, and ADL outcome measures. Limited tools are available to ‘crosswalk’
between endpoints within cognitive and functional domains,
109
making direct comparisons to
clinical trial results less reliable. Additionally, our work estimates direct health care costs and
benefits, such as the reduction in nursing home care expenditures, but we do not explicitly model
the lower demand on informal care and possible improvements in the quality of life of
caregivers, insurance value of a therapy, and its other elements of value.
110
Finally, mean
benefits from AD DMTs may differ depending on the composition of the patient pool of
individual insurers given that risk factors and comorbidities are not uniformly distributed across
regions and demographic groups. Our analysis estimates mean benefits under select treatment
scenarios based on nationally representative data.
Conclusions
We find that AD DMTs may provide patients with different levels of clinical benefit
depending on age and treatment effects in the cognitive and functional domains. Our modeling
indicates that payment at time of treatment for these therapies may result in challenges to access
given the net negative benefit private payers are expected to incur in some patients under 65
years. We study two potential payment model solutions: constant and performance-based
80
installments. Both avoid the net loss accrual to private payers. Depending on the long-term
benefits and costs of AD treatments, such contracts may be negotiated, requiring the
coordination of private and public payers in the coverage of AD DMTs. Other solutions, such as
an AD-specific risk pool may be considered. Ultimately, the success of emerging DMTs for
Alzheimer’s disease will depend on patients’ timely access to them. Our work suggests that new,
innovative approaches may be necessary, particularly for patients covered by private payers.

81
Appendix
Overview on Treatment Scenario Specification
The Future Elderly Model is a representative microsimulation of U.S. adults of 51 years
of age and older. We conduct disease-specific modeling to approximate possible treatment effect
sizes by domain. Given the impact of treatment on both cognitive and functional domains, we
impose cognitive and functional improvements simultaneously, calibrating the functional
improvements to match the intended treatment effect conditional on the imposed cognition
effect.
Transition Models

The FEM uses the Telephone Interview for Cognitive Status (TICS), a scale of 0-27
where 27 is the highest score, from the HRS to assess a person’s cognitive status. Using
Crimmins et al. (2011), we restrict our sample to patients with a TICS score of over 7, in
addition to a predicted CDR score of 0.5, which we use to select patients with cognitive
impairment, no dementia (CIND). The ordinary least squares model for TICS-27 is a function of:
TICS-27 in the previous survey wave, race/ethnicity, education, gender, gender*race,
gender*education, age splines with knots at 65 and 75, and previous chronic disease indicators
for cancer, diabetes, heart attack, heart disease, hypertension, chronic lung disease, and stroke.
The FEM uses Activities of Daily Living (ADLs) as a measure of functional status, aside
from other measures (such as Instrumental Activities of Daily Living (IADLs) and nursing home
status). Functional limitations are modeled with an ordered probit for the number of functional
limitations (0, 1, 2, 3 or more). This model is a function of previous functional limitations,
demographics, chronic diseases, risk factors, and TICS-27. Note that functional decline depends
on cognition, but cognitive decline does not depend on function.
82

Joint Treatment Effect Calibration

To estimate joint effect scenarios, such as a 20% reduction in cognitive decline and a
40% reduction in functional decline, we proceed as follows. A calibration parameter is inserted
into the ADL transition model. We impose the 20% reduction in cognitive decline, which has
spillover effects to the prevalence of ADLs in the simulation due to the role of TICS-27 in the
ADL transition model. We then adjust the calibration parameter to yield a 40% reduction in
acquisition of functional limitations in the first two-year period of the simulation.
83
Table 3.1: Demographic Characteristics of the Simulated Population
Cohort
Simulated
observations
at entry
Mean age
at entry
(years)
Median life
expectancy
(years)
Male % at
entry
Mean years
education at
entry
Mean TICS
at entry
Any ADL at
entry
Any IADL
at entry
51-55 2,288 53.8 27.0 26.4% 12.74 11.80 17.4% 11.0%
56-60 16,432 58.6 23.0 25.1% 11.90 11.60 26.5% 15.0%
61-65 20,488 62.9 19.0 27.5% 11.79 11.10 28.2% 16.5%
66-70 37,960 68.3 15.0 31.5% 11.68 11.62 20.6% 8.1%
71-75 63,752 73.0 13.0 34.4% 11.77 11.79 19.3% 8.2%
76-80 63,752 78.0 9.0 33.2% 12.17 11.73 20.1% 10.2%
81-85 53,040 83.0 7.0 41.2% 12.43 12.26 24.1% 14.0%
All 257,712 71.3 13.0 32.5% 11.96 12.06 22.2% 11.2%
Note: ADL = Activities of Daily Living, IADL = Instrumental Activities of Daily Living

84
Figure 3.1: Notional Allocation of Treatment Costs by Payment Model

85
Figure 3.2: Share of Discounted QALY Gain Under and Over 65 Years of Age, by
Treatment Scenario and Age Cohort

- 0.1 0.2 0.3 0.4 0.5 0.6
51-55 years
56-60 years
61-65 years
66-70 years
71-75 years
76-80 years
81-85 years
Discounted QALY gain per patient
Cohort
Scenario 1: 20% cognitive, 40% functional
effect
<65 years 65+ years
- 0.1 0.2 0.3 0.4 0.5 0.6
51-55 years
56-60 years
61-65 years
66-70 years
71-75 years
76-80 years
81-85 years
Discounted QALY gain per patient
Cohort
Scenario 2: 10% cognitive, 20% functional
effect
<65 years 65+ years
- 0.1 0.2 0.3 0.4 0.5 0.6
51-55 years
56-60 years
61-65 years
66-70 years
71-75 years
76-80 years
81-85 years
Discounted QALY gain per patient
Cohort
Scenario 3: 40% cognitive 20% functional
effect
<65 years 65+ years
- 0.1 0.2 0.3 0.4 0.5 0.6
51-55 years
56-60 years
61-65 years
66-70 years
71-75 years
76-80 years
81-85 years
Discounted QALY gain per patient
Cohort
Scenario 4: 20% cognitive, 10% functional
effect
<65 years 65+ years
86
Figure 3.3: Medical Expenditures Estimates and Value-Based Cost of Therapy, by
Treatment Scenario and Age Cohort

-$16,000
-$12,000
-$8,000
-$4,000
$0
$4,000
$8,000
51-55
years
56-60
years
61-65
years
66-70
years
71-75
years
76-80
years
81-85
years
Discounted costs or savings
Age Cohort
Scenario 1: 20% cognitive, 40% functional
effect
Medical Expenditure Savings Value-Based Cost of Therapy
-$16,000
-$12,000
-$8,000
-$4,000
$0
$4,000
$8,000
51-55
years
56-60
years
61-65
years
66-70
years
71-75
years
76-80
years
81-85
years
Discounted costs or savings
Age Cohort
Scenario 2: 10% cognitive, 20% functional
effect
Medical Expenditure Savings Value-Based Cost of Therapy
-$16,000
-$12,000
-$8,000
-$4,000
$0
$4,000
$8,000
51-55
years
56-60
years
61-65
years
66-70
years
71-75
years
76-80
years
81-85
years
Discounted costs or savings
Age Cohort
Scenario 3: 40% cognitive, 20% functional
effect
Medical Expenditure Savings Value-Based Cost of Therapy
-$16,000
-$12,000
-$8,000
-$4,000
$0
$4,000
$8,000
51-55
years
56-60
years
61-65
years
66-70
years
71-75
years
76-80
years
81-85
years
Discounted costs or savings
Age Cohort
Scenario 4: 20% cognitive, 10% functional
effect
Medical Expenditure Savings Value-Based Cost of Therapy
87
Figure 3.4: Net Costs or Benefits Accrued to Private and Public Payer under Treatment
Scenarios, by Age Cohort and Payment Model

