1 Value Creation and Value Capture in Open Innovation Master Thesis MSc ETH MTEC Olivier Esseiva Tutored by Andreas Schneider May 2013 Chair of Strategic Management and Innovation Prof. Georg von Krogh Department Management, Technology and Economics, ETH Zürich
78
Embed
Value Creation and Value Capture in Open Innovation · 2014-09-29 · Open Innovation Value Capture in Open Innovation Linking Value Creation and Value Capture in Open Innovation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
1
Value Creation and Value Capture in Open Innovation
Master Thesis MSc ETH MTEC
Olivier Esseiva
Tutored by Andreas Schneider
May 2013
Chair of Strategic Management and Innovation
Prof. Georg von Krogh
Department Management, Technology and Economics, ETH Zürich
2
Acknowledgements
I would like to thank Andreas Schneider for his kind supervision of this thesis. The discussions, inputs,
and feedback from him were extremely helpful. Also I would like to thank all the other group members
of the Chair of Strategic Management for their support and the fruitful discussions with them. Namely
Zeynep Erden and Renato Sydler were of great help to me. Special thanks go to Fotini Pachidou for
showing me how to access Bloomberg data.
I would like to thank David Klang and Maria Wallnöfer for the helpful discussions on business models and
their view on this topic.
3
Abstract
This master thesis investigates what strategies companies pursue to profit from Open Innovation and
how value creation mechanisms interact with value capture. This thesis is unique in the current
literature, since it investigates aspects of both value creation and value capture in Open Innovation and
provides empirical ramification of collaborative innovation strategies that have gained widespread
acceptance in recent years. To answer the research question, the patenting and publishing activity
during the last ten years of the ten largest pharmaceutical companies is investigated. The data-mining
analysis reveals that Open Innovation practices change the appropriability regime by diminishing the
importance of the patent portfolio. By engaging in interfirm alliances, firms can increase R&D
productivity. The main drivers of openness in the value creation process are the need to fill knowledge
gaps, technological demand, and strategic decisions to boost certain business areas.
4
List of Figures
Figure 1: Visual representation how theory section is built up .................................................................. 10
Figure 2: The Closed Innovation model. ..................................................................................................... 11
Figure 3: Open Innovation model ................................................................................................................ 12
Figure 4: Possible value distribution of an innovation among different stakeholders ............................... 16
Figure 5: Profiting from Innovation framework .......................................................................................... 17
Figure 6: Distinction between generic, specialized and co-specialized complementary assets according to
their mutual dependence ................................................................................................................. 18
Figure 7: From innovation strategy to financial performance. .................................................................. 22
Figure 8: Framework for determining openness based on value creation and value capture drivers. ..... 23
Figure 9: Evolution of total number and openness in publications from 2002 to 2012. ........................... 34
Figure 10: Type of co-authorship in published papers ................................................................................ 34
Figure 11: Annual growth of R&D intensity (R&D expenditure divided by total sales) and annual growth of
openness from 2002 to 2012. ........................................................................................................... 36
Figure 12: Share of inter-firm alliances and and openness from 2002 to 2012. ......................................... 37
Figure 13: Number of patents and average openness of patents of the top10 pharmaceutical companies
from 2002 to 2012. ........................................................................................................................... 39
Figure 14: Type of collaboration in publications and patents from 2002 to 2012. ..................................... 39
Figure 15: Annual change of patent share of different disease areas against the change in openness of
List of Figures ................................................................................................................................................. 4
List of Tables .................................................................................................................................................. 5
1.2 Structure of Thesis ............................................................................................................................... 9
2. Theory ...................................................................................................................................................... 10
2.1 Open Innovation ................................................................................................................................ 10
2.1.1 Conceptual Framework of Open Innovation .............................................................................. 10
2.1.2 Open Innovation and Business Models ...................................................................................... 13
2.2 Value Capture from Innovation ......................................................................................................... 15
