PHARMACOECONOMICS 749

A Matrix to Determine Health Economic

Michiel Hemels, Drs,
MSc, CBA
Global Health Economics
and Outcomes Research
Department, Novo Nordisk
A/S, Copenhagen, Denmark
Michael Wolden, MSc
Global Health Economics
and Outcomes Research
Department, Novo Nordisk
A/S, Copenhagen, Denmark
Thomas R. Einarson, PhD
Leslie Dan Faculty of
Pharmacy, University of
Toronto, Toronto, Canada
Viability Throughout Product Development:
A Pharmaceutical Industry Perspective
Developing drugs is a risky process. This article
presents a conceptual model that guides in iden-
tifying the economic viability of a compound
throughout its development. Examples are
based on the class of antidepressants (ATC
Class N06). The characteristics of a compound
in development determine the data requirements
(clinical, economic, and quality-of-life data) to
ensure maximum support for potential reim-
bursement. The matrix is formed by two drug
characteristics of primary importance: indica-
tion and mechanism of action. Four scenarios
arise from this matrix: new market entry, prod-
uct development, market expansion, and market
penetration. Economic viability incorporates
the following parameters: efficacy, safety, pa-
tient-reported outcomes, pricing, effectiveness,
and formulation. Consequently, development
can be made efficient, with reduced waste of re-
sources and funds. The model should ultimately
aid in determining how to optimize drug viabili-
ty and help product teams choose the optimal
health economic strategy subject to the market
scenario.

Key Words
Health economic; Economic;
Development; Matrix; INTRODUCTION
Viability There has been a drastic change in the business
Correspondence Address model of pharmaceutical companies. After an
Dr. Michiel Hemels, Novo
Nordisk A/S, Novo Allé, DK- era of immense growth in revenues and prof-
2880 Bagsværd, Denmark itability, creation of hundreds of new medica-
(email:
mhnh@novonordisk.com). tions, and targeting physicians as key decision
makers, pharmaceutical companies are experi-
encing demands for cost containment and risk
management. The industry is furthermore char-
acterized by outsourcing, mergers and acquisi-
tions (with layoffs becoming commonplace),
and an intensive focus on the payer as key deci-
sion maker.
Health economics involve the application of
the principles of economics to health-related
products, programs, and services (1). When ap-
plied to drugs, it is referred to as pharmacoeco-
nomics (2). Awareness of the discipline of health
economics within pharmaceutical companies
has greatly improved recently (3,4). As well, its
value in communicating to external decision
makers (eg, payers, policymakers, patients) the
value of their products and gaining market ac-
cess or reimbursement approval has been well
recognized (5–8). Most of the applications of
health economics, however, have involved the
production of these analyses late in the product
cycle and almost all of them have been aimed
largely at marketing newly launched products.
It has been suggested that health economics
should also be positioned earlier in the develop-
ment process where its role can become more
influential (9–13). In fact, both Sculpher et al.
(14) and Grabowski (12) recommended that
health economic analyses be an iterative
process that starts in phase 1 and continues to
phase 4, producing firmer estimates along the
way. The purpose of health economics, states
Miller (13), is “to generate information that will
influence customer groups, gain regulatory and
reimbursement approval, and lead to enhanced
commercial success.” Ultimately, companies
seek to maximize their return on investment. At
the same time, an undeveloped area in this field
exists in product development. Clemens et al.
(11) suggested that health economic research
be expanded to “include assessing the implica-
tions of projected outcomes and costs of phar-
maceutical products for the decision whether to
continue or stop development of a drug, and for
global pricing strategies.” Miller (13) pointed
out that, based on value for money, customers
generally have a preference for cost-effective
products.
According to Miller (13), health economics

