Beruflich Dokumente
Kultur Dokumente
research supports improved delivery, access, and cost. Few previous analyses have compared
the United States with other developed countries.
Supplemental content at
jama.com
OBJECTIVES To quantify total public and private investment and personnel (economic inputs)
and to evaluate resulting patents, publications, drug and device approvals, and value created
(economic outputs).
EVIDENCE REVIEW Publicly available data from 1994 to 2012 were compiled showing trends
in US and international research funding, productivity, and disease burden by source and
industry type. Patents and publications (1981-2011) were evaluated using citation rates and
impact factors.
FINDINGS (1) Reduced science investment: Total US funding increased 6% per year
(1994-2004), but rate of growth declined to 0.8% per year (2004-2012), reaching $117 billion
(4.5%) of total health care expenditures. Private sources increased from 46% (1994) to 58%
(2012). Industry reduced early-stage research, favoring medical devices, bioengineered
drugs, and late-stage clinical trials, particularly for cancer and rare diseases. National Insitutes
of Health allocations correlate imperfectly with disease burden, with cancer and HIV/AIDS
receiving disproportionate support. (2) Underfunding of service innovation: Health services
research receives $5.0 billion (0.3% of total health care expenditures) or only 1/20th of
science funding. Private insurers ranked last (0.04% of revenue) and health systems 19th
(0.1% of revenue) among 22 industries in their investment in innovation. An increment of
$8 billion to $15 billion yearly would occur if service firms were to reach median research
and development funding. (3) Globalization: US government research funding declined from
57% (2004) to 50% (2012) of the global total, as did that of US companies (50% to 41%),
with the total US (public plus private) share of global research funding declining from 57% to
44%. Asia, particularly China, tripled investment from $2.6 billion (2004) to $9.7 billion
(2012) preferentially for education and personnel. The US share of life science patents
declined from 57% (1981) to 51% (2011), as did those considered most valuable, from 73%
(1981) to 59% (2011).
CONCLUSIONS AND RELEVANCE New investment is required if the clinical value of past
scientific discoveries and opportunities to improve care are to be fully realized. Sources could
include repatriation of foreign capital, new innovation bonds, administrative savings, patent
pools, and public-private risk sharing collaborations. Given international trends, the United
States will relinquish its historical international lead in the next decade unless such measures
are undertaken.
Key Questions
We address 3 major trends:
1. Diminished funding in the United States from both public and private sponsors at a time when scientific opportunity has never
been greater but when support for sustained, long-term investments is limited and short-term performance is rewarded disproportionately
2. Establishing strong incentives for investment in health service and
delivery innovations and better ways to deliver care
3. The implications of globalization
Better understanding of these factors is required if the full promise of the cumulative investment in biomedical science and opportunity for improved services are to be realized.
Information in 8 areas has been assembled to inform the discussion (Figure 1). Two areas involve the current and historical landscape in the United States of investment and employment in medical research, placing the United States in an international context.
Two areas examine funding on biomedical and health services research separately. Four areas quantify the value of that investment
as judged by resulting patents, publications, drug and device approvals, and public market performance of life science and health
service companies.
Methods
To describe and document the current anatomy and historical trends
of medical research, we assembled an array of information from various data sources. We relied on publicly available data, recalculated
those data for display when necessary, reconciled inconsistent
sources, and included years for which data are complete (in general, from 1994 to 2012). The Box contains a list of the included and
supplementary figures and tables.
Methods were similar to those we have used previously.1-3 Additionally, in this study, the 40 largest developed nations were ex-
Biomedical research
Historical funding trends
Funding by phase of
research
Funding by therapeutic area
jama.com
175
slowed (medical device, 6.6% to 6.2% in 1994-2004 vs 20042012; biotechnology, 14.1% to 4.6% in 1994-2004 vs 20042012), or declined (pharmaceutical firms, 6.8% to 0.6% in 19942004 vs 2004-2012).