2.2.1 Value Capture from Innovation in General ................................................................................ 15
2.2.2 Value Capture Aspects of Open Innovation ............................................................................... 18
2.3 Interplay of Openness, Value Creation, and Value Capture .............................................................. 21
2.3.1 Openness and Innovative Performance ..................................................................................... 21
2.3.2 Drivers of Openness in Value Creation and Value Capture ........................................................ 22
2.4 Research Gap and Research Question .............................................................................................. 23
3.1 Research approach ............................................................................................................................ 25
3.2 Study sample ..................................................................................................................................... 25
3.2.1 Pharmaceutical Industry ............................................................................................................. 25
3.2.2 Ten largest companies ................................................................................................................ 27
3.2.3 Using Publications and Patents to Study Value Creation and Value Capture ............................ 27
3.3 Extracting relevant information from dataset .................................................................................. 28
7
3.3.1 Extracting company affiliation .................................................................................................... 28
3.3.2 Extracting disease area ............................................................................................................... 29
3.3.3 R&D Intensity data ..................................................................................................................... 32
7.7 Summary of Value Capture Data ....................................................................................................... 78
8
1. Introduction
1.1 Motivation
Innovation is almost any company’s bread and butter nowadays if they want to achieve their ambitious
growth targets. Decades ago, innovators faced a completely different economic environment and large
corporate research labs such as AT&T’s Bell Labs and Xerox’s PARC flourished. Today, the innovator faces
ever shorter product life cycles, global competitive threats, and customers demanding more and more
while paying less and less.
How can a company simultaneously bring new ideas to market while making sure that it can capture at
least part of the value that is created? Years ago, these questions would have appeared downright
obvious to many companies and hardly any scholar or practitioner would have bothered to answer these
questions in great detail. Nowadays and especially through the advent of modern information and
communication technologies, it is not that easy anymore to defend an innovated product or process
against imitators or competitors. Put differently, making the pie bigger does not automatically lead to a
larger slice of the pie.
In response to the dramatic change in the business landscape, several innovation strategies have
emerged to tackle today’s challenges. Notably, it has been suggested that organizations need to tap into
sources of knowledge that lie outside the traditional boundary of the company and that the highly
inward focusing innovation practices of many companies lie at the heart of an insufficient innovation
performance (Chesbrough, 2003; Von Hippel & Von Krogh, 2003). Concepts that were previously known
only to a few experts of some isolated industries such as Open Innovation and Open-Source, have
become household words in today’s management language. At the heart of modern innovation
strategies lies the idea that companies need to become more open and need to engage with their
external environment because labor mobility has increased, venture capital is abundant, knowledge is
widely dispersed, and product life cycles have contracted dramatically (Chesbrough, 2003). Or simply
put: “Not every smart person works for us. We need to engage with people inside and outside the
company” (Chesbrough, 2003).
Prompted by the internet and with new strategic concepts at hand, many companies have indeed
become more open in the way how they innovate (Enkel, Gassmann, & Chesbrough, 2009). Companies
engage in venturing activities, contract research, license in and out intellectual property, involve
customers in their innovation process, participate in research consortia, form public-private partnerships
9
and engage in many more activities that could be termed open (Chesbrough, 2012). However, being
open is not the answer to everything and psychological and organizational barriers impede companies to
become even more open than they are today. Some of these barriers are behavioral in nature such as
the not-invented-here syndrome, which describes the resistance that R&D departments often show
towards outside ideas (Katz & Allen, 1982). But beyond these psychological factors, the main limitation
toward openness is the more strategic question on how an innovator can profit from its innovation
(Teece, 1986). While earlier work that urged companies to become more open focused on the aspect of
value creation, later work has put the emphasis on value capture (Fischer & Henkel, 2012). However,
there is no systematic study that has investigated value creation and value capture aspects of Open
Innovation concomitantly.
