Analysis Group, Inc.
Health Economics, Outcomes Research,
and Epidemiology Practice Areas
September 13, 2012
BOSTON
CHICAGO
DALLAS
DENVER
LOS ANGELES
MENLO PARK
MONTREAL
NEW YORK
SAN FRANCISCO
WASHINGTON
A Snapshot of Our Firm
 Analysis Group is the largest privately held economic
consulting firm in North America
 Our clients include major law firms, Fortune 100 companies,
and government agencies
 We have been growing steadily since our launch over 30
years ago and now have 10 offices in the U.S. and Canada
 Our firm has over 500 professionals, many with graduate
degrees in economics, finance, accounting, management, or
law
 We have been ranked as one of the top ten consulting firms
by Vault for the past three years
 We have been named one of the ‘Top Places to Work’ in
Massachusetts by The Boston Globe for the past four years
2
Our Professional Team
Interdisciplinary team with expertise in:
 Health economics
 General economics
 Econometrics
 Biostatistics
 Epidemiology
 Pharmacy
 Medicine
 Finance
 Business Administration
 Accounting
Over 120 health care specialists with advanced degrees: PhD, PharmD,
MD, and ScD
3
Our Clients
 Pharmaceutical manufacturers
 Medical device manufacturers
 Drug delivery companies
 Biotechnology companies
 Payers (insurers, employers)
 General and specialty hospitals, integrated delivery networks, joint ventures,
and physician practices
 Equipment and medical product companies
 State and federal government
4
Our Integrated Approach to Health Care
BUSINESS DIRECTION





Therapeutic area selection and
strategy
Product portfolio optimization
Partnering, licensing,
outsourcing, acquisitions
Organizational structuring to
achieve growth
Technology and
commercialization assessment
COMMERCIAL RESULTS



Brand strategy and
positioning
Market analysis and
segmentation
Global pricing and
reimbursement

Managed care contracting

Competitive assessment

SCIENTIFIC OUTCOMES






R&D health care economics
Health outcomes and costeffectiveness
Econometrics
Public/regulatory hearings
Epidemiology and
biostatistics
Clinical trials optimization
LITIGATION






Intellectual property
Antitrust
Product fraud
Off-label investigations
General commercial
business litigation
Contract disputes
Lifecycle and patent expiry
management strategy
5
Selected AG Health Care Studies in the News
Votrient
February 22, 2011
AG epidemiologists and Harvard affiliates used novel
statistical methods to assess survival data for economic
models demonstrating Votrient’s cost-effectiveness.
NICE issued a positive Final Appraisal Determination for
Votrient as a first-line treatment for advanced or
metastatic renal cell carcinoma.
Celexa
May 2, 2007
When the maker of the blockbuster antidepressant
Celexa was threatened with safety concerns and
additional black box warnings by the U.S. FDA, AG
epidemiologists helped demonstrate that the drug was
no less safe than class competitors and averted
differential labeling.
Humira
June 2008
AG pharmacoeconomists provided the maker of Humira with pivotal support in modeling and developing health
economic submission packages for commercial and national payers. NICE has recommended Humira and not
competing TNFs in certain indications.
6
Our Research Generates Timely and Impactful Scientific
Evidence
Gaps in
Scientific
Research
 Innovative research
 Cutting-edge health data
 World-class analytics
 Academic thought leaders
Client
Needs
Decision
Maker
Needs
 Publications
7
Analysis Group Assists Health Care Clients with HEOR
Challenges Over the Product Life Cycle
Discovery and
Development
Product
Approval and
Launch
Post-Launch
Optimization
Patent Expiry
and Late Life
Cycle

Early stage modeling and GAP analysis

Personalized medicine analysis to help refocus clinical trial

Increase awareness of diseases (e.g., prevalence, burden of illness)

Identify, design, and validate appropriate PRO instrument

Treatment patterns of conventional therapy (adherence, dose escalation)