Drug Information Journal, Vol. 43, pp. 749–756, 2009 • 0092-8615/2009 Submitted for publication: October 8, 2008
Printed in the USA. All rights reserved. Copyright © 2009 Drug Information Association, Inc. Accepted for publication: February 5, 2009
750 PHARMACOECONOMICS
can assist in determining what information is
collected and when. It informs decisions of
product selection and development as well as
termination and it helps determine optimal
pricing strategy. In applying health economics
during development, there is a need to address
both technical and allocative efficiency. Alloca-
tive efficiency refers to maximizing output by se-
lecting the best choices from among the invest-
ment options, that is, selecting the best drug or
aspect of a drug to develop. Technical efficiency
refers to identifying the best way to achieve a
goal once it has been selected (ie, needs with re-
gards to type of data, appropriate analyses, com-
parators, sample sizes, etc).
As noted above, health economics analyses
have traditionally been done late in the product
cycle. It appears that there are distinct advan-
tages to introducing them into all phases of
product development. This admonition is espe-
cially true when viewed within the context of the
model presented below.
In this article, we introduce the concept of
health economic viability, which refers to an as-
sessment of the reimbursement potential of a
compound in development from a health eco-
nomic point of view. To assess health economic
viability, a 2 × 2 matrix is presented as a guide
throughout product development (ie, preclini-
cal and clinical) for optimal positioning for mar-
ket access and pricing strategies. The matrix
provides direction and strategy that can be de-
termined early in the process and thereby pro-
duce efficiencies that reduce costs and increase
revenues.
H E A LT H E C O N O M I C V I A B I L I T Y
ASSESSMENT
MATRIX MODEL
To gain competitive advantage, companies need
to establish the health economic viability of the
product and adapt development plans effective-
ly to meet market access requirements. As stated
earlier, health economic viability is based on the
opportunity for a compound to offer value for
money relative to the existing and entering
compounds (ie, between the product’s go deci-
sion and its launch). To describe health eco-
Hemels, Wolden, Einarson
nomic viability, value for money is expressed by
using the following six parameters: efficacy, safe-
ty, patient reported outcomes (PROs), pricing,
effectiveness, and formulation (Table 1). In or-
der to establish the product’s health economic
viability, a product team should evaluate, based
on type of scenario as well as these six parame-
ters, the health economic challenges ahead that
must be overcome to achieve reimbursement
success. Table 1 highlights the fact that market
competition increases the emphasis on these
parameters due to increased pressure on the
compound and company to differentiate them
from the current market situation. Please note
that the emphasis gradations are arbitrarily
graded to demonstrate the associations between
these parameters as well as across the market
scenarios.
The matrix model presented here (Figure 1) is
designed to aid in determining relatively easily
the health economic viability of a product. In
the early stage of development (phase 0), the
model can be used to create a framework for re-
quirements and benchmarking. Following this,
it may be applied to any stage of development to
assess a compound’s value using the six parame-
ters, as well as identifying potential strengths or
pitfalls that could guide the decision whether to
discontinue the development of a product in
the early stages of development as its reimburse-
ment potential becomes unlikely.
The model uses the concept of pharmaceuti-
cal thrust to categorize potential new market
entrants. It identifies which product is to be
used in which indication and what unique ad-
vantages it possesses. Pharmaceutical thrust is
based on the Anatomic, Therapeutic, Chemical
(ATC) System of the World Health Organization
(15). In our model, Anatomic and Therapeutic
classifications identify the indication (which
can be either novel or existing) and Chemical
identifies the mechanism of action, which also
can be either novel or existing. Based on
whether an entity has a novel or already existing
indication and mechanism of action in the cur-
rent and future market, four possible scenarios
of health economic viability are presented (Fig-
ure 1). To explain the functioning of the matrix,
A Matrix to Determine Health Economic Viability PHARMACOECONOMICS 751

Product Differentiators and Their Emphases Within Each Market Scenario

TABLE 1
Scenario
I II III IV
Differentiators Enter New Market Product Development Market Expansion Market Penetration
Efficacy ++ ++/+++ ++ +++
Safety ++ ++/+++ + +++
Patient reported outcome + ++ + ++
Pricing + +/++ + ++
Formulation NA +/++ + ++
Effectiveness NA +/++ +/++ ++
NA, not applicable. Emphasis gradations are arbitrarily graded: + for low; ++ for medium, +++ for high.