Biotechnology firmsc
Research Funding
Other federalb
Pharmaceutical firms
Biomedical Research
jama.com
140
The distribution of investments across the types of medical research changed from 2004 to 2011. Pharmaceutical companies
shifted funding to late-phase clinical trials and away from discovery activity such as target identification and validation. The share
of pharmaceutical industry funding (including that by US companies outside of the United States) spent on phase 3 trials increased
by 36% (5%/year growth rate) from 2004 to 2011 (Figure 4), and
the share of investment in prehuman/preclinical activities decreased by 4% (2%/year average decline). This shift toward clinical
research and development reflects the increasing costs, complexity, and length of clinical trials but may also reflect a deemphasis of
early discovery efforts by the US pharmaceutical industry. While industry has shifted funding to clinical trials, the share of NIH contributions dedicated to basic science and clinical research was unchanged (eTable 2 in the Supplement), with the majority of funds
still focused on basic research. These data may not accurately reflect the true division of NIH investment for basic science vs diseasefocused research, as a growing proportion of NIH expenditures is
for projects having potential clinical application in many diseases or
organ systems.7
In real terms, venture capital investment in biotechnology companies steadily increased from $1.5 billion in 1995 to a peak of $7.0
billion in 2007 (eFigure 3 in the Supplement). During that period,
investment in biotechnology companies as a share of total venture
capital investment increased from 10% to 18%, and the number of
investments increased from 176 to 538. Investment levels and the
number of transactions of biotechnology decreased following the
financial crisis in 2008-2009, declining to a low of $4.3 billion in
2009. Venture capital investment still has not recovered to its pre2008 levels, with only $4.5 billion invested in 2013. Size of investment per transaction (median, $11 million, inflation adjusted) has remained unchanged for 2 decades.
Public funding of medical research by condition was only marginally associated with disease burden in the United States in 2010
(eFigure 4 in the Supplement). A set of 27 diseases that account
for 84% of US mortality, 52% of years of life lived with disability,
84% of years of life lost, and 70% of disability-adjusted life-years
receive 48% of NIH funding (R2 = 0.26) (eTable 3 in the Supplement). Several factors other than disease burden may influence
funding, including the quality of research, scientific opportunity,
portfolio diversification, or building of infrastructure, and the combination of these factors complicates the relationship of funding to
particular conditions.8,9 Cancer and HIV/AIDS were funded well
above the predicted levels based on US disability alone (eFigure 4
in the Supplement), with cancer accounting for 16% ($5.6 billion)
of total NIH funding and 25% of all medicines currently in clinical
trials (Figure 5).
Rare diseases have emerged for industry as a preferential area
of therapeutic development, with nearly as many compounds in
trials as analgesics and antidiabetic drugs (Figure 5). Industry
favors rare diseases because they are commercially attractive due
120
100
80
60
40
20
0
1994
1996
1998
2000
2002
2004
Year
2006
2008
2010
2012
Includes
ARRA Fundingb
Data were adjusted to 2012 dollars using the Biomedical Research and
Development Price Index.4
The National Institutes of Health and other federal sources include stimulus
provided by ARRA in 2009 and 2010.
Compound annual growth rate (CAGR) supposing that year A is x and year B is
y, CAGR = (y/x){1/(BA)}1. The CAGR was calculated separately for 2 different
periods with a single overlapping year: 1994-2004 and 2004-2012. The cut
point was chosen at 2004 given the changes seen in funding from the
National Institutes of Health in that year.
Health services research, which examines access to care, the quality and cost of care, and the health and well-being of individuals,
communities, and populations, accounted for between 0.2% and
(Reprinted) JAMA January 13, 2015 Volume 313, Number 2
177
slowed (medical device, 6.6% to 6.2% in 1994-2004 vs 20042012; biotechnology, 14.1% to 4.6% in 1994-2004 vs 20042012), or declined (pharmaceutical firms, 6.8% to 0.6% in 19942004 vs 2004-2012).
Biotechnology firmsc
Research Funding
Other federalb
Pharmaceutical firms
Biomedical Research
jama.com
140
The distribution of investments across the types of medical research changed from 2004 to 2011. Pharmaceutical companies
shifted funding to late-phase clinical trials and away from discovery activity such as target identification and validation. The share
of pharmaceutical industry funding (including that by US companies outside of the United States) spent on phase 3 trials increased
by 36% (5%/year growth rate) from 2004 to 2011 (Figure 4), and
the share of investment in prehuman/preclinical activities decreased by 4% (2%/year average decline). This shift toward clinical
research and development reflects the increasing costs, complexity, and length of clinical trials but may also reflect a deemphasis of
early discovery efforts by the US pharmaceutical industry. While industry has shifted funding to clinical trials, the share of NIH contributions dedicated to basic science and clinical research was unchanged (eTable 2 in the Supplement), with the majority of funds
still focused on basic research. These data may not accurately reflect the true division of NIH investment for basic science vs diseasefocused research, as a growing proportion of NIH expenditures is
for projects having potential clinical application in many diseases or
organ systems.7
In real terms, venture capital investment in biotechnology companies steadily increased from $1.5 billion in 1995 to a peak of $7.0
billion in 2007 (eFigure 3 in the Supplement). During that period,
investment in biotechnology companies as a share of total venture
capital investment increased from 10% to 18%, and the number of
investments increased from 176 to 538. Investment levels and the
number of transactions of biotechnology decreased following the
financial crisis in 2008-2009, declining to a low of $4.3 billion in
2009. Venture capital investment still has not recovered to its pre2008 levels, with only $4.5 billion invested in 2013. Size of investment per transaction (median, $11 million, inflation adjusted) has remained unchanged for 2 decades.