This master thesis addresses the gap that has been left by the literature on Open Innovation, value
capture, and value creation. Specifically, this thesis explores the interplay between value creation and
value capture in Open Innovation settings. Since no established theory exists for this specific question,
the thesis will be rather explorative in nature and will focus on developing suggestions rather than
formally testing hypothesis. Therefore, the research question is an open question with the goal of
formulating more specific and testable hypotheses.
1.2 Structure of Thesis
This thesis begins by reviewing the existing literature on Open Innovation, focusing on value creation and
value capture aspects of innovation. The theory section is built of three parts. First it reviews the work on
value creation on Open Innovation which comprises the original work of Henry Chesbrough on this topic.
Second, it discusses the literature on value capture aspects. Third, it investigates the existing literature
on the connection between value creation and value capture in Open Innovation settings. Figure 1
depicts how the theory section is built up.
10
Value Creation in Open Innovation
Value Capture in Open Innovation
Linking Value Creation and
Value Capture in Open Innovation
Research Gap
Research Question
Figure 1: Visual representation how theory section is built up
The theory section leads to the formulation of a research question. The next chapter describes the
methodology of this thesis. Since the research topic is very explorative, new methods for collecting,
analyzing, and interpreting the data were developed. These methods include a new procedure to
measure the openness of a publication and a patent. After this, the result section highlights the most
important results, visualizes, describes and interprets key points. Following the result section, the
discussion session shows the contribution to the existing literature, formulates hypotheses, highlights
managerial implications, discusses the limitations of this study and shows avenues for future research.
2. Theory
2.1 Open Innovation
2.1.1 Conceptual Framework of Open Innovation
In his book “Open Innovation”, Henry Chesbrough argued that companies should become more open to
external knowledge and ideas (Chesbrough, 2003). He distinguished between a Closed Innovation model
and an Open Innovation Model. The Closed Innovation Model is depicted in Figure 2. The innovation
process is described via a funnel structure; ideas originate on the left hand side (research part), some of
these ideas are pursued further while some are discarded (development part) and a few ideas are
subsequently brought to market (narrow part of the funnel). Every step of the innovation process is
conducted within the confines of the company that originated the idea. Therefore, the closed innovation
paradigm represents a fully vertically integrated model of innovation (Gassmann, Enkel, & Chesbrough,
11
2010). Under the premises of Closed Innovation, there are no “exit mechanisms” for knowledge and
ideas that would allow them to be pursued by a different firm. As a consequence, the ideas that
originate from the company’s R&D are either brought to market or are put on the shelf. Furthermore,
there are no mechanisms for ideas to flow from the outside world into the firm’s R&D process. The
innovating company relies solely on its own internal knowledge base to create new ideas and its
development capabilities to bring them to market. The firm’s boundaries are impermeable and are,
therefore, depicted as solid lines in Figure 2. Classical examples of companies that relied on the Closed
Innovation model were AT&T (with the famous Bell Labs), Xerox, and IBM in the 1970ies and 1980ies
that produced outstanding research, but often failed to exploit their ideas commercially (Chesbrough,
2003)
Internal technology
base
Externaltechnology
base
The Market
= Ideas
R DResearch Development
New Products / Services
Figure 2: The Closed Innovation model. Ideas and knowledge originate at the left-hand side, go through the development funnel and are brought to market. Source: (Chesbrough, 2003)
In contrast to the Closed Innovation paradigm, Figure 3 shows the Open Innovation Model. There are
two significant changes to the Closed Innovation model. First, ideas that were spawned within the
internal technology base do not necessarily stay within the company. As an alternative to being
developed and commercialized by the inventing company, ideas can flow out a company’s boundary at
any stage and could be pursued by another company (Inside-Out) (Chesbrough, 2003). Possible inside-
out mechanisms are spin-offs, out-licensing, free revealing of ideas, etc. The second difference to the
Closed Innovation model is that externally conceived and developed ideas are allowed to flow into the
12
company for commercialization (Outside-In) (Chesbrough, 2003). Acquisitions and in-licensing, for
example, are outside-in mechanisms. These two fundamental mechanisms, Inside-Out and Outside-In,
can be pursued concomitantly, for example by cross-licensing technologies, joint ventures, private-public
partnerships, strategic alliances, co-marketing arrangements and so forth (Chesbrough, 2012). Since
ideas can flow quite freely between a firm’s R&D and the external environment, the boundaries are no
longer depicted as solid lines, but as dotted lines that allow passing of ideas and knowledge. The Open
Innovation paradigm can be summarized as “… the use of purposive inflows and outflows of knowledge
to accelerate internal innovation, and expand the markets for external use of innovation, respectively”
(Chesbrough & Schwartz, 2007).