Note: P4P = pay for performance

-$10,000
$0
$10,000
$20,000
$30,000
$40,000
$50,000
$60,000
$70,000
$80,000
Private
Payer
Public
Payer
Private
Payer
Public
Payer
Private
Payer
Public
Payer
Time of
Treatment
Payment
Constant
Installments
P4P
Installments
Discounted net costs/benefits
Scenario 1: 20% cognitive, 40% functional
effect
51-55 years 56-60 years 61-65 years 66-70 years
71-75 years 76-80 years 81-85 years
-$10,000
$0
$10,000
$20,000
$30,000
$40,000
$50,000
$60,000
$70,000
$80,000
Private
Payer
Public
Payer
Private
Payer
Public
Payer
Private
Payer
Public
Payer
Time of
Treatment
Payment
Constant
Installments
P4P
Installments
Discounted net costs/benefits
Scenario 2: 10% cognitive, 20% functional
effect
51-55 years 56-60 years 61-65 years 66-70 years
71-75 years 76-80 years 81-85 years
-$10,000
$0
$10,000
$20,000
$30,000
$40,000
$50,000
$60,000
$70,000
$80,000
Private
Payer
Public
Payer
Private
Payer
Public
Payer
Private
Payer
Public
Payer
Time of
Treatment
Payment
Constant
Installments
P4P
Installments
Discounted net costs/benefits
Scenario 3: 40% cognitive, 20% functional
effect
51-55 years 56-60 years 61-65 years 66-70 years
71-75 years 76-80 years 81-85 years
-$10,000
$0
$10,000
$20,000
$30,000
$40,000
$50,000
$60,000
$70,000
$80,000
Private
Payer
Public
Payer
Private
Payer
Public
Payer
Private
Payer
Public
Payer
Time of
Treatment
Payment
Constant
Installments
P4P
Installments
Discounted net costs/benefits
Scenario 4: 20% cognitive, 10% functional
effect
51-55 years 56-60 years 61-65 years 66-70 years
71-75 years 76-80 years 81-85 years
88
Appendix Table 3.1: Discounted QALY Benefit Under and Over 65, by Treatment Effect
Scenario and Cohort

Cohort
20% cognitive, 40%
functional effect
10% cognitive, 20%
functional effect
40% cognitive, 20%
functional effect
20% cognitive, 10%
functional effect
<65 years 65+ years <65 years 65+ years <65 years 65+ years <65 years 65+ years
51-55 years 0.18 0.34 0.10 0.17 0.17 0.32 0.08 0.14
56-60 years 0.12 0.43 0.06 0.23 0.09 0.41 0.04 0.20
61-65 years 0.04 0.48 0.02 0.25 0.02 0.44 0.01 0.22
66-70 years - 0.44 - 0.23 - 0.40 - 0.20
71-75 years - 0.40 - 0.20 - 0.35 - 0.17
76-80 years - 0.35 - 0.18 - 0.31 - 0.16
81-85 years - 0.30 - 0.15 - 0.25 - 0.13
89
Appendix Table 3.2: Discounted Medical Expenses, by Treatment Effect Scenario and
Cohort
Cohort
20% cognitive, 40%
functional effect
10% cognitive, 20%
functional effect
40% cognitive, 20%
functional effect
20% cognitive, 10%
functional effect
51-55 years -$3,143 -$2,468 -$4,311 -$3,151
56-60 years -$3,201 -$2,529 -$5,721 -$4,956
61-65 years -$4,287 -$3,656 -$7,424 -$6,206
66-70 years -$4,450 -$3,839 -$8,017 -$7,020
71-75 years -$4,230 -$3,523 -$7,691 -$6,559
76-80 years -$3,631 -$3,036 -$6,825 -$6,275
81-85 years -$3,557 -$3,112 -$6,558 -$6,051
Note: These are medical expenditures per patient by treatment effect scenario and age cohort,
before taking the cost of therapy into account.

90
Appendix Table 3.3: Value-Based Cost of Therapy, by Treatment Effect Scenario and
Cohort
Cohort
20% cognitive, 40%
functional effect
10% cognitive, 20%
functional effect
40% cognitive, 20%
functional effect
20% cognitive, 10%
functional effect
51-55 years $15,703 $3,057 $14,677 $6,765
56-60 years $16,465 $3,673 $14,827 $7,406
61-65 years $15,517 $3,425 $14,007 $7,008
66-70 years $13,347 $2,873 $12,083 $6,119
71-75 years $11,880 $2,509 $10,526 $5,235
76-80 years $10,570 $2,233 $9,396 $4,654
81-85 years $8,871 $1,877 $7,493 $3,756

91
Appendix Table 3.4 Sensitivity Analysis (Durability of Treatment Effect)

Effect Durability: 2 years
10% cognitive, 20% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years $2,752 $804 $3,234 $322 $2,520 -$1,046
56-60 years $2,258 $1,243 $2,880 $621 $2,306 -$1,571
61-65 years $2 $3,936 $856 $3,082 $648 $804
66-70 years $3,676 $3,676 $1,641
71-75 years $3,986 $3,986 $2,362
76-80 years $4,556 $4,556 $3,328
81-85 years $4,656 $4,656 $3,806
20% cognitive, 40% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years $4,417 $1,115 $5,140 $392 $4,724 $808
56-60 years $4,383 $2,364 $5,551 $1,195 $4,917 $1,829
61-65 years $121 $7,213 $1,702 $5,632 $1,502 $5,832
66-70 years $7,460 $7,460 $7,460
71-75 years $7,698 $7,698 $7,698
76-80 years $8,750 $8,750 $8,750
81-85 years $8,962 $8,962 $8,962
20% cognitive, 10% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years $2,093 $175 $2,372 -$103 $2,136 $133
56-60 years $1,733 $1,233 $2,245 $721 $2,007 $959
61-65 years -$1 $3,335 $697 $2,638 $622 $2,712
66-70 years $3,258 $3,258 $3,258
71-75 years $3,593 $3,593 $3,593
76-80 years $4,175 $4,175 $4,175
81-85 years $4,388 $4,388 $4,388
40% cognitive, 20% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years $4,365 $1,149 $5,095 $419 $4,679 $836
56-60 years $3,455 $1,715 $4,337 $832 $3,839 $1,331
61-65 years $0 $6,256 $1,336 $4,919 $1,188 $5,067
66-70 years $6,310 $6,310 $6,310
71-75 years $6,732 $6,732 $6,732
76-80 years $7,944 $7,944 $7,944
81-85 years $8,160 $8,160 $8,160

92
Effect Durability: 4 years
10% cognitive, 20% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$4,579 $912 $5,300 $192 $4,829 -$938
56-60 years
$3,644 $3,127 $4,822 $1,949 $4,503 $314
61-65 years
-$673 $8,342 $979 $6,689 $850 $5,209
66-70 years
$7,410 $7,410 $6,226
71-75 years
$7,695 $7,695 $6,893
76-80 years
$8,786 $8,786 $8,507
81-85 years
$8,732 $8,732 $8,738
20% cognitive, 40% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$8,086 $1,839 $9,387 $538 $8,596 $1,329
56-60 years
$6,968 $6,095 $9,210 $3,853 $8,264 $4,799
61-65 years
-$1,120 $15,367 $1,922 $12,324 $1,795 $12,451
66-70 years
$14,253 $14,253 $14,253
71-75 years
$14,784 $14,784 $14,784
76-80 years
$16,639 $16,639 $16,639
81-85 years
$16,695 $16,695 $16,695
20% cognitive, 10% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$3,792 $320 $4,303 -$191 $3,866 $246
56-60 years
$2,995 $2,836 $3,989 $1,842 $3,595 $2,236
61-65 years
-$603 $7,334 $790 $5,941 $753 $5,979
66-70 years
$6,687 $6,687 $6,687
71-75 years
$7,088 $7,088 $7,088
76-80 years
$8,069 $8,069 $8,069
81-85 years
$8,115 $8,115 $8,115
40% cognitive, 20% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$7,805 $1,658 $9,027 $436 $8,235 $1,228
56-60 years
$5,752 $5,099 $7,602 $3,249 $6,830 $4,021
61-65 years
-$1,109 $13,784 $1,540 $11,135 $1,456 $11,218
66-70 years
$12,848 $12,848 $12,848
71-75 years
$13,471 $13,471 $13,471
76-80 years
$15,572 $15,572 $15,572
81-85 years
$15,427 $15,427 $15,427