INDICATION
FIGURE 1
NOVEL EXISTING

Matrix for determining a

Scenario I: Enter new market Scenario II: Product development new drug’s health eco-
nomic viability.
Focus: Drug should be safe and Focus: Safety, efficacy and
effective cost-effectiveness

- First in class/indication - Partial me too

- Demonstrate efficacy, safety - Improve product on at least 1 key
NOVEL

- Free pricing differentiator

- Establish cost/QALY - Demonstrate cost-effectiveness
benchmark - Measure patient reported outcomes

Example: fluoxetine Example: escitalopram (S-enantiomer

of citalopram)
MECHANISM OF ACTION

Scenario III: Market expansion Scenario IV: Market penetration

Focus: Safety, efficacy and Strong focus: Efficacy/effectiveness,

cost-effectiveness safety, patients’ quality of life, price
compared to existing treatments
- Increase label
- Demonstrate efficacy - Me too
EXISTING

- Fixed pricing: risk price reduction - efficacy and safety comparable to

- Establish cost/QALY benchmark
- Interest in effectiveness
Example: duloxetine (in treatment
of depression, stress urinary
incontince drug, pain etc.),
bupropion (smoking cessation)
existing products
- Improve product on at least 2 key
differentiators
- Demostrate cost-effectiveness
- Measure patient reported outcomes
Example: any drug second in class
(eg, citalopram)