Internal technology
base
Externaltechnology
base
DifferentMarket
Market
Inside-Out (Spin-offs,
out-licensing
Outside-in (Acquisitions, in-licensing)
= Ideas
R DResearch Development
Figure 3: Open Innovation model. Source: (Chesbrough, 2003)
The Open Innovation model is inherently focused on creating value through externalizing unused
technologies and ideas and by internalizing outside useful knowledge. However, the aspect of capturing
value and making a profit out of these innovations has received less focus (Fischer & Henkel, 2012). In
order to explore the question of value capture in Open Innovation, the next section will explore the
concept of business models which is a statement of how a company intends to capture value. After the
discussion of business models, the next chapter discusses the literature on value capture. Then, we
explore how the literature on value capture applies to Open Innovation.
13
2.1.2 Open Innovation and Business Models
In order to successfully implement an Open Innovation strategy, a company needs to have a clearly
defined business model. The term “Business Model” is a relatively recent concept that emerged in
strategic and entrepreneurial literature, but no generally accepted definition has yet been established
(for a comprehensive review on business models see: (Zott, Amit, & Massa, 2011)) Chesbrough defines
the business model as a framework that links technology and ideas to economic outcomes and, thereby,
encompasses two important functions: Value Creation and Value Capture (Chesbrough, 2006). The six
main aspects are depicted in Table 1.
Aspect of Business Model Authors mentioning similar aspect in Business
Controlling variables:- Product/Service oriented- Firm size- Number of competitors
Val
ue
Cre
atio
n a
spe
cts
Val
ue
Cap
ture
asp
ect
s
Figure 8: Framework for determining openness based on value creation and value capture drivers. Adapted from Drechsler and Natter 2012.
This section ends the review on the current literature of value creation and value capture aspects of
Open Innovation. The next section summarizes the research gap and leads to the research question.
2.4 Research Gap and Research Question
Open Innovation is a fairly new concept and has seen a staggering growth in interest both from academia
and business (Chesbrough, 2012). However, most of the work has focused on questions of value creation
while value capture has by and large been neglected (Fischer & Henkel, 2012). Even though some studies
in recent years have tried to shed light on the issue of value capture in open innovation (Fischer &
24
Henkel, 2012; Van De Vrande et al., 2009), they do not provide a sufficient theoretical framework for
value creation and value capture in Open Innovation. First, they do not study aspects of value creation
and value capture concomitantly, but focus either on value creation or value capture. Second, they do
not explain why a firm would decide to be more open in one business area but not in the other even
though such behavior exists in reality (Alexy, George, & Salter, 2012; Henkel, 2006). Third, existing
literature does not provide enough differentiation between different forms of openness and why a firm
would engage in a particular form of Open Innovation activities such as joint-ventures, research
consortia, and so forth. Fourth, since many studies rely on cross-sectional surveys or decision games,
they lack a comprehensive view of the path-dependence and time evolution of Open Innovation
practices.
In order to give possible hints how these questions could be answered, an explorative research question
is stated:
What patterns of value creation and value capture can be observed in Open Innovation and how does
value creation and value capture interact in an Open Innovation setting?