93
Effect Durability: 6 years
10% cognitive, 20% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$6,424 $1,449 $7,446 $427 $7,284 -$401
56-60 years
$3,891 $5,476 $5,490 $3,877 $5,364 $2,663
61-65 years
-$1,478 $12,774 $907 $10,388 $850 $9,642
66-70 years
$11,204 $11,204 $10,851
71-75 years
$11,188 $11,188 $11,172
76-80 years
$12,390 $12,390 $12,902
81-85 years
$11,636 $11,636 $12,242
20% cognitive, 40% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$11,962 $2,646 $13,889 $718 $12,716 $1,891
56-60 years
$7,341 $11,027 $10,469 $7,899 $9,605 $8,763
61-65 years
-$2,568 $23,157 $1,793 $18,795 $1,795 $18,793
66-70 years
$21,227 $21,227 $21,227
71-75 years
$21,562 $21,562 $21,562
76-80 years
$23,458 $23,458 $23,458
81-85 years
$22,675 $22,675 $22,675
20% cognitive, 10% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$5,869 $846 $6,694 $21 $6,053 $662
56-60 years
$3,252 $5,064 $4,642 $3,673 $4,268 $4,047
61-65 years
-$1,315 $11,226 $727 $9,185 $753 $9,159
66-70 years
$10,260 $10,260 $10,260
71-75 years
$10,446 $10,446 $10,446
76-80 years
$11,402 $11,402 $11,402
81-85 years
$10,978 $10,978 $10,978
40% cognitive, 20% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$11,514 $2,433 $13,326 $621 $12,166 $1,782
56-60 years
$6,273 $9,354 $8,892 $6,735 $8,151 $7,476
61-65 years
-$2,543 $21,621 $1,412 $17,666 $1,456 $17,622
66-70 years
$19,580 $19,580 $19,580
71-75 years
$19,989 $19,989 $19,989
76-80 years
$22,303 $22,303 $22,303
81-85 years
$21,017 $21,017 $21,017

94
Effect Durability: 8 years
10% cognitive, 20% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$8,092 $2,385 $9,403 $1,074 $9,535 $534
56-60 years
$3,364 $9,110 $5,459 $7,015 $5,559 $6,296
61-65 years
-$2,191 $16,573 $844 $13,538 $850 $13,441
66-70 years
$14,227 $14,227 $14,542
71-75 years
$14,451 $14,451 $15,146
76-80 years
$15,416 $15,416 $16,580
81-85 years
$13,706 $13,706 $14,751
20% cognitive, 40% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$15,061 $5,024 $17,715 $2,369 $16,388 $3,697
56-60 years
$6,427 $17,387 $10,406 $13,408 $9,863 $13,951
61-65 years
-$3,836 $30,017 $1,681 $24,501 $1,795 $24,386
66-70 years
$27,281 $27,281 $27,281
71-75 years
$27,874 $27,874 $27,874
76-80 years
$29,165 $29,165 $29,165
81-85 years
$26,825 $26,825 $26,825
20% cognitive, 10% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$7,531 $1,812 $8,654 $688 $7,891 $1,451
56-60 years
$2,786 $8,410 $4,621 $6,575 $4,414 $6,782
61-65 years
-$1,992 $15,052 $666 $12,394 $753 $12,307
66-70 years
$13,231 $13,231 $13,231
71-75 years
$13,506 $13,506 $13,506
76-80 years
$14,264 $14,264 $14,264
81-85 years
$12,999 $12,999 $12,999
40% cognitive, 20% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$14,835 $4,213 $17,243 $1,805 $15,822 $3,227
56-60 years
$5,388 $15,651 $8,847 $12,192 $8,422 $12,617
61-65 years
-$3,773 $28,441 $1,303 $23,365 $1,456 $23,212
66-70 years
$25,578 $25,578 $25,578
71-75 years
$26,195 $26,195 $26,195
76-80 years
$28,086 $28,086 $28,086
81-85 years
$25,235 $25,235 $25,235

95
Effect Durability: 10 years
10% cognitive, 20% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$8,670 $4,105 $10,260 $2,515 $10,675 $2,254
56-60 years
$2,624 $13,100 $5,228 $10,496 $5,559 $10,287
61-65 years
-$2,811 $20,081 $789 $16,481 $850 $16,948
66-70 years
$17,223 $17,223 $18,207
71-75 years
$17,243 $17,243 $18,537
76-80 years
$17,619 $17,619 $19,275
81-85 years
$15,221 $15,221 $16,575
20% cognitive, 40% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$16,295 $7,834 $19,432 $4,697 $18,164 $5,965
56-60 years
$5,122 $24,470 $10,000 $19,592 $9,863 $19,729
61-65 years
-$5,052 $36,529 $1,573 $29,903 $1,795 $29,681
66-70 years
$33,115 $33,115 $33,115
71-75 years
$33,250 $33,250 $33,250
76-80 years
$33,619 $33,619 $33,619
81-85 years
$29,754 $29,754 $29,754
20% cognitive, 10% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$8,205 $3,441 $9,610 $2,035 $8,890 $2,755
56-60 years
$2,132 $11,973 $4,416 $9,689 $4,414 $9,691
61-65 years
-$2,586 $18,350 $613 $15,151 $753 $15,012
66-70 years
$16,277 $16,277 $16,277
71-75 years
$16,276 $16,276 $16,276
76-80 years
$16,472 $16,472 $16,472
81-85 years
$14,478 $14,478 $14,478
40% cognitive, 20% functional
Time of Treatment Payment Constant Installments P4P Installments
Private Payer Public Payer Private Payer Public Payer Private Payer Public Payer
51-55 years
$16,119 $7,479 $19,102 $4,496 $17,776 $5,822
56-60 years
$4,115 $22,555 $8,450 $18,220 $8,422 $18,248
61-65 years
-$4,953 $34,963 $1,198 $28,812 $1,456 $28,553
66-70 years
$31,404 $31,404 $31,404
71-75 years
$31,718 $31,718 $31,718
76-80 years
$32,398 $32,398 $32,398
81-85 years
$28,288 $28,288 $28,288