QALY: Quality Adjusted Life Year

Drug Information Journal

752 PHARMACOECONOMICS
N06A ANTIDEPRESSANTS
FIGURE 2
* N06AA Nonselective monoamine reuptake inhibitors
Antidepressants ATC class
- N06AA01 Desipramine
N06.
* N06AB Selective serotonin reuptake inhibitors
- N06AB03 Fluoxetine
- N06AB04 Citalopram
- N06AB10 Escitalopram
* N06AF Monoamine oxidase inhibitors, nonselective
- N06AF03 Phenelzine
*N06AG Monoamine oxidase A inhibitors
- N06AG02 Moclobemide
* N06AX Other antidepressants
- N06AX05 Trazodone
- N06AX16 Venlafaxine
- N06AX21 Duloxetine
- N06AX23 Desvenlafaxine
examples are based on the class of antidepres-
sants (ATC Class N06) as depicted in Figure 2
(15).
MODEL STRUCTURE
In the upper left quadrant of Figure 1, the phar-
maceutical thrust is a new market entry (sce-
nario I) based on having a new indication as well
as a novel mechanism of action. This scenario is,
unfortunately, becoming less common (4). It re-
quires that the drug have demonstrated safety
and efficacy to support its approval (eg, EMEA,
FDA, etc), but health economic requirements at
this stage are minimal. Such drugs are referred
to as first in class, a very desirable designation,
as this status conveys a certain degree of free-
dom in (premium) pricing.
In this scenario, quality of life (QOL) should
be measured throughout a product’s develop-
ment to establish a benchmark for the health
states involved and to allow for across-the-board
comparisons based on cost/quality-adjusted life
year (QALY) gained. The formulation of the drug
(eg, oral, intravenous, subcutaneous, etc) has a
relatively low impact on health economic viabil-
Hemels, Wolden, Einarson
ity, but could pose challenges when similar
products enter the market with the same specif-
ic indication. At that time, effectiveness data
may also become more relevant.
Taking the class of antidepressants as an ex-
ample, five agents were first in class, being
desipramine (N06AA01) in 1963, phenelzine
(N06AF03) in 1981, moclobemide (N06AG02)
in 1993, trazodone (N06AX05) in 1985, and flu-
oxetine, better known as Prozac (N06AB03), in
1990.
When products fall into the upper right quad-
rant, the pharmaceutical thrust is product de-
velopment (scenario II), meaning that the drug
is purported to have a different (and preferably
unique) mechanism of action from an existing
indication (ie, a partial “me-too”). Consequent-
ly, all six parameters need to be taken into ac-
count (Table 1). In most cases, parameters such
as efficacy and safety need to be at least compa-
rable (unless one greatly offsets the other, re-
sulting in a very efficacious but more toxic, or
less efficacious and less toxic drug). It has been
recommended that one of the parameters of
QOL, that is, the cost per QALY should not ex-
ceed £30,000, or the formulation should have
some clear clinical advantage for full market
penetration (16). As there is competition within
this scenario, PRO assessments, appropriate for
the product’s indication, are possible differen-
tiators.
There are two clear examples of products in
the class of antidepressants, being escitalo-
pram (N06AB10), the S-stereoisomer (enan-
tiomer) of the earlier Lundbeck drug citalopram
(N06AB04), and desvenlafaxine (N06AX23),
the synthetic form of a known active metabo-
lite of the earlier Wyeth drug venlafaxine
(N06AX16).
In the lower left quadrant, the pharmaceutical
thrust is market expansion (scenario III), also
called label expansion, which is similar to the
product development scenario. However, next to
parameters such as efficacy and safety that must
be comparable, the pricing will be determined
based on its previous indication. This limitation
may have drastic implications for the product’s
price and it is therefore of great importance to
A Matrix to Determine Health Economic Viability
determine the sequence of launch and to
change the development program accordingly.
This activity may sound easier than it is in reali-
ty, as the clinical development program is usual-
ly based on likelihood of success combined with
the duration of the development for a particular
indication. Different formulations for different
indications could result in pricing exemptions,
although support for these decisions should be
well documented and effectively presented to
authorities. For this scenario, QOL should be
measured throughout its development to estab-
lish a benchmark and, allowing for across-the-
board comparisons on cost/QALY gained and
depending on its use in the real-life setting, in-
terest in effectiveness data may exist.
Several antidepressants have obtained multi-
ple indications. Among the most recent prod-
ucts is duloxetine (N06AX21), manufactured by
Eli Lilly, used for major depressive disorder, gen-
eralized anxiety disorder, pain related to diabet-
ic neuropathy and fibromyalgia, and in some
countries for stress urinary incontinence.
Finally, in the lower right quadrant, the phar-
maceutical thrust is market penetration (sce-
nario IV). It is probably among the most likely
scenarios in current product development, but
poses the greatest challenges for health eco-
nomic viability. The product is considered to be
a me-too, and therefore there is an increased fo-