25
3. Methodology
3.1 Research approach
To answer the research question, an explorative research method has been chosen. It has been noted in
the literature that research needs to achieve a fit between the level of maturity of the field and the
research methodology (Edmondson & McManus, 2007). As it was shown in the literature review, no fully
developed theory for value creation and value capture in Open Innovation exists at the moment. Given
the nascent state of the field, we cannot build hypothesis and confirm them using statistical methods
since the resulting correlations would not be backed up by a solid theory (Edmondson & McManus,
2007). Rather is it essential to develop detailed knowledge of the available data and to achieve intimacy
with the many layers of patterns that occur (Behrens & Yu, 2003). This approach is termed “Exploratory
Data Analysis” and was developed by John Tukey and is a method that tries to extract and interpret
patterns that occur in the data (Tukey, 1977). The goal of this method is to construct a terse,
mathematical description of a problem that eventually leads to hypothesis formulation (Behrens & Yu,
2003). Conversely, traditional confirmatory data analysis focuses on verification of testable hypothesis
using rigorous and well-defined statistical methods (Behrens & Yu, 2003).
This master thesis provides an explorative study of value creation and value capture aspects in Open
Innovation. The next section details the nature and scope of the data set and justifies the reason for this
choice.
3.2 Study sample
3.2.1 Pharmaceutical Industry
In order to understand the value creation and value capture aspects of Open Innovation, this study
analyses the publications and patents that were published and issued from 2002 to 2012 of the ten
largest pharmaceutical companies. The ten largest pharmaceutical companies are shown in Table 4. In
this section, it is first explained why the pharmaceutical industry has been chosen and in a second step
why the research was limited to the ten largest firms of the industry.
26
Company Sales Profits Assets Market Value
Pfizer 67.4 10.0 188.0 165.4
Novartis 58.6 9.1 117.5 150.4
Sanofi 43.2 7.4 125.3 103.3
Merck & Co 48.0 6.3 105.1 115.8
Roche Holding 45.3 10.0 62.9 152.0
GlaxoSmithKline 42.5 8.2 59.4 111.8
Abbott Laboratories 38.9 4.7 60.3 93.4
AstraZeneca 32.4 9.6 51.4 58.0
Eli Lilly & Co 24.3 4.3 33.7 46.6
Bristol-Myers Squibb 21.2 3.7 33.0 56.0
Table 4: The ten largest pharmaceutical companies based on a weighted average of sales, profits, assets and market value. Values in billion US Dollars and based on 2011 financial reports. Source: (Forbes, 2012)
The pharmaceutical industry has been chosen for several reasons. First, the pharmaceutical industry has
suffered in recent years from a declining R&D productivity (Garnier, 2008). The cost of developing a new
molecular entity, i.e. a drug containing an active ingredient that has not been approved previously, has
risen to 1.8 billion US dollars (Paul et al., 2010). Several authors have suggested more open approaches
to drug development (Munos, 2010; O’Hagan & Farkas, 2009; Tralau-Stewart et al., 2009). Second,
compared to other industries, patents in the pharmaceutical industry are seen as an effective means of
IP production (Teece, 1986). Hence, it provides a suitable proxy for value capture in research; especially
given the readily availability of patent filings online. Data on intellectual property protection such as
patent filings can easily be retrieved from public and private databases (USPTO Web Patent Database,
Thomson Reuters’ Derwent Innovations Index, etc.). Conversely, in industries where patenting activity is
spurious and heterogeneous across companies, one cannot rely on patent filings as a proxy for value
capture (Drechsler & Natter, 2012). Third, pharmaceutical companies are highly complex endeavors that
require the integration of many specialists and technologies. Therefore, it provides a rich data set with
many different forms of collaboration and institutional arrangements that can be studied. Fourth,
pharmaceutical companies possess a high amount of specialized complementary assets such as
investments in research facilities and human capital. Previous research has shown that great differences
in patenting and licensing activity exist between companies with no complementary assets such as start-
ups and companies with a lot of complementary assets (Arora & Ceccagnoli, 2006).