96
Conclusions

The advent of DMT interventions may provide patients with AD an opportunity to enjoy
a longer and healthier life. With their arrival, there are concerns revolving equitable access, as
well as concerns regarding barriers to coverage – especially for the privately insured. There also
remain questions over the etiology and risk factors for AD, especially the potential vascular
mechanisms at play.
In Chapter 2, we find evidence that DMT treatment – in concert with timely diagnosis
and access – favors those with lower SES and may offer compression of health disparities
between educational groups. This is due in part to the earlier age of incident MCI patients that
have less educational attainment as well as their poorer baseline cognitive and functional status –
essentially their greater burden of disease. While these patients receive the greatest benefits from
timely diagnosis and access, these patients with lower SES also expect to see greater
consequence from delayed diagnosis and treatment. Given this, delays in diagnosis and treatment
pose a risk that may diminish or erase the equalizing force of new therapies.
There are a number of reasons why there is greater delayed diagnosis in patients with less
education. Such delays may be driven by barriers to healthcare access such as cost or insurance
coverage,
30
cultural or language barriers,
31,32
and limited awareness of dementia as a disease
(rather than as a natural consequence of aging).
30
And even though Medicare has included
cognitive screening in its annual wellness visits, only 23% of fee-for-service beneficiaries
actually receive a cognitive assessment at their wellness visit each year, and notably, this rate is
30% in Medicare Advantage patients – who tend to be higher SES.
83
As such, better strategies
need to be developed to increase earlier detection, particularly in lower SES individuals.
97
We also find that the aggregate social value benefit from a hypothetical DMT – under
Treatment Effect II – falls by $5.18 billion (16%) due to a 2-year delay in diagnosis and access,
and $10.28 billion (32%) due to a 4-year delay. These losses from delay are considerable, since
on average it costs $985 million alone to bring a new drug to market.
84
We find this forgone
social value benefit to be more concentrated in the less educated, since they experience greater
losses from delay and represent a majority of incident MCI cases. Strategies should improve
earlier detection in lower SES patients, to help minimize losses in social value alongside
compressing health disparities.
While Chapter 2 centered on baseline educational attainment as one measure of SES,
future research may want to consider the consequences of delayed diagnosis and access by other
factors such as income and sex. Other interesting subgroups worth exploring may include age of
onset (i.e., early versus late onset of disease) and genetic status.
In Chapter 3, we identified a potential barrier – upfront payment – that could delay or
prevent access to treatment. We find that even if DMTs are priced at a fraction of their overall
welfare gain, they may result in net losses accrued to private insurers, especially for patients who
are closer to the Medicare age of eligibility (i.e., 65 years old). If potential DMTs are to require
upfront payment at the time of treatment, private insurers may choose to delay the treatment of
such patients. Alternatively, insurers may increase insurance premiums or put such drugs on high
copay formulary tiers, which could result in patients shouldering a larger share of the cost of
treatment. Either of these approaches may result in suboptimal outcomes since fewer patients
receive timely access to these emerging treatments.
Constant and pay-for-performance installment payment models provide possible
solutions by spreading the cost of treatment over the remainder of a patient’s lifetime. Under
98
these models, private insurers would no longer face the disincentive to provide care given that
their share of the treatment cost would no longer make the net benefit of treatment negative. Pay-
for-performance installments may serve as the first-best solution from an actuarial perspective;
however, their implementation costs may outweigh their benefits. Constant installments result in
net positive benefits to private insurers for all the age cohorts we studied. However, in a deeper
dive, we find that treatment durability of 2 years or less would eliminate the need for innovative
payment contracts, but treatment durability of 4 years or longer would create misaligned
incentives. As such, the uncertainty about benefit accrual poses challenges to payment design.
103
If alternative payment models are pursued, public and private insurers would need to cooperate
and agree on a mechanism to transfer patients’ payment obligations between one another when
they switch plans or become eligible for Medicare.
104
As this study was based on stylized
treatment effects since the treatment effects and persistence of AD DMTs still remain unclear,
future research should consider the real-world effects and durability of AD DMTs.
While timely diagnosis and access, as well as alternative payment models for AD DMTs
are important areas to study since DMTs may offer a valuable clinical benefit, it remains
worthwhile to study the risk factors for AD, as these non-therapeutic options may have the
potential to modify disease progression. Furthermore, because the etiology of AD remains
unclear. In Chapter 1, we find supportive evidence that aortic valve replacement surgeries may
delay the onset of ADRDs. In patients with severe AS, SAVR and TAVR were associated with a
protective effect and reduced the risk of ADRDs in the earlier years post-index. However, over
time this reduction in risk diminished and the median time-to-ADRD diagnosis was not
significantly different between the SAVR and MM treatment groups, or between the TAVR era
and pre-TAVR era. This finding suggests there may be an association between heart valve
99
disease and ADRDs. As this analysis – the first of its kind to study the long-term effects of heart
valve replacement surgery on ADRD outcomes – utilized administrative claims data to identify
severe AS based on a validated claims algorithm, future follow-on research should consider use
of registries or electronic medical record (EMR) data to understand the long-term cognitive
effects of heart valve replacement surgery in this population. Future research may also wish to
consider cognitive scoring as well, which may be available through survey data.
Given the rapidly aging US population, in which older adults 85+ represent the fastest-
growing segment of the population,
57
along with the greater incidence of dementia at older age,
58

it is important to understand the mechanisms that give rise to these diseases for better prevention
and/or management. Our research indicates that there may be a link between heart valve disease
and ADRDs, and highlights the value of studying the vascular mechanisms behind cognitive
outcomes.
Concluding Remarks
The arrival of DMT interventions may provide patients with an effective therapy for a
condition where no other course-altering options previously existed. However, as these treatment
options emerge, it is important to be aware of their implications for vulnerable populations,
consider payment innovations necessary to optimize treatment access – especially in younger
patients, and continue investigating the mechanisms that give rise to AD.

100
References

1. Ashraf GM, Chibber S, Mohammad, et al. Recent Updates on the Association Between
Alzheimer's Disease and Vascular Dementia. Med Chem. 2016;12(3):226-37.
doi:10.2174/1573406411666151030111820

2. Kumar A, Sidhu J, Goyal A, Tsao JW. StatPearls: Alzheimer Disease. Accessed
September 9, 2022, https://www.ncbi.nlm.nih.gov/books/NBK499922/

3. Neumann PJ, Kuntz KM, Leon J, et al. Health utilities in Alzheimer's disease: a cross-
sectional study of patients and caregivers. Med Care. Jan 1999;37(1):27-32.
doi:10.1097/00005650-199901000-00005

4. Rajan KB, Weuve J, Barnes LL, McAninch EA, Wilson RS, Evans DA. Population
estimate of people with clinical Alzheimer's disease and mild cognitive impairment in the United
States (2020-2060). Alzheimers Dement. May 27 2021;doi:10.1002/alz.12362

5. Aging NIo. Alzheimer's Disease Fact Sheet. Accessed August 12th, 2022,
https://www.nia.nih.gov/health/alzheimers-disease-fact-
sheet#:~:text=Alzheimer's%20disease%20is%20currently%20ranked,of%20dementia%20among
%20older%20adults.