Public funding of medical research by condition was only marginally associated with disease burden in the United States in 2010
(eFigure 4 in the Supplement). A set of 27 diseases that account
for 84% of US mortality, 52% of years of life lived with disability,
84% of years of life lost, and 70% of disability-adjusted life-years
receive 48% of NIH funding (R2 = 0.26) (eTable 3 in the Supplement). Several factors other than disease burden may influence
funding, including the quality of research, scientific opportunity,
portfolio diversification, or building of infrastructure, and the combination of these factors complicates the relationship of funding to
particular conditions.8,9 Cancer and HIV/AIDS were funded well
above the predicted levels based on US disability alone (eFigure 4
in the Supplement), with cancer accounting for 16% ($5.6 billion)
of total NIH funding and 25% of all medicines currently in clinical
trials (Figure 5).
Rare diseases have emerged for industry as a preferential area
of therapeutic development, with nearly as many compounds in
trials as analgesics and antidiabetic drugs (Figure 5). Industry
favors rare diseases because they are commercially attractive due
120
100
80
60
40
20
0
1994
1996
1998
2000
2002
2004
Year
2006
2008
2010
2012
Includes
ARRA Fundingb
Data were adjusted to 2012 dollars using the Biomedical Research and
Development Price Index.4
The National Institutes of Health and other federal sources include stimulus
provided by ARRA in 2009 and 2010.
Compound annual growth rate (CAGR) supposing that year A is x and year B is
y, CAGR = (y/x){1/(BA)}1. The CAGR was calculated separately for 2 different
periods with a single overlapping year: 1994-2004 and 2004-2012. The cut
point was chosen at 2004 given the changes seen in funding from the
National Institutes of Health in that year.
Health services research, which examines access to care, the quality and cost of care, and the health and well-being of individuals,
communities, and populations, accounted for between 0.2% and
(Reprinted) JAMA January 13, 2015 Volume 313, Number 2
177
120
1994
2004
2012
1994-2004
2004-2012
2.6 (4)
3.9 (4)
4.2 (4)
4.2
1.0
3.9 (7)
5.9 (5)
6.3 (5)
4.1
0.9
Other federal
8.0 (13)
4.8 (4)
7.1 (6)
4.9
5.0
17.6 (29)
35.6 (33)
30.9 (27)
7.3
1.8
6.2
Funding Source
100
80
60
40
20
3.8 (6)
7.1 (6)
11.5 (10)
6.6
Biotechnology firms
3.7 (6)
13.7 (12)
19.6 (17)
14.1
4.6
Pharmaceutical firms
20.0 (34)
38.6 (35)
36.8 (32)
6.8
0.6
Overall
59.5
6.3
0.8
116.5
109.7
0
1994
2004
2012
Year
b
Compound annual growth rate (CAGR) supposing that year A is x and year B is
y, CAGR = (y/x){1/(BA)}1.
2011
2004-2011
50
Uncategorizedc
4.2 (9)
1.7 (3)
11.9
40
Phase 4
6.4 (13)
4.8 (10)
3.9
Approval
4.5 (9)
4.1 (8)
1.2
Phase 3
12.6 (26)
17.6 (36)
4.9
20
Phase 2
4.9 (10)
6.2 (13)
3.3
10
Phase 1
3.2 (7)
4.3 (9)
4.1
Prehuman/preclinical
12.5 (26)
10.6 (22)
2.3
Overall
48.3
49.3
Phase of Research
30
2004
2011
Compound Annual
Growth Rate, % b
Data were adjusted to 2012 dollars using the Biomedical Research and
Development Price Index.4
Compound annual growth rate (CAGR) supposing that year A is x and year B is
y, CAGR = (y/x){1/(BA)}1.
ing, or 1.7%-2.5% of revenue). Health insurers may provide additional health services research funding that cannot be distinguished from the insurance industry as a whole, although these
funds are small and unlikely to change the results for industry
funding (Figure 7).
0.3
jama.com
or 9.4% per year, with especially large increases in China, India, South
Korea, and Singapore.