27
3.2.2 Focus on Large Companies
The pharmaceutical industry has been chosen because of its characteristics that make it interesting to
study Open Innovation. This study also limits its scope to the ten largest companies of the
pharmaceutical industry for several reasons. First, the exploratory nature of this study makes it more
difficult to control for variables such as firm size and fundamentally different patenting policies. The ten
largest companies of an industry constitute a rather homogeneous sample that allows to infer
conclusions more easily. Second, it is assumed that the publishing and patenting of large firms remains
fairly stable over time while a larger variance would be observed for smaller firms. In addition, due to a
high Mergers & Acquisitions activity, extreme variations could occur if two similar sized companies
merge and double their output of publications and patents basically overnight (LaMattina, 2011). Third,
by limiting the sample to the ten largest pharmaceutical companies, the size of the data set is more
manageable. Since all patents and publications over ten years are retrieved, a very large data set is
obtained, amounting to almost 1 gigabyte of data. For computational reasons, the data analysis process
is much easier if the data set does not exceed a certain size
3.2.3 Using Publications and Patents to Study Value Creation and Value Capture
In order to operationalize value creation and value capture, publications and patents are used as proxies.
Publications represent a contribution to the existing body of knowledge and they signify a better
understanding of the nature of human diseases. Even though the increase in knowledge cannot be
stated as value as such, it is obvious that knowledge will eventually create value by allowing the
development of better drugs and therapies. Therefore, publications are used as proxy for value creation
in this thesis. Patents on the other hand, represent value capture since they have an intrinsic economic
value and can be sold or licensed to other companies. It has been noted that patents can have a wide
range of uses from fending of patent trolls to revenue generating mechanisms (Cohen, Nelson, & Walsh,
2000). Patents provide a good proxy for value capture because they require not only being non-obvious,
but they must also have a commercial value for the patenting company. Else, they would not choose to
patent the innovation and pay the patenting fees.
In order to uncover the underlying evolution of value creation and value capture mechanisms in Open
Innovation, a time-frame of 10 years (from 2002 to 2012) was chosen. The publications were retrieved
from the ThomsonReuters WebOfScience database and the patents from ThomsonReuters
DerwentPatent Index. The search queries depicted in Table 5 were used to generate the data set. The
search queries were designed to include all publications and patents by the major legal entities of the
ten largest pharmaceutical companies.
28
Company Publications-Query Patents-Query #Publications #Patents
Pfizer AD=PFIZER AE=PFIZER 11280 2604
Novartis AD=NOVARTIS AE=NOVARTIS 9471 3597
Sanofi
AD=SANOFI OR AD=AVENTIS AE=SANOFI or AE=AVENTIS 4378 2988
Merck & Co AD=MERCK AE=MERCK 10833 5364
Roche Holding AD=ROCHE AE=ROCHE 6836 2772
GlaxoSmithKline AD=GLAXOSMITHKLINE or
AD=SMITHKLINE
AE=GLAXO or
AE=SMITHKLINE 10431 3285
Abbott Laboratories AD=ABBOTT AE=ABBOTT 4293 4154
AstraZeneca AD=ASTRAZENECA or
AD=ZENECA
AE=ASTRAZENECA or
AE=ZENECA 7535 2371
Eli Lilly & Co AD=ELI LILLY or AD = LILLY AE=ELI or AE=LILLY or
Table 5: Queries used in WebOfScience and Derwent Patent database to generate the data set. AD stands for “Address”, AE for “Assignee Name + Code” and AC for “Assignee Code” Mergers and name changes that happened during 2002 and 2012 or just before, such as the merger between Sanofi and Aventis in 2004, required to search for two entities instead of one in some cases.
Since we are interested in the authors’ affiliation, the query was performed on the authors’ address for
publications (Query code “AD”) and on the assignee (Query code “AE”) for patents. The query was
performed on the 24th of February 2013. After eliminating empty of ill-formatted records, a data set
containing 73840 publication records and 29620 patent records was obtained.