6. Mayo Clinic. Alzheimer's Disease: Symptoms and Causes. Accessed August 12th, 2022,
https://www.mayoclinic.org/diseases-conditions/alzheimers-disease/symptoms-causes/syc-
20350447

7. Hurd MD, Martorell P, Delavande A, Mullen KJ, Langa KM. Monetary costs of dementia
in the United States. N Engl J Med. Apr 4 2013;368(14):1326-34. doi:10.1056/NEJMsa1204629

8. Zissimopoulos J, Crimmins E, St Clair P. The Value of Delaying Alzheimer's Disease
Onset. Forum Health Econ Policy. Nov 2014;18(1):25-39. doi:10.1515/fhep-2014-0013

9. U.S. Food and Drug Administration. FDA Grants Accelerated Approval for Alzheimer’s
Drug. https://www.fda.gov/news-events/press-announcements/fda-grants-accelerated-approval-
alzheimers-drug

10. Cummings J, Lee G, Nahed P, et al. Alzheimer's disease drug development pipeline:
2022. Alzheimers Dement (N Y). 2022;8(1):e12295. doi:10.1002/trc2.12295

11. U.S. National Library of Medicine. Dominantly Inherited Alzheimer Network Trial: An
Opportunity to Prevent Dementia. A Study of Potential Disease Modifying Treatments in
Individuals at Risk for or With a Type of Early Onset Alzheimer's Disease Caused by a Genetic
Mutation. Master Protocol DIAN-TU001 (DIAN-TU). Updated June 2021.
https://clinicaltrials.gov/ct2/show/NCT01760005

101
12. Boustani M, Peterson B, Hanson L, Harris R, Lohr KN, Force USPST. Screening for
dementia in primary care: a summary of the evidence for the U.S. Preventive Services Task
Force. Ann Intern Med. Jun 03 2003;138(11):927-37. doi:10.7326/0003-4819-138-11-
200306030-00015

13. Bradford A, Kunik ME, Schulz P, Williams SP, Singh H. Missed and delayed diagnosis
of dementia in primary care: prevalence and contributing factors. Alzheimer Dis Assoc Disord.
2009 Oct-Dec 2009;23(4):306-14. doi:10.1097/WAD.0b013e3181a6bebc

14. Kotagal V, Langa KM, Plassman BL, et al. Factors associated with cognitive evaluations
in the United States. Neurology. Jan 06 2015;84(1):64-71.
doi:10.1212/WNL.0000000000001096

15. Taylor DH, Østbye T, Langa KM, Weir D, Plassman BL. The accuracy of Medicare
claims as an epidemiological tool: the case of dementia revisited. J Alzheimers Dis.
2009;17(4):807-15. doi:10.3233/JAD-2009-1099

16. Prince M, Bryce R, Ferri C. World Alzheimer's Report 2011: The Benefits of Early
Diagnosis and Intervention. 2011.

17. The Alzheimer's Foundation of America (AFA). I CAN: Investigating Caregivers’
Attitudes and Needs. Accessed July 24, 2022, 2022. https://alzfdn.org/wp-
content/uploads/2017/04/Caregiver_Survey_1_2006.pdf

18. Albert MS, DeKosky ST, Dickson D, et al. The diagnosis of mild cognitive impairment
due to Alzheimer's disease: recommendations from the National Institute on Aging-Alzheimer's
Association workgroups on diagnostic guidelines for Alzheimer's disease. Alzheimers Dement.
May 2011;7(3):270-9. doi:10.1016/j.jalz.2011.03.008

19. Petersen RC. Mild Cognitive Impairment. Continuum (Minneap Minn). Apr 2016;22(2
Dementia):404-18. doi:10.1212/CON.0000000000000313

20. Alzheimer's Association. 2018 Alzheimer's disease facts and figures. Alzheimer's
Dement. 2018;14(3):367‐429.

21. Rasmussen J, Langerman H. Alzheimer's Disease - Why We Need Early Diagnosis.
Degener Neurol Neuromuscul Dis. 2019;9:123-130. doi:10.2147/DNND.S228939

22. McCarten J, Anderson P, Kuskowski M, Jonk Y, Dysken M. Changes in outpatient costs
following screening and diagnosis of cognitive impairment. presented at: International
Conference on Alzheimer’s Disease; 2010; Honolulu, HI.

23. de Vugt ME, Verhey FR. The impact of early dementia diagnosis and intervention on
informal caregivers. Prog Neurobiol. Nov 2013;110:54-62. doi:10.1016/j.pneurobio.2013.04.005

102
24. Robinson RL, Rentz DM, Andrews JS, et al. Costs of Early Stage Alzheimer's Disease in
the United States: Cross-Sectional Analysis of a Prospective Cohort Study (GERAS-US)1. J
Alzheimers Dis. 2020;75(2):437-450. doi:10.3233/JAD-191212

25. Weimer DL, Sager MA. Early identification and treatment of Alzheimer's disease: social
and fiscal outcomes. Alzheimers Dement. May 2009;5(3):215-26. doi:10.1016/j.jalz.2009.01.028

26. Moritz DJ, Petitti DB. Association of education with reported age of onset and severity of
Alzheimer's disease at presentation: implications for the use of clinical samples. Am J Epidemiol.
Feb 15 1993;137(4):456-62. doi:10.1093/oxfordjournals.aje.a116694

27. Ross GW, Abbott RD, Petrovitch H, et al. Frequency and characteristics of silent
dementia among elderly Japanese-American men. The Honolulu-Asia Aging Study. JAMA. Mar
12 1997;277(10):800-5.

28. Holm E, Jacobsen KK, de Lony TB, et al. Frequency of missed or delayed diagnosis in
dementia is associated with neighborhood socioeconomic status. Alzheimers Dement (N Y).
2022;8(1):e12271. doi:10.1002/trc2.12271

29. Qian W, Schweizer TA, Fischer CE. Impact of socioeconomic status on initial clinical
presentation to a memory disorders clinic. Int Psychogeriatr. Apr 2014;26(4):597-603.
doi:10.1017/S1041610213002299

30. Ortiz F, Fitten LJ. Barriers to healthcare access for cognitively impaired older Hispanics.
Alzheimer Dis Assoc Disord. 2000 Jul-Sep 2000;14(3):141-50. doi:10.1097/00002093-
200007000-00005

31. Boise L, Camicioli R, Morgan DL, Rose JH, Congleton L. Diagnosing dementia:
perspectives of primary care physicians. Gerontologist. Aug 1999;39(4):457-64.
doi:10.1093/geront/39.4.457

32. Hinton L, Franz C, Friend J. Pathways to dementia diagnosis: evidence for cross-ethnic
differences. Alzheimer Dis Assoc Disord. 2004 Jul-Sep 2004;18(3):134-44.
doi:10.1097/01.wad.0000127444.23312.ff

33. Jones R, Chow T, Gatz M. Asian Americans and Alzheimer’s disease: assimilation,
culture, and beliefs. J Aging Stud. 2006;20(1):11–25.

34. Ikegami N. Fee-for-service payment - an evil practice that must be stamped out? Int J
Health Policy Manag. Feb 2015;4(2):57-9. doi:10.15171/ijhpm.2015.26

35. Culyer A. The dictionary of health economics. Cheltenham (UK); 2005.

36. Slocomb T, Werner M, Haack T, Valluri S, Rader B. New Payment and Financing
Models for Curative Regenerative Medicines. Updated July/August 2017.
103
https://invivo.pharmaintelligence.informa.com/IV005132/New-Payment-And-Financing-
Models-For-Curative-Regenerative-Medicines

37. Justin BN, Turek M, Hakim AM. Heart disease as a risk factor for dementia. Clin
Epidemiol. 2013;5:135-45. doi:10.2147/CLEP.S30621

38. A Armstrong R. Risk factors for Alzheimer's disease. Folia Neuropathol. 2019;57(2):87-
105. doi:10.5114/fn.2019.85929

39. Corder EH, Ervin JF, Lockhart E, Szymanski MH, Schmechel DE, Hulette CM.
Cardiovascular damage in Alzheimer disease: autopsy findings from the Bryan ADRC. J Biomed
Biotechnol. Jun 30 2005;2005(2):189-97. doi:10.1155/JBB.2005.189

40. Oliveira-Filho J, Massaro AR, Yamamoto F, Bustamante L, Scaff M. Stroke as the first
manifestation of calcific aortic stenosis. Cerebrovasc Dis. 2000 Sep-Oct 2000;10(5):413-6.
doi:10.1159/000016099

41. de la Torre JC. Cardiovascular risk factors promote brain hypoperfusion leading to
cognitive decline and dementia. Cardiovasc Psychiatry Neurol. 2012;2012:367516.
doi:10.1155/2012/367516