These trends resulted in the restructuring of the share of total
global investment (eFigure 5 in the Supplement). As a percentage
of global funding, the United States declined by approximately 13%
from 2004 to 2012, and Asian economies increased by approximately the same share (13% in 2004 to 20% in 2011). Although absolute growth of Asian investment from 2004 to 2011 reached $24
billion, the United States remained the leading sponsor of global
medical research in 2011 (44% share), and Europe the next largest
sponsor (33% share).
Overall growth was slightly greater for industry outside the
United States compared with public sources (4.3% vs 2.2%), and industry accounted for two-thirds of funds in 2011. However, US contributions increased slowly from both public (0.1%/year) and industry sources (1.7%/year).
Public funding in the United States decreased to a 49% share
of the worlds public research investment by 2011, down from 57%
in 2004 (Figure 8). United States industry, which accounted for
nearly half of global industry medical research expenditures in 2004,
declined to 41% of global industry funding in 2011 (Figure 8). Japan
demonstrated the greatest increase in the worlds share of industry funding (+3.9%), and European countries gained the most in public investment (+3.5%). Despite decreases in the US share of investment, the United States remained the worlds leading sponsor for
both public and industry medical research funding in 2011.
Although China led the world in the overall size of their science
and technology workforce, it had only 1.9 science and technology
workers per 100 000 full-time equivalents, the lowest among the
countries included in the analysis (Figure 9). The United States employed 8.1 science and technology workers per 100 000 full-time
equivalents in its total workforce, or the median for the 10 largest
workforces in the world.
The investment in capital terms and in labor terms differ widely
across countries and regions. The United States contributes 44.2%
of global medical research funding but comprises only 21.2% of the
200
400
600
Data for the number of compounds in development were from the Citeline
Pharma R&D Annual Review 2014.10 Data for rare diseases were from the
Pharmaceutical Research and Manufacturers of America.11
a
Rare diseases were defined as those affecting 200 000 or fewer people in the
United States.
Funding source
6
2004
2011
2004-2011
653 (18)
1352 (27)
11.0
AHRQ
365 (10)
1018 (20)
15.8
1158 (32)
1189 (24)
0.4
Other federald
442 (12)
494 (10)
1.6
Foundationse
1034 (28)
967 (19)
1.0
Overall
3652
NIH
Compound Annual
Growth Rate, % b
5019
4.6
2004
2011
AHRQ indicates Agency for Healthcare Research and Quality; NIH, National
Institutes of Health. Data were calculated according to methods outlined in
eTable 5 in the Supplement.
a
Compound annual growth rate (CAGR) supposing that year A is x and year B is
y, CAGR = (y/x){1/(BA)}1.
jama.com
health services research and quality improvement initiatives for the US health
care services industry.
d
Other federal funding includes the Centers for Disease Control and Prevention,
Centers for Medicare & Medicaid Services, Veterans Health Administration,
Health Resources and Services Administration, and Patient Centered
Outcomes Research Institute (in 2011 only).
Foundation funding includes total giving from the Robert Wood Johnson
Foundation, California Endowment, Pew Charitable Trusts, W. K. Kellogg
Foundation, and Commonwealth Fund.
179
Figure 7. Research and Development Investment Ranking of Industrial Sectors Among US-Based Companies, 2011
Total research and development fundinga
Medical devices
Machinery
Medical devices
Chemicals
Median
Chemicals
Architectual engineering
Telecommunications
Telecommunications
Architectual engineering
Utilities
Utilities
Domestic
Insurance carriers
Transportation services
Foreign
0.2
Transportation services
0
0.04
Insurance carriers
20
40
60
80
0.04
0
10
15
global science and technology workforce (eFigure 6 in the Supplement). Conversely, China contributes only 1.8% of global funding for
medical research but comprises 22.3% of the global science and technology workforce. This difference in investment represents a natural experiment in productivity management and has broad implications for patents and intellectual property ownership, which will
evolve over the next few years.
China filed 30% of global life science patent applications in 2011, increasing from 1% of global applications in 1991 (Figure 10). This includes applications from a number of patenting offices throughout
the world, including offices in China, the United States, and the
European Union. The United States followed with 24% of patent filings globally, increasing from an 11% share in 1991.
United States inventors led in the number of life science patent filings in both the United States and EU, where China accounted for less than 2% of filings in both regions (Figure 11 and
180
Median
Publications
The United States led the world with 33% of published biomedical
research articles in 2009 (Figure 13A). In the United States, the number of biomedical research articles increased at 0.6% per year from
2000 to 2009. During the same period, the number of articles published in China increased by 18.7% annually.
The United States also leads the world in its share of the most
highly cited biomedical research articles, with 63% of the top cited
jama.com
Publicb
Industryc
Globale
United
States
Europe
Japan
China
Other
Asiaf
Canada
Australia
37.8 (14)
4.9 (1.2)
9.7 (4)
3.1 (1.2)
3.8 (1.4)
Publicb
17.0 (17)
1.3 (2)
2.4 (2)
1.8 (2)
2.8 (3)
Industryc
20.8 (13)
3.6 (0.8)
7.3 (4)
1.3 (0.8)
1.0 (0.6)
4.5
9.3
3.5
1.0
4.1
Since 2003, drug approvals by the US Food and Drug Administration (FDA) have remained unchanged with an average of 26 approvals per year. Although drug approvals increased slightly in 2011 and
2012, they returned closer to average in 2013 with 27 approvals (eFigure 9 in the Supplement). United States device approvals have also
remained relatively constant over the last decade. While the number of approvals steadily increased from 15 approvals in 2009 to 39
approvals in 2012, only 22 new devices were approved in 2013.
During the same period, the European Medicines Agency (EMA)
averaged a higher number of both applications (55/year) and approvals (42/year) than the FDA (eFigure 9). In 2013, the EMA received 22 more applications and approved 16 more drugs than the
FDA.
Life Sciences Market Performance
Equity (stock) markets reflect broad public perception of one industrys value in comparison with others. Since 2003, market return for the entire health care industry (including medical device,
jama.com
6.8
16.9
20.8
Compound annual growth rate (CAGR) supposing that year A is x and year B is
y, CAGR = (y/x){1/(BA)}1.
Global total for medical research funding includes research and development
expenditures from 36 major world countries across 4 continents.
pharmaceutical, and biotechnology companies as well as hospitals, nursing homes, and other health service suppliers) as measured by the Dow Jones US Health Care Index increased 8.2% annually, closely trailing the Standard & Poors 500 (8.3%) (Figure 14).
Market returns for biotechnology and health insurance companies
outperformed the market, growing at 18.5% and 13.8% per year, respectively. Medical device companies, pharmaceutical companies,
and hospital chains underperformed compared with the Standard
& Poors 500, increasing annually at 7.3%, 6.8%, and 5.8%, respectively. The financial crisis of 2008 led to a decrease in market performance for all life sciences industries. Generally, all sectors recovered in the years following, and biotechnology companies, hospital
chains, and health insurance companies performed exceptionally well
since their decline in 2008-2009.
181
1400
1996
1200
2011
1000
800
600
400
200
0
Chinac
United
States
Japan
Russian
Federation
Germany
United
Kingdom
Korea
France
Canada
Spain
6.0
2.7
0.4
1.5
2.6
3.7
7.4
3.2
3.8
6.4
12
1996
10
2011
8
6
4
2
0
Korea
Japan
France
Canada
United
States
Germany
6.2
0.5
2.5
2.2
1.8
2.2
2.9
2.1
Spain
Chinac
4.1
5.2
The sizes of national science and technology workforces were obtained from
the Organisation for Economic Co-operation and Development.16
Compound annual growth rate (CAGR) supposing that year A is x and year B is
y, CAGR = (y/x){1/(BA)}1.
United
Russian
Kingdom Federation
The information assembled demonstrates that 3 factors, wavering financial support for science, underinvestment in service innovation, and globalization, pose the chief challenges of the current era.
Biomedical Research
New knowledge about disease has a 15- to 25-year gestation from
basic discovery to clinical application, an interval that may be
lengthening.22,23 Hence, the cumulative investment in biomedical
research of the past 3 decades will soon mature. Therefore, ensuring sufficient support for its clinical development is a pressing need.
Equally important are stable academic institutions and companies
along with skilled researchers that have the capability to organize
the research process and to sustain the innovation cycle,24 particularly since the size of research teams and scale of activities have
grown. Year to year variability in funding is a threat to that stability.
jama.com
Figure 10. Global Life Science Patent Applications by Country of Origin, 1981-2011
No. of patent family applications
in life sciencea
400 000
Other
Other
Germany
Germany
80
Japanb
300 000
Japanb
Russia
Percentage
Russia
No.
Taiwan
India
200 000
Australia
60
Taiwan
India
Australia
40
Canada
100 000
Canada
Korea
Korea
20
United States
United States
China
0
1981
1991
2001
China
2011
1981
1991
Year
2011
Year
The number of patent family applications by country filed was calculated based
on data obtained from Thomson Innovation.17 Only the most recent patent
application in a patent family was counted for this analysis. Data are included for
all countries available in the Thomson data set.
a
2001
Only patent grants, not all patent applications, are counted for Japan, which
tends toward patent applications with narrower definitions and therefore
much greater numbers relative to the number of patents ultimately granted.
Percentage distribution
by country of inventor
100
Other
Netherlands
Great Britain
60
Switzerland
40
France
20 000
Germany
France
Germany
20
Japan
United States
0
1981
1991
2001
2011
Taiwan
Percentage
No.
Korea
40 000
Great Britain
China
Taiwan
Switzerland
Korea
Netherlands
80
China
60 000
Other
Japan
United States
0
1981
Year
1991
2001
2011
Year
In the United States and Europe, private companies will not likely
have the latitude from their investors, or governments the political
will, to continue to make long-term investments at historical levels.
Todays political and commercial environment leads to this conclusion. Many new basic discoveries that have probable clinical value
are stymied by financial constraints at the critical proof-of-concept
stage, where utility in humans is demonstrated. That number can
be expected to increase once platform technologies (such as highresolution mapping of the central nervous system, analysis of complex biological systems and networks, or insights into developjama.com
183
Figure 12. Highly Valuable US Life Science Patents by Country of Origin, 1981-2011
No. of life science patent applications
in top 10% of patents by inventor countrya,b
8000
Other
Other
China
China
80
Netherlands
6000
Netherlands
Korea
Percentage
Korea
No.
Canada
Switzerland
4000
France
60
Canada
Switzerland
France
40
Great Britain
2000
Great Britain
Germany
Germany
20
Japan
Japan
United States
0
1981
1991
2001
United States
2011
1981
1991
Year
2011
Year
2001
Top 10% of patents ranked by year using BCG Quality Index. The BCG Quality
Index is made up of 3 components; specifically, forward citations of a patent in
newer patents adjusted for the patents age, the number of patent claims, and
the strength of a patents backward citations. The components and
corresponding weights used by the quality index are a product of proprietary
Boston Consultng Group research.
Figure 13. Medical Research Articles and Citations by Selected Countries/Regions, 2000-2010
400 000
300 000
2000
2009
Otherb
49 946
63 483
Other Asiac
10 029
20 790
8.4
3937
18 399
18.7
No.
China
100 000
2.7
26 755
21 477
2.4
114 970
120 421
0.5
United States
116 156
122 659
0.6
Overall
321 795
367 229
1.5
2000
2000-2009
European Uniond
Japan
200 000
Annual
Growth Rate, % a
No. of Medical
Research Articles
2009
Year
No.
2000
Citation Index
of Highly
Cited Articles
Compound
Annual Growth
Rate (Citation
Index), % a
2000
2010
2000-2009
763
1034
0.57
0.59
0.4
Other Asiac
20
113
0.1
0.22
8.6
8000
China
16
82
0.22
0.22
0.3
6000
Japan
345
294
0.5
0.45
1.0
4000
European Uniond
2079
2936
0.68
0.86
2.5
United States
5402
5729
1.67
1.63
0.2
Overall
8626
10 189
NA
NA
NA
12 000
Otherb
10 000
2000
0
2000
2010
2009
Year
NA indicates not available. Medical research was defined as the life sciences and
psychology, excluding agricultural science. Article counts reported by the
National Science Foundation were from the Thomas Reuters Science Citation
Index and Social Science Citation Index,18 classified by year of publication and
assigned to countries on the basis of institutional addresses listed on each
article. Articles were counted on a fractional basis; ie, for articles with
collaborating institutions from multiple countries, each country received
fractional credit on the basis of proportion of its participating institutions.
Citations were based on a 3-year period with 2-year lag; eg, citations for 2000
are references made in articles in 2000 to articles published in 1996-1998. The
citation index of highly cited articles was defined as the share of the worlds top
184
Compound annual growth rate (CAGR) supposing that year A is x and year B is
y, CAGR = (y/x){1/(BA)}1.
Other includes the remaining 159 nations of the world within the original
database.
jama.com
Figure 14. Market Performance of Publicly Traded Life Sciences and Health Care Companies, 2003-2013
NYSE Arca Biotechnology Index
Health insurance
Standard & Poor 500 index
Dow Jones US health care index
Dow Jones US medical equipment index
NYSE Arca Pharmaceutical Index
Hospital chains
700
600
500
400
300
200
100
0
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Year
analysis. This trend will likely continue. A combination of the limited recent record of industry research and development and the unpredictability of outcomes and length of time required to observe
results produces uncertain returns on investment, which are not tolerated in an economy that values short-term performance disproportionally.
Therefore, altogether new funding sources are required. As we
and others have proposed previously,25,26 a variety of new financing vehicles are feasible and attractive. These might include
Foreign capital repatriation, with new tax provisions that allow companies to return funds held outside the United States if used for
research.27 Because of the size of these holdings, a yearly increment of 25% to 50% of total research funding is feasible over the
next decade.
Biomedical research bonds, analogous to those used to finance
sports stadiums and airports, could be issued by federal, state, or
local governments, with amortization from patent royalties or converted to equity in new companies created. Historically, bonds have
funded infrastructure investments but are now being adapted for
environmental and green projects, which have economics that
resemble medical research.28
Research innovation trusts could be formed to allow private and
public entities to join forces for innovation in high-priority diseases or those of high public health importance, in return for tax
credits (not deductions), as have been used previously to preserve land, create parks, and build factories. These new trusts can
be structured to permit investments by public and company pension funds or individual retirement programs, which are currently
precluded from most early-stage, speculative investments. Australia, Canada, and the United Kingdom have used such research
and development trusts effectively.
Tax checkoffs, whereby individuals can specify a portion of their
tax payment to be diverted to research, as is currently the practice for public funding of elections. A few states (eg, California,
Maryland, New York, and Oregon) have made science a priority
using tax checkoffs.
Each of these financial innovations could be invoked without direct federal or state funding. They potentially can mobilize new private sources of funds without requiring tax increases or direct public appropriations.
jama.com
Productivity
New science and technology have been slow to address the morbidity and cost of chronic diseases and the growing number of elderly persons. Consequently, some have suggested that changing
patient behavior and education (for adherence and lifestyle modification), not technology, should become the priority.29 Others focus on changing the NIH mission to emphasize prevention and clinical evaluation rather than basic scientific discovery, or altering
incentives of industry to encourage their investment in highprevalence, high-cost conditions rather than lucrative niches such
as cancer and orphan diseases.8,9,12 Some observers have even suggested that expectations for science and technology be reduced,
given the long cycle time from discovery to clinical application.30
Declining productivity is at the root of many of these dissatisfactions. Therefore, greater attention is required to introduce methods that enhance the pace of research with few additional costs.
Improve the scientific process. As our analysis confirms, research
is costly, capital intensive, and, above all, collaborative. Moreover, researchers depend on one another for a source of new ideas,
as well as access to material, reagents, clinical information, samples,
and ultimately patients who are willing to participate in clinical trials.
Therefore, recent efforts have been aimed at facilitating those critical interactions. An example is the Accelerating Medicines
Partnership31 (between companies and the NIH), which identified common diseases for which few effective therapies exist but
science is especially promising. Four conditions meeting those criteria were selected: Alzheimer disease, type 2 diabetes, rheumatoid arthritis, and systemic lupus erythematosus. In each, specific
biological questions were identified that can best be answered using
resources from industry, the NIH, and academic investigators combined, who would otherwise be limited if working on their own.
Enhance benefits of large-scale, industrialized biology and smallscale investigators laboratories. The past 2 decades investment
in large projects, such as sequencing the human genome and its
successors for proteins, the microbiome, or the nervous system
connectome, is unlikely to realize its full value without interpretation and application by skilled individual scientists. Many astute observers suggest that the desirable balance has not yet been struck
between industrial-scale and individual-inspired laboratories.21 Specifically, further experimentation with new organizational models
(Reprinted) JAMA January 13, 2015 Volume 313, Number 2
185
should be aimed at obviating current limitations of the existing balkanized corporate, venture capital, NIH, and university practices.
Examples of new models are the Broad Institute in Cambridge, Massachusetts (genomics), BioDesign Institute at Arizona State University, Tempe (biomedical engineering), and Allen Institute for
Brain Sciences, Seattle, Washington (neurological and psychiatric
disease). Each of these seeks to optimize individual and institutional contributions while ensuring funding. Each orchestrates external relationships.
jama.com
tral Europe, and India), raise barriers to the diffusion of clinical innovations between countries.
Two areas are of particular concern: erosion of the publics support for science in the United States and hesitancy to reform the patent system.
Public Opinion
Recent polls show erosion of public support for biomedical research compared with other priorities. Support has declined steadily
since 2000 and is now well behind concerns about the economy,
domestic security, immigration, crime, and the US role in international affairs.46,47 The trend is not confined to the United States but
is also evident in Europe. Despite the demonstrable successes of earlier decades, the primacy of science as the source of improved health
is today questioned because of the convergence of several forces.