cus on added value relative to existing products.
Due to the strong competition in this scenario,
next to superior efficacy and safety (or at least
one of the two), it is recommended that the
product should possess another benefit (eg,
PRO) for differentiation. Cost-effectiveness data
are warranted and, at a later stage, effectiveness
data should support the evidence derived
from efficacy studies and economic modeling.
Clemens et al. (11) recommended that compa-
nies avoid producing “economic me-toos,” as
they will not fill an unmet need. Such drugs are
in the same class but offer no distinct advantage
over existing products. Others have considered
that economic me-toos fall into three cate-
gories: efficacy and cost equal to existing drugs,
but with a different (ie, worse) adverse effect
profile; additional efficacy but at additional cost
Drug Information Journal
PHARMACOECONOMICS 753
not acceptable to formulary decision makers;
and much better efficacy but not cost effective
due to excessively high acquisition cost.
For the antidepressants, these products are
any product after the first in class, meaning all
products in, for example, ATC class N06AA,
N06AF, N06AG, N06AX, or N06AB.
At all stages of development, one should be
aware of new entities entering the market and
assess the implications of these agents relative
to the product in development.
DATA REQUIREMENTS
In order to truly assess the business opportuni-
ty, all members within a product team of a phar-
maceutical company need to continuously as-
sess the true value of their product throughout
its development. Within these product teams,
the temptation to construct an unduly opti-
mistic case has to be avoided at all times. In the
ideal situation, the medical team, which is re-
sponsible for the clinical development of the
compound, reports timely updates on clinical
trial progress with its clinical results bench-
marked with competitor data. Marketing per-
sonnel must communicate clearly the product
profile targets for clinical safety and efficacy,
thereby setting the scenario for clinical devel-
opment related to returns on investment.
Health economists challenge the health eco-
nomic viability by communicating the needs
and requirements for global as well as national
reimbursement. The following paragraph briefly
describes the health economic processes relat-
ed to the six parameters (ie, efficacy, safety, PRO,
pricing, effectiveness, and formulation) that
should be addressed in these early stages of de-
velopment.
Efficacy (in relation to safety) is the key pa-
rameter in reimbursement decisions. Although
its responsibility lies in medical teams, it is the
health economist’s role to ensure that the clini-
cal trial designs (eg, clinical outcomes, assess-
ment scales, time of measurements, etc) meet
the criteria for reimbursement set by authori-
ties. In this case, “demonstrate efficacy” would
refer to the case where the drug is comparable
to existing competing products, while “high effi-
754 PHARMACOECONOMICS
cacy and safety” would refer to the case where
the new product has a clear advantage over ex-
isting choices. In trials of a drug having little
chance of benefit and a high risk of toxicity, the
QOL of the patient in the final stages of the dis-
ease may be of importance to the reimburse-
ment authorities. In most cases, QOL is used for
benchmarking (eg, cost/QALY) or for competi-
tor differentiation. QOL assessments, however,
could also be extremely valuable in assessing pa-
tients’ attitude and its relative value toward po-
tential reimbursement for debilitating side ef-
fects (eg, hand lesions associated with some
cancer treatments). In order to use QOL claims
in their submissions, pharmaceutical compa-
nies have been consulting authorities such as
the FDA to incorporate the agency’s regulatory
approach to these claims. Within the company,
it is advised to discuss early in the product’s de-
velopment which PROs (eg, QOL assessment
scales, resource utilization, etc) are the most rel-
evant for inclusion in phase 2 or phase 3 trials.
Through early involvement of the product team,
an infrastructure (ie, budget allocation, ade-
quate sample size in trial, acceptable patient
and investigator burden) is created.
Pricing studies investigating the market, bar-
riers to entry, and product differentiation can
be conducted early in the development process.
However, once more information on the prod-
uct profile has become available, more relevant
questions can be asked and more specific infor-
mation related to the product could be gathered
to answer those questions.
Effectiveness studies evaluate treatments in
the settings where they will be applied. They
could be carried out through database analyses,
observational studies, or effectiveness trials.
Although associated with methodological chal-
lenges (eg, loss to follow-up, uncontrolled influ-
ences and confounders, etc), statistical adjust-
ments (eg, allowances for dropouts, missing data,
etc) can be made. Prelaunch, effectiveness stud-
ies could very well be used to define the unmet
clinical need. Postlaunch, these studies could be
used for price negotiations.
Drug formulation is becoming increasingly
relevant for positioning new products. Formula-