Drug safety

Limitations of standard care and competitors (unmet needs)

Demonstrate product value relative to conventional therapy

Cost of illness studies

HTA/reimbursement submission: CEA, BIA, Dossier

Treatment pattern and adherence

Support market expansion (burden of under-treatment, delayed treatment and
delayed diagnosis)

Identify the patient subgroup populations who most likely to benefit from a given
treatment
8
Examples of Our Work:
Matching-Adjusted Indirect Comparisons
PAGE 9
Indirect Comparisons in Comparative Effectiveness
Research
 Health care decision makers face choices among a growing number of
alternative treatments
 Policy makers recognize the importance of strengthening the evidence base
for medical decisions (e.g., renewed U.S. public investment in CER)
 Comparative evidence is valuable, but difficult to obtain
 Head-to-head randomized trials provide the gold standard, but are not always
available, especially for new drugs
 A costly gap can result if treatment decisions and product strategies are
developed without accounting for true product superiorities
Indirect comparisons help to fill such gaps
10
Goals of Indirect Comparison
 Make a fair and credible comparison
 Need to adjust for differences between trials
 Provide timely results
 Need to inform decisions as they are made; can’t wait for more
data
 Draw robust conclusions
 Use all available clinical trial data
 Appreciated by payers
11
Step 1: Study and Sample Selection
Inclusion/exclusion criteria for comparator trial need to be
equivalent to or nested within the inclusion criteria for the IPD trial
B vs. placebo (published trial)
Increasing Age 
Increasing Age 
A vs. placebo (IPD available)
Increasing Disease Severity 
Increasing Disease Severity 
12
Step 1: Study and Sample Selection
Inclusion/exclusion criteria for the published comparator trial can
then be imposed on the IPD to create comparable populations
B vs. placebo (published trial)
Increasing Age 
Increasing Age 
A vs. placebo (IPD available)
Increasing Disease Severity 
Increasing Disease Severity 
13
Step 2: Reweighting Patients
Using robust statistical methodology, patients in Trial A can be reweighted to match the baseline characteristics of those in Trial B
Re-weighted Trial
of A vs. placebo
Trial of A vs. placebo
70
70
50
60
60
50
50
40
40
%
30
20
%
60
40
%
Trial of B vs. placebo
30
30
20
10
20
10
10
0
0
Males
Females
0
Males
Females
Males
Females
Increase weight for
females relative to males
14
Results: Comparison of Outcomes between Matched
Trials
Mean Improvement in Symptoms
P < 0.001
Our drug’s efficacy after matching
Competitor drug’s published efficacy
Balanced placebo-arm efficacy
from our trial (after matching) and
the competitor’s trial (as published)
15
Examples of Our Work:
Personalized Medicine
PAGE 16
Personalized Medicine
Comparative Effectiveness Research (CER): “What treatment works
best for which patient population?”
Traditional CER: Given a patient population, find the treatment that works
best
Individualized CER: Given a treatment, find the patient population for
whom it works best
17
Identification of Optimal Patient Subpopulations
Any subpopulation involves a tradeoff between two goals
Maximize value proposition
Payers want to get the most
value while limiting costs
Maximize market size
We want to access the largest
market consistent with labeling
 Payers do not systematically consider this tradeoff when designing access
restrictions
 The optimal subpopulation is the largest subpopulation in which the value
proposition is acceptable to payers
How can we identify this optimal subpopulation?
18
Identification of Optimal Patient Subpopulations
In the full trial population, 21% more patients achieve response with
active drug vs. placebo
100
90
Efficacy
80
70
60
50
40
30
20
All Patients
10
0
20
40
60
80
100
Subpopulation Size
Hypothetical data
19
Identification of Optimal Patient Subpopulations