First, despite bold promises, advances visible to the public have
been less frequent because solutions to many conditions like autism, Alzheimer disease, and most cancers remain elusive, with neither effective prevention nor treatment, despite intensive research. Second, drug discovery has proven more difficult and less
predictable than many had expected, with a decline over the past 2
decades in altogether new classes of drugs, new registrations, and
drugs in clinical trials. Third, the economics of medical advances are
being scrutinized as a source of added insurance cost, with growing pressure to justify clinical value using objective criteria, formal
tools of technology assessment, and consideration of quality-oflife measures separately from those that affect mortality. Some technology skeptics have even urged that the United States take a technology holiday for a decade, suggesting that the money saved be
spent on ensuring that everyone receives existing preventive and
therapeutic means, even if this slows scientific discovery.48
Such tensions are perhaps inevitable, given the high cost and
poor performance of US health care as judged by international mortality comparisons. Skepticism of medical research is evident in recent US budget discussions, which have favored the physical sciences as faster, reliable, and more predictable routes to US
competitiveness than the uncertainties of medicine. Also, medical
devices and new manufacturing practices for large-molecule biopharmaceuticals are heavily driven by engineering advances, which
in turn depend more on the physical sciences and less on the biological sciences. These trends imply that pressure will mount to divert resources away from challenging but high-potential avenues in
biology.
Conclusions
As this analysis demonstrates, at the same time support for biomedical research in the United States has wavered, global interest
in biomedical research is increasing.49 Asia and Europe are now on
par with the United States in the relative number of researchers, and
Asia, especially China, is making rapid gains in life science patents
and highly cited publications. Although the United States is far from
losing its preeminent role in biomedical research, similar historical
changes have occurred in other industries (eg, electronics, automobiles, industrial manufacturing) that over time reshaped the countrys competitiveness. Many in the United States applaud the new
interest in other countries as a reflection of the truly international
reach of science, since discoveriesmade anywherecan be applied here. This optimistic view neglects the strong barriers crejama.com
187
ARTICLE INFORMATION
Author Contributions: Dr Moses had full access to
all of the data in the study and takes responsibility
for the integrity of the data and the accuracy of the
data analysis. Dr George and Mr Palisch contributed
equally.
Study concept and design: Moses, Matheson,
Cairns-Smith, Dorsey.
Acquisition, analysis, or interpretation of data: All
authors.
Drafting of the manuscript: Moses, Matheson,
Cairns-Smith, George, Palisch.
Critical revision of the manuscript for important
intellectual content: All authors.
Statistical analysis: Moses, George, Palisch.
Administrative, technical, or material support:
Moses, Matheson, Cairns-Smith, Dorsey.
Study supervision: Moses, Matheson, Cairns-Smith,
Dorsey.
Conflict of Interest Disclosures: All authors have
completed and submitted the ICMJE Form for
Disclosure of Potential Conflicts of Interest. Dr
Moses reports membership in a variety of
foundation and company boards in health care and
financial services. Dr Moses, Messrs Matheson and
Palisch, and Dr Cairns-Smith report providing
management consulting services to hospital
systems, insurers, foundations, and
pharmaceutical, device, and IT companies. Dr
Dorsey reports consultancy for Amgen, Avid
Radiopharmaceuticals, Clintrex, Lundbeck,
Medtronic, the National Institute of Neurological
Disorders and Stroke, and Transparency Life
Sciences; a filed patent related to telemedicine and
neurology; and stock/stock options in Grand
Rounds (a second opinion service). No other
disclosures were reported.
REFERENCES
1. Moses H III, Matheson DHM, Dorsey ER, George
BP, Sadoff D, Yoshimura S. The anatomy of health
care in the United States. JAMA. 2013;310(18):19471963.
2. Moses H III, Dorsey ER, Matheson DH, Thier SO.
Financial anatomy of biomedical research. JAMA.
2005;294(11):1333-1342.
188
The analysis underscores the need for the United States to find
new sources to support medical research, if the clinical value of its
past science investment and opportunities to improve care are to
be fully realized. Substantial new private resources are feasible,
though public funding can play a greater role. Both will require nontraditional approaches if they are to be politically and economically
realistic. Given global trends, the United States will relinquish its historical innovation lead in the next decade unless such measures are
undertaken.
jama.com
//www.people-press.org/2009/07/09/public
-praises-science-scientists-fault-public-media/.
Accessed April 15, 2014.
jama.com
189