Hemels, Wolden, Einarson
tion has been associated with adherence. It can
also play a major role in product pricing, espe-
cially in the market expansion scenario. Im-
plications of formulation affect the product
launch strategy, particularly if the product price
is impacted. Early in development plans, careful
analysis of the possible impact of the drug’s for-
mulation on price related to the sequence of its
anticipated launch should be carried out.
DISCUSSION
ECONOMIC VIABILITY
There is an increased incentive to determine
economic viability. Development costs have
been steadily rising since the 1960s (17). In
1991, DiMasi et al. (18) estimated that it cost
$231 million (in 1987 USD) to develop a drug.
That value rose to $403 million (in 2000 USD)
in their 2003 publication (17). Although indus-
try numbers have been challenged (19–21), the
costs remain nonetheless very high. A com-
pounding problem is the fact that the costs of
drug development increase for each phase of
the cycle. One reason is that each phase re-
quires studying an increasing number of sub-
jects. DiMasi et al. (18) estimated that the costs
for phase 2 were double those of phase 1 (at the
time, $2 million in 1987 USD) and phase 3 costs
were six times those of phase 1. Using the same
approach in 2003, DiMasi et al. (17) found the
same cost ratios, but absolute costs had in-
creased by a factor of 7, with the cost of phase 1
at that time over $15 million (measured in 2000
USD). Grabowski (12) noted that the majority of
costs arose in later stages of the product cycle
and that go/no-go decisions were made just pri-
or to phase 3.
In 1997, the suggestion was made that meth-
ods such as health economics could be intro-
duced earlier to help make those decisions. Fur-
thermore, Grabowski (12) indicated that the
objective of early health economics is to identi-
fy drugs that will not become successes and
cause them to fail sooner than if they were al-
lowed to progress and fail later. In that way, re-
sources could be more efficiently deployed in
developing drugs with higher potential to suc-
ceed and generate revenues. Since the majority
A Matrix to Determine Health Economic Viability
of costs are incurred in later stages of develop-
ment, it would be feasible to use these markers
of economic viability early in the cycle and
thereby avoid huge capital losses. The saved re-
sources could then be deployed more efficiently
in developing more viable products.
MATRIX MODEL
The model presented here serves as a conceptu-
al framework and its purpose is to guide employ-
ees in pharmaceutical organizations to deter-
mine the health economic viability of their
products throughout all stages of development.
The factors presented in the model are recom-
mendations for which factors to focus on and
should be linked to Table 1. The aim of the mod-
el is not to compare the outcomes, as the drug
maker will face different scenarios depending
on the drug type, such as first in class versus me-
too drug. The aim of the model is to help identi-
fy and put emphasis on important health eco-
nomic factors subject to the market scenario
that will aid in increasing the drug’s viability.
MODEL LIMITATIONS
As with all models, there are some limitations.
For example, the examples provided in the ma-
trix may not match for all drugs in development.
Furthermore, caution should be taken when us-
ing this model that the point of reference (ie,
standard of care) may change over time, espe-
cially in indications where several pharmaceuti-
cal companies, often due to high unmet medical
needs, are developing a new compound for the
same indication. Ideally, a cost-effectiveness
model would be used to identify possible chal-
lenges related to future market access of a drug.
A rough estimation of drug efficacy can be ob-
tained through modeling incorporating all as-
pects of the experimental design, from its begin-
ning to its completion, allowing estimation of
the treatment outcomes (eg, safety and efficacy)
with sufficient statistical power, or at least that
is what is assumed (22). However, with in-licens-
ing, due to market circumstances, timelines for
decision making are often relatively short. Con-
sequently, cost-effectiveness modeling could
only be done when sufficient data as well as an
Drug Information Journal
PHARMACOECONOMICS 755
existing (eg, plug and play) decision model are
readily available, like the CORE diabetes model
(23). When constructing these models, it is es-
sential to also take into account products that
are under development or are in the process of
launching. If the drug development is undertak-
en in-house, there is an opportunity for design-
ing a decision model de novo. Furthermore, one
would expect more data availability within the
company, as the choice for development had
been made at an earlier stage.
Vernon (16) described the mathematical ap-
proach to cost-effectiveness modeling for
go/no-go decisions. In that article, they used a
hypothetical example to demonstrate health
economic viability of a product. Their model in-
corporated both deterministic and stochastic
assumptions using both analytic and simulation
methods. However, the methodology is quite
complex and requires making many assump-
tions. It also requires knowing the payer’s will-
ingness to pay for a product at various different
levels of efficacy, which is far from certain at
present.
In order to optimize product development,
one could suggest that within product teams of
pharmaceutical companies, members from vari-
ous disciplines (eg, regulatory affairs, product
development) should establish and present sim-
ilar viability assessment models. This way, an
ever more thorough understanding of total
product viability could be achieved.
CONCLUSION
Product managers and their development teams
need to identify early the properties of the
products under development and classify them
into one of the four scenarios. If a product en-
ters a new market, the highest probability for
success will be obtained by focusing mainly on
its efficacy. The unmet medical need for the
product in this indication will balance most of
the safety concerns. Product development and
market expansion both have similar challenges
in health economic viability; however, the em-
phasis should be directed toward not only effi-
cacy and safety, but also cost effectiveness.
Compounds that fall into the market penetra-
756 PHARMACOECONOMICS
tion category have a strong focus on all parame-
ters due to requirements for product differenti-
ation, particularly in demonstrating value for
money. It could be useful to consider changing
product development into a niche strategy, tar-
geting a more limited population, however, at a
price that could ensure a reasonable return on
investment. This model should assist health
economists and product teams in making the
drug development process more focussed, effi-
cient, and successful.
REFERENCES
1. Drummond M, Mooney G. Essentials of health
economics: part I (continued). What is econom-
ics? Br Med J (Clin Res Ed). 1982;285:1024–
1025.
2. McGhan W. Pharmacoeconomic series: part I.
Pharmacoeconomics and the evaluation of drugs
and services. Hosp Formul. 1993;28:365–368.
3. DiMasi J. Risks in new drug development: ap-
proval success rates for investigational drugs.
Clin Pharmacol Ther. 2001;69:297–307.
4. Grabowski H. Are the economics of pharmaceu-
tical research and development changing?: pro-
ductivity, patents and political pressures. Pharma-
coeconomics. 2004;22(2 Suppl 2):15–24.
5. Drummond M, Mason A. European perspective
on the costs and cost-effectiveness of cancer
therapies. J Clin Oncol. 2007;25:191–195.
6. Grabowski H. The role of cost-effectiveness
analysis in managed-care decisions. Pharmaco-
economics. 1998;14(Suppl 1):15–24.
7. Grabowski H, Mullins C. Pharmacy benefit man-
agement, cost-effectiveness analysis and drug
formulary decisions. Soc Sci Med. 1997;45:535–
544.
8. Taylor R, Drummond M, Salkeld G, Sullivan S. In-
clusion of cost effectiveness in licensing require-
ments of new drugs: the fourth hurdle. BMJ.
2004;329:972–975.
9. Mauskopf J, Schulman K, Bell L, Glick H. A strate-
gy for collecting pharmacoeconomic data during
phase II/III trials. Pharmacoeconomics. 1996;9:
264–277.
10. Annemans L, Genesté B, Jolian B. Early modelling
The authors report no relevant relationships to disclose.
Hemels, Wolden, Einarson
for assessing health and economic outcomes of
drug therapy. Value Health. 2000;3:427–434.
11. Clemens K, Garrison L, Jones A, Macdonald F.
Strategic use of pharmacoeconomics research in
early drug development and global pricing. Phar-
macoeconomics. 1993;4:315–322.
12. Grabowski H. The effect of pharmacoeconomics
on company research and development deci-
sions. Pharmacoeconomics. 1997;11:389–397.
13. Miller P. Role of pharmacoeconomic analysis in
R&D decision making: when, where, how? Phar-
macoeconomics. 2005;23:1–12.
14. Sculpher M, Drummond M, Buxton M. The itera-
tive use of economic evaluation as part of the
process of health technology assessment. J Health
Serv Res Policy. 2007;2:26–30.
15. World Health Organization. Anatomical Thera-
peutic Chemical (ATC) Classification System.
WHO Collaborating Centre for Drug Statistics
Methodology. Available at: http://www.whocc
.no/atcddd/ (accessed September 25, 2008).
16. Vernon J. Mathematical modeling and pharma-
ceutical pricing: analyses to inform in-licensing
and go/no-go decisions. Health Care Manag Sci.
2005;8:167–179.
17. DiMasi J, Hansen R, Grabowski H. The price of
innovation: new estimates of drug development
costs. J Health Econ. 2003;22:151–185.
18. DiMasi J, Hansen R, Grabowski H, Lasagna L. The
cost of innovation in the pharmaceutical indus-
try. J Health Econ. 1991;10:107–142.
19. Frank R. New estimates of drug development
costs. J Health Econ. 2003;22:325–330.
20. Love J. Who pays for what in drug development.
Nature. 1999;397:202.
21. Adams C, Brantner V. Estimating the cost of new
drug development: is it really 802 million dol-
lars? Health Aff (Millwood). 2006;25:420–428.
22. Girard P, Cucherat M, Guez D. Round Table No. 2
GX. Clinical trial simulation in drug develop-
ment. Therapie. 2004;59:287–304.
23. Palmer A, Roze S, Valentine W, et al. The CORE
Diabetes Model: projecting long-term clinical
outcomes, costs and cost-effectiveness of inter-
ventions in diabetes mellitus (types 1 and 2) to
support clinical and reimbursement decision-
making. Curr Med Res Opin. 2004;20:5–26.