42. Belohlavek M, Jiamsripong P, Calleja AM, et al. Patients with Alzheimer disease have
altered transmitral flow: echocardiographic analysis of the vortex formation time. J Ultrasound
Med. Nov 2009;28(11):1493-500. doi:10.7863/jum.2009.28.11.1493

43. Reitz C, Brickman AM, Luchsinger JA, Wu WE, Small SA, Tang MX. Frequency of
subclinical heart disease in elderly persons with dementia. Am J Geriatr Cardiol. 2007 May-Jun
2007;16(3):183-8. doi:10.1111/j.1076-7460.2007.06511.x

44. Chung HE, Chen J, Ghosalkar D, et al. Aortic Valve Calcification Is Associated with
Future Cognitive Impairment. J Alzheimers Dis Rep. Apr 29 2021;5(1):337-343.
doi:10.3233/ADR-200253

45. Wolters FJ, Bos D, Vernooij MW, et al. Aortic Valve Calcification and the Risk of
dementia: A Population-Based Study. J Alzheimers Dis. 2017;55(3):893-897. doi:10.3233/JAD-
160871

46. Ramlawi B, Ramchandani M, Reardon MJ. Surgical Approaches to Aortic Valve
Replacement and Repair-Insights and Challenges. Interv Cardiol. Mar 2014;9(1):32-36.
doi:10.15420/icr.2011.9.1.32

47. Hu PP. TAVR and SAVR: Current Treatment of Aortic Stenosis. Clin Med Insights
Cardiol. 2012;6:125-39. doi:10.4137/CMC.S7540

104
48. Khan MM, Herrmann N, Gallagher D, et al. Cognitive Outcomes After Transcatheter
Aortic Valve Implantation: A Metaanalysis. J Am Geriatr Soc. 02 2018;66(2):254-262.
doi:10.1111/jgs.15123

49. Gu S, Coakley D, Chan D, et al. Does Transcatheter Aortic Valve Implantation for Aortic
Stenosis Impact on Cognitive Function? Cardiol Rev. 2020 May/Jun 2020;28(3):135-139.
doi:10.1097/CRD.0000000000000279

50. Oldham MA, Vachon J, Yuh D, Lee HB. Cognitive Outcomes After Heart Valve
Surgery: A Systematic Review and Meta-Analysis. J Am Geriatr Soc. 12 2018;66(12):2327-
2334. doi:10.1111/jgs.15601

51. Talbot-Hamon C, Afilalo J. Cognitive Function After Transcatheter Aortic Valve
Replacement: Reassuring Findings For Now. J Am Geriatr Soc. 02 2018;66(2):227-228.
doi:10.1111/jgs.15185

52. National Institute on Aging. What Causes Alzheimer's Disease? Accessed August 23,
2021, https://www.nia.nih.gov/health/what-causes-alzheimers-disease

53. Armstrong R. Risk factors for Alzheimer's disease. Folia Neuropathol. 2019;57(2):87-
105. doi:10.5114/fn.2019.85929

54. Clark MA, Arnold SV, Duhay FG, et al. Five-year clinical and economic outcomes
among patients with medically managed severe aortic stenosis: results from a Medicare claims
analysis. Circ Cardiovasc Qual Outcomes. Sep 01 2012;5(5):697-704.
doi:10.1161/CIRCOUTCOMES.112.966002

55. Chao A, Picard MH, Passeri JJ, Cui J, Nethery RC, Wasfy JH. Effect of Availability of
Transcatheter Aortic-Valve Implantation on Survival for all Patients With Severe Aortic
Stenosis. Am J Cardiol. 06 15 2021;149:72-77. doi:10.1016/j.amjcard.2021.03.016

56. Wolters FJ, Chibnik LB, Waziry R, et al. Twenty-seven-year time trends in dementia
incidence in Europe and the United States: The Alzheimer Cohorts Consortium. Neurology. 08
04 2020;95(5):e519-e531. doi:10.1212/WNL.0000000000010022

57. National Institute on Aging. Why population aging matters: A global perspective. Vol.
07–6134. 2007.

58. Brookmeyer R, Gray S, Kawas C. Projections of Alzheimer's disease in the United States
and the public health impact of delaying disease onset. Am J Public Health. Sep
1998;88(9):1337-42. doi:10.2105/ajph.88.9.1337

59. Knopman DS, Jones DT, Greicius MD. Failure to demonstrate efficacy of aducanumab:
An analysis of the EMERGE and ENGAGE trials as reported by Biogen, December 2019.
Alzheimers Dement. 04 2021;17(4):696-701. doi:10.1002/alz.12213

105
60. Alexander GC, Emerson S, Kesselheim AS. Evaluation of Aducanumab for Alzheimer
Disease: Scientific Evidence and Regulatory Review Involving Efficacy, Safety, and Futility.
JAMA. May 04 2021;325(17):1717-1718. doi:10.1001/jama.2021.3854

61. Qiu C, Bäckman L, Winblad B, Agüero-Torres H, Fratiglioni L. The influence of
education on clinically diagnosed dementia incidence and mortality data from the Kungsholmen
Project. Arch Neurol. Dec 2001;58(12):2034-9. doi:10.1001/archneur.58.12.2034

62. Chan MY, Han L, Carreno CA, et al. Long-term prognosis and educational determinants
of brain network decline in older adult individuals. Nat Aging. Nov 2021;1(11):1053-1067.
doi:10.1038/s43587-021-00125-4

63. Sharp ES, Gatz M. Relationship between education and dementia: an updated systematic
review. Alzheimer Dis Assoc Disord. 2011 Oct-Dec 2011;25(4):289-304.
doi:10.1097/WAD.0b013e318211c83c

64. Goldman DP, Zheng Y, Girosi F, et al. The benefits of risk factor prevention in
Americans aged 51 years and older. American Journal of Public Health. Nov 2009;99(11):2096-
101. doi:10.2105/AJPH.2009.172627

65. Goldman DP, Cutler DM, Shang B, Joyce GF. The Value of Elderly Disease Prevention.
Forum Health Econ Policy. 2006;9(2)doi:10.2202/1558-9544.1004

66. Goldman D. The Economic Promise of Delayed Aging. Cold Spring Harbor Perspectives
in Medicine. Feb 2016;6(2)a025072. doi:10.1101/cshperspect.a025072

67. Goldman DP, Cutler D, Rowe JW, et al. Substantial health and economic returns from
delayed aging may warrant a new focus for medical research. Health Affairs. Oct
2013;32(10):1698-705. doi:10.1377/hlthaff.2013.0052

68. Goldman DP, Olshansky SJ. Delayed Aging Versus Delayed Disease: A New Paradigm
for Public Health. Public Policy & Aging Report. 2013;23(4):16-18.

