[Pharm-policy] fixed: Allocation of R&D costs, the 1991 DiMasi study
James Love
love@cptech.org
Thu, 03 Feb 2000 18:03:31 -0500
This is a spell checked version! Jamie
--------------
The following are numbers from Joe DiMasi, et. all's 1991 paper
on drug development costs. (DiMasi, Hansen, Grabowski and
Lasagna, "Cost of Innovation in the pharmaceutical industry,"
Journal of Health Economics 10 (1991)). This is the most
famous of all cost studies, and the one that nearly everyone
uses as a basis for cost estimates. Joe has kindly looked at
this note already.
In Table 5 of his paper, Joe and his co-authors presented
sensitivity analysis of the costs of developing a drug,
using different assumptions regarding the costs of capital.
Joe DiMasi's 1991 paper, using 1987 dollars
Table 5, using .23 success
with with with
Capital Capital Capital
out of pocket Costs Costs Costs
per approval @ 5% @ 9% @ 15%
------------ ------------ ------------ -----------
Preclinical 66 57.9% 107 63.3% 156 67.5% 269 72.7%
Clinical 48 42.1% 62 36.7% 75 32.5% 101 27.3%
Total 114 100.0% 169 100.0% 231 100.0% 370 100.0%
Given some technical issues that Joe and I are talking about, I am not
getting into the (Phase I, II and III, before and after risk adjustment)
detail of of clinical expenses. Only the "per approval" numbers are
reported for pocket expenses. This is a risk adjusted number, that
takes account of failures. The risk adjusted out of pocket costs for
clinical trials were $48 million, and the risk adjusted pre-clinical
expenditures were $66 million.
Joe assumed 9 percent was the cost of capital, so the $231 million
total was quoted by people for a while, until OTA updated his numbers
in 1990 dollars, and looked at different capital costs, for sensitivity
analysis. Note that Joe estimated that 58 percent of all out pocket
(per- approved drug) costs were for the pre-clinical phase, and when
capital costs were assumed to be 9 or 15 percent, the pre-clinical
share is even larger: 68 to 73 percent. Joe's figures only apply
to "self originated" drugs, and not to drugs licensed from a
University or government. One point to keep in mind, is that if
Yale licenses d4T to BMS, most if not all of the pre-clinical work
is paid for already (often by the US government).
Joe suggested I present these in 1998 dollars, using
an inflation factor of 35.6 percent:
In 1998 dollars:
with with with
Capital Capital Capital
out of pocket Costs Costs Costs
per approval @ 5% @ 9% @ 15%
------------ ------------ ------------ ------------
Preclinical 89 57.9% 145 63.3% 212 67.5% 365 72.7%
Clinical 65 42.1% 84 36.7% 102 32.5% 137 27.3%
Total 155 100.0% 229 100.0% 313 100.0% 502 100.0%
Next, I wanted to present the capital costs as a separate item from the
per approval out of pocket expenses. (This is the total cost from
above, minus the out of pocket expenses). As indicated below,
capital costs (the profits you need to justify the investments),
are 51 percent with a 9 percent real return is assumed, and 69
percent of the total when a 15 percent real return is assumed.
out of pocket and capital costs compared
1998 dollars, capital costs at 9 percent
out of Capital out of Capital
pocket Costs pocket Costs
p/app @9% total p/app @9% total
--------------------- --------------------
Preclinical 89 122 212 28% 39% 67%
Clinical 65 37 102 21% 12% 32%
Total 155 159 313 49% 51% 100%
out of pocket and capital costs compared
1998 dollars, 15 percent capital costs
out of Capital out of Capital
pocket Costs pocket Costs
p/app @15% total p/app @15% total
----------------------- --------------------
Preclinical 89 276 365 18% 55% 73%
Clinical 65 72 137 13% 14% 27%
Total 155 347 502 31% 69% 100%
Finally, here are a few interesting facts on the out of pocket cost
numbers.
DiMasi's 1987 unweighted average cost for clinical trials, for
drugs that were approved by the FDA. (Before risk adjustments).
$28 million
The 1987 clinical trial costs, after risk adjustments:
$48 million
The risk adjusted (per approval) private sector expenditures
on clinical trials, for Orphan Drugs, based upon the
IRS data, nominal data, for 1983 to 1993:
$2.3 million.
There are, of course, a lot of different issues that
one might raise in these studies, and I'll address
some of these later, such as assumptions regarding
the time to NDA.
Jamie
--
=======================================================
James Love, Director | http://www.cptech.org
Consumer Project on Technology | mailto:love@cptech.org
P.O. Box 19367 | voice: 1.202.387.8030
Washington, DC 20036 | fax: 1.202.234.5176
=======================================================