3.3 Extracting Relevant Information from Dataset
3.3.1 Extracting Company Affiliation
The first step in extracting relevant information from the data was to determine to which of the ten
largest pharmaceutical companies a publication or patent should be assigned. This was done by looking
for the company name in the address field in the publications record and the assignee in the patent
records. To accomplish this quickly for >100’000 records and to take into account slight name variations
(e.g. “Bristol Myers” and “Bristol-Myers”) a function was written in Visual Basic for Applications (VBA)
which is a scripting language within several Microsoft applications such as Microsoft Excel. The function
takes the address field and, optionally, the reprint author as an input and yields one or more company
names as an output. The VBA function is given in appendix 7.1.1.
29
3.3.2 Extracting Disease Area
The second step in getting useful information out of the data was to assign each patent and publication
to one or more therapeutic areas based on two newly developed procedures. Both publications and
patents are assigned to categories; publications in WebOfScience via the subject areas and patents via
the International Patent Classification (IPC) code. The subject area classification is a proprietary
classification algorithm by Thomson Reuters that assigns each article to one or more categories out of a
list of more than 250 categories (Thomson-Reuters, 2013). The IPC code is maintained by the World
Intellectual Property Organization (WIPO) and classifies each patent to a category according to the
technology to which it pertains (WIPO, 2013).
In order to have a consistent and comparable classification between the WebOfScience subject area and
the IPC patent classification, each relevant subject area has been linked to one or more IPC codes by
manually comparing the classifications and mapping them individually to each other. Out of the >250
subject area classifications, 34 categories have been initially deemed relevant for pharmacological and
medical research. These categories and the IPC codes which have been mapped to them are depicted in
Table 6. The most relevant IPC code for pharmaceutical purposes is A61P: “Specific therapeutic activity
of chemical compounds or medicinal preparations”. Almost all A61P categories were linked to at least
one subject area based on individual judgment which category would fit best. A detailed table is given in
the appendix.
Life Sciences & Biomedicine Corresponding IPC code(s)
Total Number 1'797 3'153 2'866 419 3'073 1'106 1'281 2'364 2'868 7'005 4'430 333 906 1'585
Table 8: Overlap matrix of publication data
Especially for patents, the initial overlap between certain categories was extremely high. This is probably
due to the fact that patents are usually filed with very broad claims, for example to block competitors
32
from patenting the same compound for a different indication. In order to reduce highly overlapping
categories (such as allergy and respiratory systems) a method by Agarwal and Searls to orthogonalize the
classification scheme was used (Agarwal & Searls, 2009). It consists of merging highly overlapping
categories (e.g. Allergy and Respiratory Systems) into one single category and to subtract some
categories from each other (e.g. changing the category from (Gastroenterology) to (Gastroenterology
BUT NOT hematology). The initial overlap matrices before orthogonalization are depicted in the
appendix.
3.3.3 R&D Intensity Data
Information on R&D intensity was retrieved from a Bloomberg terminal. The data extracted includes all
R&D intensities, i.e. total R&D expenditure divided by total sales, from 2002 to 2012. The figures for
2012 are preliminary figures since the annual reports for 2012 had not been published at the time of
extraction.
3.4 Calculating Openness
In order to calculate the degree of collaboration, it is important to distinguish between different forms of
collaboration. Collaboration can mean between individuals, departments, institutions, industries, or
geographic locations (Katz & Martin, 1997). In the context of Open Innovation where the boundaries of a
company are relevant, it makes sense to focus on inter-institutional collaboration. To measure inter-
institutional collaboration, co-authorship based on the author’s affiliation has been taken as a measure.
Co-authorship has been shown to be a reliable proxy for research collaboration in literature (Schummer,
2004).
To calculate openness, an “Openness measure” has been developed. The openness measure describes
the nature of collaboration that is embodied in a publication or patent. It takes the value of zero if
authors are only affiliated with one institution, hereby called “home institution”. If the article features
authors affiliated with three institutions (home institution, Univ A, and Univ B), the value would be 0.67.