69. Yu J, Tysinger B, Piano Mortari A, et al. The Returns to Preventing Chronic Disease in
Europe and the United States Journal of Human Capital. 2022;16(1)

70. Lakdawalla DN, Goldman DP, Shang B. The health and cost consequences of obesity
among the future elderly. Health Affairs. 2005;24 Suppl 2:W5R30-41.
doi:10.1377/hlthaff.w5.r30

71. Hlávka JP, Tysinger B, Yu JC, Lakdawalla DN. Access to Disease-Modifying
Alzheimer's Therapies: Addressing Possible Challenges Using Innovative Payment Models.
Value Health. Jul 05 2022;doi:10.1016/j.jval.2022.06.003

72. National Academies of Sciences E, and Medicine,. The Growing Gap in Life Expectancy
by Income: Implications for Federal Programs and Policy Responses. 2015:182. Consensus
106
Report. 978-0-309-31707-8. https://www.nap.edu/catalog/19015/the-growing-gap-in-life-
expectancy-by-income-implications-for

73. Auerbach AJ, Charles KK, Coile CC, et al. How the Growing Gap in Life Expectancy
May Affect Retirement Benefits and Reforms. The Geneva Papers on Risk and Insurance.
2017;42(3):475-499. doi:10.1057/s41288-017-0057-0

74. Goldman DP, Orszag PR. The Growing Gap in Life Expectancy: Using the Future
Elderly Model to Estimate Implications for Social Security and Medicare. Am Econ Rev. May
2014;104(5):230-233. doi:10.1257/aer.104.5.230

75. Neumann PJ, Cohen JT, Weinstein MC. Updating cost-effectiveness--the curious
resilience of the $50,000-per-QALY threshold. N Engl J Med. Aug 28 2014;371(9):796-7.
doi:10.1056/NEJMp1405158

76. Leaf DE, Tysinger B, Goldman DP, Lakdawalla DN. Predicting quantity and quality of
life with the Future Elderly Model. Health Econ. Oct 2020;doi:10.1002/hec.4169

77. Zissimopoulos JM, Tysinger BC, St Clair PA, Crimmins EM. The Impact of Changes in
Population Health and Mortality on Future Prevalence of Alzheimer's Disease and Other
Dementias in the United States. J Gerontol B Psychol Sci Soc Sci. 04 16 2018;73(suppl_1):S38-
S47. doi:10.1093/geronb/gbx147

78. Crimmins EM, Kim JK, Langa KM, Weir DR. Assessment of cognition using surveys
and neuropsychological assessment: the Health and Retirement Study and the Aging,
Demographics, and Memory Study. J Gerontol B Psychol Sci Soc Sci. Jul 2011;66 Suppl 1:i162-
71. doi:10.1093/geronb/gbr048

79. Yu JC, Hlávka JP, Joe E, Richmond FJ, Lakdawalla DN. Impact of non-binding FDA
guidances on primary endpoint selection in Alzheimer's disease trials. Alzheimers Dement (N Y).
2022;8(1):e12280. doi:10.1002/trc2.12280

80. Hlávka JP, Yu JC, Lakdawalla DN. Crosswalk between the Mini-Mental State
Examination and the Telephone Interview for Cognitive Status (TICS-27/30/40). Alzheimers
Dement. Feb 01 2022;doi:10.1002/alz.12569

81. Galvin JE, Sadowsky CH, NINCDS-ADRDA. Practical guidelines for the recognition
and diagnosis of dementia. J Am Board Fam Med. 2012 May-Jun 2012;25(3):367-82.
doi:10.3122/jabfm.2012.03.100181

82. Cordell CB, Borson S, Boustani M, et al. Alzheimer's Association recommendations for
operationalizing the detection of cognitive impairment during the Medicare Annual Wellness
Visit in a primary care setting. Alzheimers Dement. Mar 2013;9(2):141-50.
doi:10.1016/j.jalz.2012.09.011

107
83. Jacobson M, Thunell J, Zissimopoulos J. Cognitive Assessment At Medicare's Annual
Wellness Visit In Fee-For-Service And Medicare Advantage Plans. Health Aff (Millwood). 11
2020;39(11):1935-1942. doi:10.1377/hlthaff.2019.01795

84. Wouters OJ, McKee M, Luyten J. Estimated Research and Development Investment
Needed to Bring a New Medicine to Market, 2009-2018. JAMA. Mar 03 2020;323(9):844-853.
doi:10.1001/jama.2020.1166

85. Whittington MD, Campbell JD, Rind D, Fluetsch N, Lin GA, Pearson SD. Cost-
Effectiveness and Value-Based Pricing of Aducanumab for Patients With Early Alzheimer
Disease. Neurology. 03 01 2022;98(9):e968-e977. doi:10.1212/WNL.0000000000013314

86. Ross EL, Weinberg MS, Arnold SE. Cost-effectiveness of Aducanumab and Donanemab
for Early Alzheimer Disease in the US. JAMA Neurol. 05 01 2022;79(5):478-487.
doi:10.1001/jamaneurol.2022.0315

87. Sinha P, Barocas JA. Cost-effectiveness of aducanumab to prevent Alzheimer's disease
progression at current list price. Alzheimers Dement (N Y). 2022;8(1):e12256.
doi:10.1002/trc2.12256

88. Rabinovici GD, Gatsonis C, Apgar C, et al. Association of Amyloid Positron Emission
Tomography With Subsequent Change in Clinical Management Among Medicare Beneficiaries
With Mild Cognitive Impairment or Dementia. JAMA. 04 02 2019;321(13):1286-1294.
doi:10.1001/jama.2019.2000

89. National Council on Disability (NCD). Quality-Adjusted Life Years and the Devaluation
of Life with Disability: Part of the Bioethics and Disability Series. 2019.

90. Heßmann P, Seeberg G, Reese JP, et al. Health-Related Quality of Life in Patients with
Alzheimer's Disease in Different German Health Care Settings. J Alzheimers Dis.
2016;51(2):545-61. doi:10.3233/JAD-150835

91. Eli Lilly and Company. Lilly Reports Robust Third-Quarter 2021 Financial Results as
Pipeline Success Strengthens Future Growth Potential. Updated October 26, 2021. Accessed
January 14, 2022. https://investor.lilly.com/news-releases/news-release-details/lilly-reports-
robust-third-quarter-2021-financial-results

92. Eisai Co. L. Eisai Initiates Rolling Submission to the U.S. FDA for Biologics License
Application of Lecanemab (BAN2401) for Early Alzheimer’s Disease Under the Accelerated
Approval Pathway. Updated September 28, 2021. Accessed January 15, 2022.
https://www.eisai.com/news/2021/news202177.html

93. Cummings J, Lee G, Zhong K, Fonseca J, Taghva K. Alzheimer's disease drug
development pipeline: 2021. Alzheimer's & Dementia: Translational Research & Clinical
Interventions. 2021;7(1):e12179. doi:https://doi.org/10.1002/trc2.12179

108
94. Hlavka JP, Mattke S, Wilks A. The Potential Benefits of Deferred Payment for a
Hypothetical Gene Therapy for Congestive Heart Failure: A Cost-Consequence Analysis. Appl
Health Econ Health Policy. Oct 2020;18(5):669-677. doi:10.1007/s40258-020-00563-y

95. Lin GA WM, Synnott PG, McKenna A, Campbell J, Pearson SD, Rind DM.
Aducanumab for Alzheimer’s Disease: Effectiveness and Value (Final Evidence Report and
Meeting Summary). 2021. August 5. https://icer.org/wp-
content/uploads/2020/10/ICER_ALZ_Final_Report_080521.pdf

96. Herring WL, Gould IG, Fillit H, et al. Predicted Lifetime Health Outcomes for
Aducanumab in Patients with Early Alzheimer's Disease. Neurol Ther. Dec 2021;10(2):919-940.
doi:10.1007/s40120-021-00273-0

97. Goldman DP, Shang B, Bhattacharya J, et al. Consequences of health trends and medical
innovation for the future elderly. Health Aff (Millwood). 2005;24 Suppl 2:W5R5-17.
doi:10.1377/hlthaff.w5.r5

98. Liu-Seifert H, Siemers E, Price K, et al. Cognitive impairment precedes and predicts
functional impairment in mild Alzheimer’s disease. Journal of Alzheimer's Disease.
2015;47(1):205-214.

99. Liu-Seifert H, Siemers E, Sundell K, et al. Cognitive and functional decline and their
relationship in patients with mild Alzheimer's dementia. Journal of Alzheimer's Disease.
2015;43(3):949-955.