Subpopulations defined by specific patient characteristics may show
greater efficacy but may have limited market potential
100
90
Efficacy
80
70
60
Age < 50
50
≥ 2 prior therapies
40
30
Weight < 100 kg
20
All Patients
10
0
20
40
60
80
100
Subpopulation Size
Hypothetical data
20
Identification of Optimal Patient Subpopulations
By combining multiple patient characteristics, the personalized
medicine approach identifies the largest market size at any level of
efficiency (the efficiency frontier)
100
Efficiency Frontier
90
Efficacy
80
70
60
Age < 50
50
≥ 2 prior therapies
40
30
Weight < 100 kg
20
All Patients
10
0
20
40
60
Subpopulation Size
80
100
Hypothetical data
21
Examples of Our Work:
Drug Adherence
PAGE 22
Impact of Non-Adherence to Anti-Epileptic Drugs on
Health Care Utilization Resources and Mortality
3.5
3.32
3
2
1.5
Hazard Ratio
Incidence Rate Ratio
2.5
1.76
1.39
1.19
0.93
1
0.5
0
Hospitalizations
Inpatient Days
Emergency
Room Visits
Outpatient Visits
Mortality
Source: Duh MS, et al., ISPOR 2008; Duh MS, et al., ISPE 2008.
23
Examples of Our Work:
Cost of Illness Studies
PAGE 24
Annual Costs per ADHD Patient and Non-ADHD Family
Member
ADHD Patients
Cost per Year
$3,000
$2,495
$2,000
$1,000
$0
Non-ADHD Family Members
$1,574
$1,289
$541
Control Patients
Hospital Outpatient
ADHD Patients
Prescription Drug
Family Members in
Control Family
Sample
Hospital Inpatient
Provider's Office
Non-ADHD Family
Members in ADHD
Family Sample
Work Loss
Other
Source: Swensen A, Birnbaum H, et al., J Am Acad Child Adoles Psychiatry 2003; 42(12) 1415-1423.
25
Your Career at Analysis Group
PAGE 26
Your Career Path - Positions
Analyst
Senior
Analyst
Associate
Manager
Vice
President
Managing
Principal
Career advancement is based on individual contributions in distinct areas:
Career progression reflects:
 Casework
 Project Management
 Business Development
 Teamwork
 Overall contribution to the firm
27
What Can You Expect as an Associate at Analysis
Group?
 Ability to work across practice areas
 Exposure to different industries and areas of expertise
 A supportive environment where you will grow
professionally:
 Advisors and peer-mentors
 Flexible case assignments
 Collaborative/open-door policy
 Lack of hierarchy
 Highly motivated, highly skilled colleagues
 Work closely with managing principals and academic
affiliates; interact directly with clients
28
An Ideal Candidate
 Advanced degree in quantitative sciences, such as
health economics, biostatistics, econometrics,
statistics, epidemiology, psychometrics
 Outstanding track record of applying quantitative
methods to real-world research problems, preferably
in health care research
 Proficiency in at least one statistical programming
language (e.g., SAS, R, S-PLUS)
 Excellent organizational and communication skills
 Comfortable interacting with clients and key opinion
leaders
 Ability to work independently and with a team
29
Your Interview:
What We Want to Know About You
 Interest in health economics, outcomes research, and
epidemiology
 Your academic experience and career goals
 Industry experience or specialized expertise
 Data-driven papers or projects you’ve worked on
 Analytical and technical skills; conceptual capabilities
 Communication skills and teamwork experiences
 Displays of leadership
30
Application Information
Analysis Group, Inc. is currently conducting a resume drop for Johns
Hopkins Bloomberg School of Public Health students.
Resume Submission Deadline:
 Monday, October 15, 2012
Please submit your resume, unofficial transcript and a cover letter
indicating geographic preference(s) through the Career Services office
and the Analysis Group, Inc. website:
www.analysisgroup.com/open_positions. In order to be considered,
you MUST list Career Services/Job Posting in the source field.
For more information about Analysis Group, please visit our website at
www.analysisgroup.com
31