Transcript thomson reuters presentation template

Personalized Medicine: A topical analysis using
Thomson Reuters data and analytics
MLA annual meeting
May, 2010
Ann Kushmerick,
Manager, Research Evaluation and Bibliometric Data
What is personalized medicine?
“A form of medicine that uses information about a person’s
genes, proteins, and environment to prevent, diagnose, and
treat disease.”
-National Cancer Institute
Information about a patient's protein, gene or
metabolite profile could be used to tailor medical care
Individual’s information is being used:
–to stratify disease status
–select from among different medications
–tailor dosages
–provide a specific therapy
–initiate a preventative measure
2
U.S. to Compare Medical Treatments
February 15, 2009
The $787 billion economic stimulus bill approved by
Congress will, for the first time, provide substantial amounts
of money for the federal government to compare the
effectiveness of different treatments for the same illness.
- New York Times
UPDATE: Personalized Medicine Poised to Become Part of
US Healthcare Reform Plan
March 24, 2010
The healthcare reform bill includes a section on comparative
effectiveness research that creates the Patient-Centered
Outcomes Research Institute. Seen as a victory for
personalized medicine, the institute will study the utility of
medical products in "various subpopulations," including
groups differentiated by genetic and molecular subtypes.
- Genomeweb.com
3
How did we define personalized medicine for this
presentation?
Examples of keywords used:
• individualized health
• molecular profiling
• prospective care
• pharmacogenomic
• molecular diagnostic test*
• pharmacogenetic
• pharmacogenetic test*
• biomarker SAME gene
• protein expression profil*
• treatment individualization
• therapist/physician SAME
tailored
• personalized SAME therapy
4
What questions can Thomson Reuters data answer
about personalized medicine?
• Who are the highly cited institutions and researchers in this
field?
• What are the publication and citation trends?
• Who collaborates in this area?
• What countries have the most impact in personalized
medicine?
• What are the key journal articles/patents/book
chapters/conference proceedings on this topic?
5
Thomson Reuters: Solutions for
the entire research cycle
•Researcher ID
•Thomson Pharma
•Thomson Innovation
Manage
Research
Output
Search &
Quality
Discover
•Web of Science
•BIOSIS Previews
•BIOSIS Citation Index
•Zoological Record
•Derwent Innovations Index
•INSPEC
•CAB Abstracts and Global Health
•MEDLINE
Research Cycle
Evaluate
outcomes
•InCites
•Journal Citation Reports
•Custom data
•Web Services
Write &
Publish
•EndNote
•EndNote Web
•Scholar One
6
Thomson Reuters Web of Knowledge:
a comprehensive, biomedical discovery tool
Web of Science
- 10,000 Journals, Natural and Social Sciences
Coverage back to 1900
- 12,000 Conference Proceedings Annually
MEDLINE
5,180 Journals
BIOSIS Previews and
BIOSIS Citation Index
- 5,250 Journals
- Thousands of Meetings,
Books, Patents Annually
7
Personalized Medicine data in Web of Knowledge
MEDLINE records
8
View same article in BIOSIS for specific biological
information
9
Analyze times cited information in Web of Science
10
Use Web of Knowledge analytical tools like citation
map
Institutions that cited this paper
11
Who is funding personalized medicine research?
Number of papers by funding body
acknowledged: Web of Science
12
BIOSIS Citation Index
 A citation resource providing cited references for BIOSIS
Previews content:
– Cited references for BIOSIS unique items (beginning with
2006 production year data)
– Cited references for BIOSIS items that overlap with items
in the Web of Science (for all years, 1900 - forward)
– All citation data is contained within the BIOSIS database
 Unique BIOSIS times-cited count
 Available only on the Web of Knowledge platform
13
BIOSIS Citation Index
• BIOSIS database has been known for decades for
extensive coverage of biomedicine:
• Journals and Serials-over 5,000 titles
• Meetings, Conference
• Reviews of books, software
• Books, book chapters
• U.S. patents
• And unique, in-depth indexing:
Major Concepts
Concept Code(s)
Taxonomic Data
Disease Data
Chemical Data
Gene Name Data
Sequence Data
Geographic Data
Geologic Time Data
Methods and Equipment Data
Parts & Structure Data
Miscellaneous Descriptors
14
…Combined with the power of citation indexing in
BIOSIS Citation Index
As with Web of Science, BIOSIS Citation Index
includes a true cited reference index revealing
exactly what has been cited, and allowing one
to browse and select the exact cited items on
which to search.