100. Goldman DP, Lakdawalla D, Michaud P-C, et al. The future elderly model: Technical
documentation. University of South California. 2015;

101. Grabowski DC, Lakdawalla DN, Goldman DP, et al. The large social value resulting
from use of statins warrants steps to improve adherence and broaden treatment. Health Aff
(Millwood). Oct 2012;31(10):2276-85. doi:10.1377/hlthaff.2011.1120

102. Jena AB, Philipson TJ. Cost-effectiveness analysis and innovation. J Health Econ. Sep
2008;27(5):1224-36. doi:10.1016/j.jhealeco.2008.05.010

103. Hlavka JP, Yu JC, Goldman DP, Lakdawalla DN. The economics of alternative payment
models for pharmaceuticals. Eur J Health Econ. Jun 2021;22(4):559-569. doi:10.1007/s10198-
021-01274-4

104. Hlavka JP, Lavelle TA, Neumann PJ, Lin PJ. Addressing Challenges to Alternative
Payment Models for New Alzheimer's Disease Therapies for US Commercial Payers.
Pharmacoeconomics. May 13 2022;doi:10.1007/s40273-022-01150-w

105. Lam J, Hlavka J, Mattke S. The Potential Emergence of Disease-Modifying Treatments
for Alzheimer Disease: The Role of Primary Care in Managing the Patient Journey. J Am Board
Fam Med. Nov-Dec 2019;32(6):931-940. doi:10.3122/jabfm.2019.06.180328
109

106. Lin P-J, Cohen JT, Neumann PJ. Preparing the health-care system to pay for new
Alzheimer's drugs. Alzheimer's & dementia : the journal of the Alzheimer's Association.
2020;16(11):1568-1570. doi:10.1002/alz.12155

107. Lakdawalla DN, Neumann PJ, Wilensky GR, et al. Health Technology Assessment in the
U.S. – A Vision for the Future. 2021. https://healthpolicy.usc.edu/wp-
content/uploads/2021/02/USC_Schaeffer_FutureHTA_WhitePaper-FNL-Spreads.pdf

108. National Council on Disability. Quality-Adjusted Life Years and the Devaluation of Life
with Disability: Part of the Bioethics and Disability Series 2019.
https://ncd.gov/sites/default/files/NCD_Quality_Adjusted_Life_Report_508.pdf

109. Hlavka JP, Kinoshita AT, Fang S, Hunt A. Clinical Outcome Measure Crosswalks in
Alzheimer's Disease: A Systematic Review. J Alzheimers Dis. 2021;83(2):591-608.
doi:10.3233/JAD-210060

110. Lakdawalla DN, Doshi JA, Garrison LP, Jr., Phelps CE, Basu A, Danzon PM. Defining
Elements of Value in Health Care; A Health Economics Approach: An ISPOR Special Task
Force Report [3]. Value in Health. 2018;21(2):131-139. doi:10.1016/j.jval.2017.12.007

Abstract (if available)

Abstract Alzheimer’s disease (AD) is a degenerative disease that is characterized by loss of cognitive function, functional impairment, and neuropsychological symptoms. It is a highly debilitating disease that results in substantial tolls to patient quality-of-life, and is one of the top leading causes of death in the US. Aside from aducanumab, which was approved by the U.S. Food and Drug Administration (FDA) in June 2021, other potentially disease-modifying therapies (DMTs) for AD are currently under development. These potential DMTs under development target both early stages such as preclinical and prodromal, as well as mild and moderate dementia, and have included patients as young as 50 years old in their pivotal trials.
This dissertation encompasses three aims. These include: (1) investigating the relationship between heart valve disease and the development of Alzheimer’s disease and related dementias (ADRDs), (2) measuring the impact of delayed mild cognitive impairment (MCI) diagnosis and access on the clinical and cost benefits of DMTs – overall and by socioeconomic status, and (3) studying whether traditional upfront payments might limit coverage of DMTs and how alternative payment models could incentivize greater coverage.
We use Medicare administrative claims data to conduct survival analyses in a cohort of severe aortic stenosis patients to understand whether aortic valve replacement surgeries – a surgical treatment for heart valve disease – delay onset of ADRDs. We then utilize the Future Elderly Model (FEM) microsimulation to understand the burden of MCI and how delays in MCI diagnosis and DMT treatment may compress or exacerbate health disparities. Lastly, we use the FEM to understand the net benefits accrued to private and public payers from covering AD DMTs in patients less than 65 years of age, through upfront, constant installment, and performance-based installment payment models.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Disease modifying treatments for Alzeimer's disease: modeling, value assessment and eligible patients

PDF

Long-term blood pressure variability across the clinical and biomarker spectrum of Alzheimer’s disease

PDF

Alzheimer’s disease: dysregulated genes, ethno-racial disparities, and environmental pollution

PDF

Affective neuropsychiatric symptoms and neural connectivity in the early stages of Alzheimer’s disease

PDF

Blood-brain barrier breakdown and vessel integrity in Alzheimer’s disease

PDF

Comparison of participant and study partner predictions of cognitive impairment in the Alzheimer's disease neuroimaging initiative 3 study

PDF

The effects of a fasting mimicking diet (FMD) on mouse models of Alzheimer's and Parkinson's disease

PDF

Memory abnormalities in Alzheimer's disease and anxiety models

PDF

Dihydromyricetin as a potential therapeutic for Alzheimer's disease-like changes induced by social isolation

PDF

Phenotypic and multi-omic characterization of novel C. elegans models of Alzheimer's disease

PDF

Sex differences in the TgF344-AD rat model: Investigating the behavior, pathology, and neuroanatomical structures

PDF

Role of CD33 structure and function in Alzheimer’s disease

PDF

Alzheimer’s disease and related dementias research communication practices with minoritized racial and ethnic groups: a multiple case study

PDF

Exploring the relationship between menopausal hot flushes and Alzheimer's disease biomarkers: a cross-sectional analysis in postmenopausal women

PDF

An overview of Alzheimer’s disease and the potential for APP-beta secretase-interaction as a therapeutic target

PDF

Insulin sensitivity in cognition, Alzheimer's disease and brain aging

PDF

Adoption and implementation of innovative diagnostic tools for Alzheimer's Disease: challenges and barriers in primary care

PDF

Essays on access to prescription drugs

PDF

Neuroimaging markers of risk & resilience to brain aging and dementia

PDF

Neuroinflammation and ApoE4 genotype in at-risk female aging: implications for Alzheimer's disease

Asset Metadata

Creator Yu, Jeffrey (author)

Core Title Alzheimer's disease: risk factors, value, and alternative payment models

School School of Pharmacy

Degree Doctor of Philosophy

Degree Program Health Economics

Degree Conferral Date 2022-12

Publication Date 11/02/2022

Defense Date 08/15/2022

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag alternative payment models,Alzheimer's disease,dementia,Diagnosis,disease modifying therapies,heart valve disease,Medicare administrative claims,microsimulation,mild cognitive impairment,OAI-PMH Harvest,socioeconomic status

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Lakdawalla, Darius (committee chair), Trish, Erin (committee member), Tysinger, Bryan (committee member)

Creator Email jeffrecy@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC112212329

Unique identifier UC112212329

Identifier etd-YuJeffrey-11293.pdf (filename)

Legacy Identifier etd-YuJeffrey-11293

Document Type Dissertation

Format theses (aat)

Rights Yu, Jeffrey

Internet Media Type application/pdf

Type texts

Source 20221103-usctheses-batch-989 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright. The original signature page accompanying the original submission of the work to the USC Libraries is retained by the USC Libraries and a copy of it may be obtained by authorized requesters contacting the repository e-mail address given.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu