Transcript Health Impact Fund

Health Impact Fund
Aidan Hollis
University of Calgary
February 24, 2008
University of California seminar on Designing
Strategies for Neglected Disease Research
1
Outline





Background to the problem
The Health Impact Fund (HIF)
Characteristics of the HIF
Comparison to other prize-type
systems
Questions
2
The Access Problem

Millions of people die or receive suboptimal care because the price of
patented medicines is higher than
manufacturing cost.


This affects primarily the poor.
It occurs also in rich countries when the
patentee raises prices to levels that even
insurers and national health systems
refuse to pay.
3
Access in the US
NYTimes:
“People are having to choose between gas, meals
and medication,” said Dr. James King, the
chairman of the American Academy of Family
Physicians, a national professional group.
“I’ve seen patients today who said they stopped
taking their Lipitor, their cholesterol-lowering
medicine, because they can’t afford it.”

“Scaling Back Medications as Economy Grows Sour” Oct 22 2008
4
The Innovation Problem

The investment in developing medicines for
diseases and conditions which primarily
affect poor people is relatively low.



In general, the correlation between profitability of
a drug and its health impact is imperfect leading
to incorrect incentives
Many proposed solutions to the access
problem can aggravate the innovation
problem.
We need a solution that addresses both
problems at the same time.
◄
5
Patents

Patents are good for drug
development, but not perfect:


They incentivize the profitable, not the
medically valuable
They work, if at all, through enabling high
prices, which inhibit access
6
The economics of drug
development



Current estimates of average drug
development costs vary substantially, from
$200m per product to $1.3bn per product
Clearly, firms are only willing to invest in
R&D when they expect meaningful returns.
Any solution must accept the fact that drug
development is expensive, and that the
rewards need to be large enough to pay for
the unsuccessful products.
7
The standard solutions

Solutions which can weaken the incentive for R&D:



Compulsory licenses – esp. when applied to neglected
diseases or to emerging economies
Arbitrary price reductions
Solutions which are partially effective, but limited:

Price discrimination across countries



Arbitrage and high income sub-markets in developing countries limit
profitability of this strategy
Ineffective for neglected diseases
Solutions which rely on government’s ability to pick
winners:

Research grants

Direct government funding of research is and will continue to be
extremely important
8
Health Impact Fund



Proposal is for a fixed sum to be paid
annually for pharmaceutical innovations.
Patentees could choose to exercise their
usual monopoly rights, or to opt in to the
HIF.
Opting in would mean that the product
would be sold globally at a low price (=
variable cost), and would earn its profits
through direct payments from the HIF based
on health impact.
9
Proposed reward mechanism



The HIF would offer a fixed annual reward of
$6bn financed by governments
Each participating firm would be given a
share of the fixed reward in each of 10 years
following the introduction of its product.
The share would be equal to its share of
total QALYs generated by all participating
products


QALY: Quality-Adjusted Life-Year
The share would be re-evaluated annually
based on current data
10
How much would firms earn?

The payment for a product would depend on





the health impact of the product, and
the health impacts of other products
the total rewards available
As with the price system, earnings would
depend on the relative quality and sales of a
given product compared to others, but also
on how well it is used.
Firms would form expectations of earnings
based on past rates of reward per QALY.
11
Additional points 1




Health impact would be assessed globally,
without distinction among countries or medical
conditions.
Following the reward period, firms would be
obliged to offer global zero-priced licenses on
all intellectual property required for
manufacture and sale.
The 10-year period of reward payments
mimics revenues under a patent.
This system does not undermine patent rights
– it is simply a different way of being paid.
12
Additional points 2

In the early years, the best opportunities
for HIF-rewards are likely to be in the
traditionally neglected area of tropical
(rather than global) diseases, stimulating
innovations that would not have existed
without the HIF.
13
Determining the sale price

One way to set the sale price would be to
require the firm to seek tenders for
production.



The tenders could be repeated every year or two
as suited to the product.
The tenders could be geographically specific.
The expectation is that the drug developer
would obtain no profits from selling the drug:
profits would obtain from health impact
rewards.
14
The importance of opting in


Firms would have a choice. Only those that
expected to make more under the HIF would
opt in.
This means that the outside opportunity
would regulate the rewards inside the HIF: if
◄
the payments from the HIF became too
generous, more firms would opt in, reducing
the payment per QALY (and vice versa).
15
Which products would be in
the HIF?

Firms would choose the HIF for any
product with potential health impact
that is high relative to the product’s
profitability with monopoly pricing.


e.g. drugs for neglected diseases
e.g. drugs for which patents did not
provide effective protection from generics
16
HIF resolves critical problems
in prize determination


Which products should be rewarded, and
how large should the prize be?
The HIF is a market-based solution:
payments are determined by competition
between all registered products for the
available rewards.


◄
A drug for malaria can directly compete against
a drug for HIV/AIDS.
This regulates relative rewards for registered
products, rewarding each at the same rate per
QALY, creating efficient incentives.
17
Assessing health impact

Health impact would be assessed compared
to the outcomes that could have been
expected to occur given the state of
technology two years before, and excluding
the firm’s own patented products.


This is tricky!
It is especially difficult to assess QALYs in
developing countries because of data
problems, weaker healthcare infrastructure,
and less compliance with therapeutic
regimen.
18
QALY confusion


There are different, conflicting approaches to
QALY assessment.
The HIF would have to choose an approach



It doesn’t have to choose a value per QALY – this
arises endogenously
The HIF process would be subject to strategic
gamesmanship … just like other systems.
The HIF’s reliance on health impact
assessment makes explicit and transparent
the emerging practice of insurers to make
◄
what they cover conditional on some
measure of value, which has been subject to
the same critiques.
19
Assessment


Health impact will be assessed annually
based on available information and
inference
Assessment will rely on data from






Clinical trials
Pragmatic or practical trials
Audited data on sales
Stratified sampling of use of the product in
different environments
Administrative databases
Global burden of disease data
20
QALYs vs. Prices


If you don’t like the idea of rewarding based
on an imperfect measure of QALYs, do you
really think that prices are a better measure
of a drug’s value?
The key question: are prices, since they
arise out of a “market” mechanism (however
imperfect), intrinsically better than QALYs,
which are estimated by an administrative
body?
21
How about prices?


Currently, firms receive a reward per pill
equal to the price less the cost of
manufacturing and distribution.
Why is that not equal to “value”?




Lack of information on the part of consumers
Consumer typically does not pay the full cost
Nor does the prescribing physician bear the cost
Willingness to pay may not reflect value
22
Assessment cost


This system would be expensive to run.
Assessment costs would probably be at
least 10% of the fund payout or $600m per
year.
But assessment of health impact is a priority
in almost all countries already.



Clinical reasons
Budgetary reasons
Assessment costs are therefore partly
balanced by collateral benefits.
23
Financing

Firms would require there to be a longterm commitment


Only governments, of affluent and
developing countries, can commit large
sums long-term.


Otherwise no innovation incentive effects
We propose a small share of GNI for each
member country.
$6bn a year is less than 1% of global
expenditures on pharmaceuticals.
24
Financing cost [1]


Currently about 90% of expenditures on
patented drugs are in developed countries;
Most expenditures are funded by
government and employment-based
insurance plans.

Thus currently, most of the cost of developing
pharmaceuticals is borne by rich countries,
through taxes and insurance premia.
◄
25
Financing cost for non-induced
products




If registered products would have been developed in
the absence of the HIF, the incremental cost of the HIF
to taxpayers is small.
Without the HIF, the product would be sold at a high
price, paid by insurer/government, funded by insurance
premium/taxes.
With the HIF, the taxpayer must contribute to the HIF,
but direct expenditure on drugs falls because of lower
price.
HIF eliminates the need to exclude potential buyers
who can afford the product at variable cost but could
not afford the much higher monopoly price.
◄
26
Financing cost for induced
products


If registered products would not have been
developed anyhow, there is an additional cost to
taxpayers.
If the products are for global diseases, the
incremental cost is for drugs which have a high
impact on health globally.


Any products in the HIF must have a higher QALY/$ ratio
than products not registered.
If the products are for tropical diseases, the
incremental cost to non-tropical countries is
performance-based foreign aid.

The rich have prudential as well as altruistic reasons for
addressing tropical diseases
27
Financing – which budget?



Most other proposals which involve real
money require that money to come from
ODA budgets.
The HIF requires a lot of funding because it
is preventing firms from profiting from high
prices in developed countries.
Since it is saving health budgets large
amounts of money, the savings should
◄
finance the HIF. Thus, there is no reason to
make ODA budgets the source of HIF
funding.
28
Too much government
involvement?


One objection is that this would allow a
multilateral quasi-governmental
institution to intrude into the market.
Mitigating factors:


Discretion is limited: impacts are
assessed and rewards allocated on the
basis of objective measures.
Governments are already in this market
as big buyers and price-setters.
29
The “last mile” problem in drug
delivery


Proper prescribing and compliance are
essential to drug effectiveness.
The HIF pays on the basis of each
medicine’s actual health impact,


as assessed not only through sales data, but
sampling of actual use and population health
data
Firms therefore have incentives to promote
appropriate use of their registered products,
as well as to develop products which are
◄
effective in resource-poor settings.
30
Everyone can win from better
institutional design



The HIF allows pharmaceutical innovators to
concentrate on promoting public health
while satisfying shareholders.
Patients win through gains in innovation,
lower prices for important new drugs, and
additional efforts by patentees to achieve
actual health impact.
Taxpayers are assured of good value for
money.
31
The HIF is a global insurance
system



Consumers contribute a co-pay based
on the variable cost of the drug.
Everybody pays a premium based on
income.
The “price” earned by the drug
innovator is based on the therapeutic
value.
32
Design Challenges




Can health impact assessment be
sufficiently consistent to be attractive to
companies and funding partners?
How will companies form initial expectations
about the reward per QALY?
How can the HIF ensure reliable funding?
How should the HIF determine who is
eligible for rewards?
33
Comparison with other prize
systems

There are other prize-type systems.
How does the HIF fit in?



AMCs
Patent pool/DEFEND proposal
Helpful first to consider what are the
basic parameters which vary across
the proposals
34
Basic parameters





Product scope
Geographic scope
Determination of prize amount
Method of ensuring low prices
Source of funding
35
Product scope



AMCs: limited to a vaccine
Patent pool: unlimited
HIF: unlimited


Whatever products have the greatest ratio of
health impact / patent profitability will be
funded under the HIF
Solves the problem of determining the reward
amount
36
Geographic scope

AMCs


Patent pool


Limited to developing countries
Limited to developing countries
HIF

Global

Implication: the HIF is not just development
aid. It can be paid for through reductions in
drug prices in rich countries.
37
Determination of prize amount

AMCs


Patent pool


Determined in advance on the basis of expected
cost?
? Perhaps based on royalty from generics
HIF

Share of assessed health impact

i.e. pay for performance
38
Method of ensuring low prices

AMCs


Patent pool


Price controls
Open licensing
HIF

Competitive tender
39
Source of funding



AMCs: ODA
Patent pool: generic royalties + ODA
HIF: reduction in drug expenditures in
◄
developed countries
40
How the HIF features fit
together


Allowing all products is necessary to
generate competition so that reward
payments based on health impact are
appropriate
Global geographic scope is important
to (a) create incentives for high value
products to register and (b) make cost
savings into the funding source
41
Progress




We are working on obtaining input from a
variety of stakeholders, including
pharmaceutical companies and
governments
We are testing consistency of health impact
assessment on existing drugs
We are running simulations
We are exploring legal issues, including
incorporation, patent issues, contracts…
42
Thank you!

More information available at
www.healthimpactfund.org
43
Supplementary slides



Mandatory licensing vs. price controls
Risk
Comparison to insurance
44
Why not require open
licensing instead of tendering?

Generic competition is often not robust


Even robust generic competition does not always
lead to low retail prices


Especially initially, and in small markets and countries
True in both developed and developing countries
Those products which did not attract enough
generic competition would be able to double up on
rewards plus high prices, reducing payments for
other drugs.

This would make the HIF very attractive for complex
biologics where non-patent barriers enabled monopolies.
The firm could register with the HIF without loss of sales
revenues in developed countries.
45
Risk


Firms appear to face extra risk since profits depend
on sales volume, on the actual impact of drugs sold,
and on the reward rate per QALY. The reward rate
is outside their control.
Risk is mitigated by





Optional nature
Other firms opting in or out will stabilize profitability
Experience from past years
Hard to know how expectations would be formed in
advance of the HIF being implemented.
Might set minimum or maximum reward/QALY?
46
Why not just rely on
insurance?

Typically insurers make payment conditional
on “cost-benefit” trade-off



HIF can always include any product; but if
benefit is low, reward is low.
HIF is different from standard insurance
mechanisms, since it doesn’t require
exclusion of some products.
The HIF is explicitly designed for the
pharmaceutical market.
47