15
BIOSIS Citation Index: Variety of research materials
on personalized medicine
16
Explore citation relationships in BIOSIS Citation Index
This book chapter was
cited 31 times across
Web of Knowledge
This book chapter
was cited 24 times
within BIOSIS
Citation Index
17
BIOSIS Citation Index- citation map visualization
Use citation report visualization
Personalized medicine:
concept code=
psychiatry-addiction alcohol,
drugs, smoking
19
Precise searching for relevant biomedical research
• Chemical=gramicidin S
• Major concept=(Pharmaceuticals)
• Concept Code=(Blood - Blood and lymph studies)
• Taxonomic data=(Hominidae)
• Literature Type=(Meeting Paper)
20
Research Analytics: tools like InCites provide
comprehensive citation metrics and analytics
• InCites is a web-based research evaluation tool designed
1.
31
to track trends on the field, country, and institution level,
and to enable detailed bibliometric analysis of the set of
papers important to you.
Types of citation metrics and what they measure
Productivity
Total
influence
# papers
Journal actual/expected
citation rate
# citations
H-index
Category actual/
expected citation rate
Avg. citation rate
Efficiency
Relative Impact/
Benchmarking
Percent of papers
cited
Can be applied to an
institution, a researcher,
a research group, etc.
Percentile in category
and mean percentile
% papers in top 10% of
their field
% papers in top 1% of
their field
Aggregated
Performance Indicator
Specialization
Disciplinarity index
Interdisciplinarity index
22
Personalized medicine: trend in publications and
citations
70-fold increase in papers using this
terminology over the 30 year period
3,500
50,000
40,000
2,500
2,000
30,000
1,500
20,000
1,000
10,000
500
Citations
Papers
0
0
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
papers
3,000
Citations increased from 723 in 1990 to
57,127 in 2009
60,000
citations
4,000
23
Research Analytics: metrics from InCites
Most highly-cited paper in personalized medicine:
• vant Veer LJ, Dai HY, van de Vijver MJ, et al.
Gene expression profiling predicts clinical outcome of breast
cancer. Nature (415), 6871. 530-536. Jan 31, 2002.
Citation metrics
Times Cited:
2,784
Journal performance ratio:
16.06 (cited over 16 times the expected
rate for the journal)
Category performance ratio:
24.96 (cited almost 25 times the
expected rate for the field of oncology)
Percentile:
Top 0.005% of oncology papers from
2002
24
Highly Cited Institutions: Personalized medicine
# citations received
Mayo Clinic & Mayo Foundation
10,350
Harvard University
10,251
National Cancer Institute
7,928
University of Washington
7,236
University of California San Francisco
6,917
Source: Web of Science and InCites
25
Top 10 countries publishing in personalized medicine
Netherlands
4% Japan
France
5%
Canada
5%
Italy
6%
Spain
3%
Australia
3%
5%
USA
48%
England
10%
Germany
11%
Source: Web of Science and InCites
26
Changing collaboration partners
2000-2004
Coauthored
Coauthored
Papers
Country
Papers
Country
2005-2009
Country
Country
Coauthored
Coauthored
Papers
Country
Papers
Country
Country
Country
83Canada
83Canada
USA
USA
210 Canada
210Canada
USA
USA
74Germany
74Germany
USA
USA
179 England
179England
USA
USA
73England
73England
USA
USA
160 Germany
160Germany
USA
USA
47Italy
47Italy
USA
USA
122 Italy
122Italy
USA
USA
34England
34England
Germany
Germany
112Netherlands
112 Netherlands
USA
USA
34France
34France
USA
USA
77Germany
77 Germany
30France
30France
Germany
Germany
77Peoples
China
USA
77 PeoplesR R
China USA
29Switzerland
29Switzerland
USA
USA
74England
74 England
Germany
Germany
28Netherlands
28Netherlands
USA
USA
74Germany
74 Germany
Switzerland
Switzerland
25England
25England
France
France
72France
72 France
USA
USA
24Australia
24Australia
USA
USA
69 Australia
69Australia
USA
USA
24England
24England
Italy
Italy
68 Japan
68Japan
USA
USA
24Japan
24Japan
USA
USA
64 Spain
64Spain
USA
USA
Netherlands
Netherlands
Pharmaceutical companies publishing in
personalized medicine
60
50
40
30
56
51
20
10
45
43
41
41
number of papers
average citation rate
25.61
17.51
16.98
12.26
13.98
13.39
0
28
Drill down to Merck papers
What topics is Merck researching?
Which are highly cited?
# citations
With whom is Merck
collaborating?
keyword
264 SURROGATE END-POINTS
# coauthored papers
231 BREAST-CANCER
187 ASSOCIATION
174 CARDIOVASCULAR-DISEASE
174 CORONARY-HEART-DISEASE
173 ALZHEIMERS-DISEASE
173 B-CELL LYMPHOMA
4
4
3
3
2
2
1
1
0
173 C-REACTIVE PROTEIN
173 IN-VIVO
173 MILD COGNITIVE IMPAIRMENT
POSITRON-EMISSION173 TOMOGRAPHY
29
168 GENE-EXPRESSION PROFILES
Most highly cited author in this topical dataset
Richard M. Weinshilboum, M.D.
Mayo Clinic
http://mayoresearch.mayo.edu/mayo/research/staff/weinshilboum_rm.cfm
30
Dr. Weinshilboum’s most cited paper
Human catechol-O-methyltransferase pharmacogenetics:
Description of a functional polymorphism and its potential
application to neuropsychiatric disorders. Pharmacogenetics.
Vol 6. 1996. (Article) Times cited: 573
Journal Expected citations: 61.78
Average citations specific to journal, year of article, and
document type.
Baseline for journal performance.
Weinshilboum’s ratio of actual cites to expected: 9.27
(573/61.78)
31
Dr. Weinshilboum’s most cited paper
Human catechol-O-methyltransferase pharmacogenetics:
Description of a functional polymorphism and its potential
application to neuropsychiatric disorders. Pharmacogenetics.
Vol 6. 1996. (Article) Times cited: 573
Category : pharmacology and pharmacy
Category Expected citations: 23.31
Average citations specific to category, year of article, and
document type.
Baseline for category performance.
Weinshilboum’s ratio of actual cites to expected: 23.04
(537/23.31)
32
Dr. Weinshilboum’s most cited paper
Human catechol-O-methyltransferase pharmacogenetics:
Description of a functional polymorphism and its potential
application to neuropsychiatric disorders. Pharmacogenetics.
Vol 6. 1996. (Article) Times cited: 573
Category : pharmacology and pharmacy
Percentile: 0.028%
Percentile position specific to field and year. The closer to
zero, the more highly cited.
Weinshilboum’s paper falls into the top 0.028% of all
pharmacology/pharmacy papers from 1996.
33
Dr. Weinshilboum’s most cited paper
Human catechol-O-methyltransferase pharmacogenetics:
Description of a functional polymorphism and its potential
application to neuropsychiatric disorders. Pharmacogenetics.
Vol 6. 1996. (Article)
Times cited: 573
34
Weinshilboum, RM-summary metrics
Productivity # papers
112
Total
influence
# citations
4,086
H-index
33
Efficiency
Relative
Impact/
Benchmarking
Avg.
36.48
citation rate
Percent of
papers
cited
81.25
Specialization
Journal
actual/expected
1.49
Category
actual/expected
2.48
Mean percentile
27.34
% papers in top 10%
of their field
37%
% papers in top 1% of
their field
8%
Disciplinarity index
0.22
Weinshilboum’s papers perform 49% above their journals’ normal
citation rates and 148% above their categories’ normal citation
rates.
His papers rank in the top 27% of their fields, according to
citation count.
35
Websites for further information:
BIOSIS Citation Index
InCites
Research Analytics
Bibliometrics white papers
Using Bibliometrics: A Guide to Evaluating Research Performance
Using bibliometrics in evaluating research