The formula is detailed below.
( )
In our data set, the openness measure was calculated for the ten largest pharmaceutical companies. In
case a paper is written by authors affiliated with two or more of these companies, the openness score
was averaged (since the openness might be different between companies). The VBA function that
calculates the Openness measure is given in the appendix.
33
The same procedure was used for determining the openness of patents. Instead of the address the
assignee name was compared.
( )
4. Results and Findings
Based on the publishing and patenting activity of the ten largest pharmaceutical companies, the
following section presents the findings of an in-depth analysis of value creation and value capture in
Open Innovation. First, it investigates the value creation which is measured by the number and nature of
publications. Second, it analyzes how the ten largest pharmaceutical companies capture value which is
measured by the patents that have been issued from 2002 to 2012. Third, it offers a comprehensive
picture of value creation and value capture, and integrates and contrasts the patterns that occur in both
value creation and value capture.
4.1 Value Creation in Open Innovation
4.1.1 Industry-wide Evolution
The analysis of publication data from 2002 to 2012 reveals that the overall openness in value creation
has steadily increased. Figure 9 shows the increase of co-authored publications of all 10 pharmaceutical
companies combined. The openness increases significantly from below 50% in 2002 to 57% in 2012. The
increase in openness parallels the increase in total numbers of publications; hence the total number of
publications and the relative share of co-authored papers have increased in lockstep.
Figure 10 indicates the type of co-authorship in the publications from 2002 to 2012. One observes a
steady decrease of purely internal research papers (neither co-authored with other firms or public
institutions) and an increase in interfirm co-authorship while the percentage of inter-firm co-authored
papers has reached 10%. Still, most of the papers (around 70%) are co-authored with researchers from
public institutions.
Figure 9 and Figure 10 aggregates data of the ten largest pharmaceutical companies. The next section
digs deeper into these numbers and looks at firm-specific patterns.
34
Figure 9: Evolution of total number and openness in publications from 2002 to 2012. The blue line represents total number of publications and the red line the openness measure as defined in section 3.4.
Figure 10: Type of co-authorship in published papers. Orange bars indicate the percentage of interfirm co-authorship between the ten largest pharmaceutical companies. Blue bars indicate the percentage of papers that were authored purely by authors affiliated to one company. The grey line shows the R&D intensity (R&D expenditure divided by Total Sales)
Interfirm alliance Purely internal research R&D intensity
35
4.1.2 Firm-specific Patterns of Value Creation
Table 9 highlights the interfirm differences in openness, R%D intensity, and number of publications. The
openness ranges from 43.4% (Merck) to 64.7% (Roche). R&D intensity and the total number of
publications do not provide an explanation for the inter-firm differences.
Company Total Publications
R&D Intensity Openness
Roche 7750 18.8% 64.7%
Sanofi 4380 15.2% 59.0%
Eli Lilly 6600 20.2% 55.9%
Novartis 9471 16.3% 55.3%
Astrazeneca 7931 16.2% 54.0%
GlaxoSmithKline 10681 14.2% 52.2%
Pfizer 11272 15.1% 48.7%
Abbott 4312 9.7% 47.4%
Bristol Myers Squibb 4099 16.6% 44.1%
Merck 10835 17.2% 43.4%
Table 9: Total number of publications, average R&D intensity (R&D expenditure divided by Total Sales) and average openness of the ten largest pharmaceutical companies from 2002 to 2012 as defined in section 3.4
Figure 11 shows the annual changes in R&D intensity and openness in the timeframe of 2002 to 2012.
Clearly, most of the companies have significantly increased their openness and slightly increased their
R&D intensity. However, the drivers of publication output seem to be complex. On the one hand, the
two largest firms in the sample, Novartis and Pfizer, have seen the highest increase in the number of
publications even though they have not significantly increased their openness. On the other hand, the
two companies with the largest increase in openness (Abbott and Bristol Myers) have also seen a
substantial increase in the number of publications.
36
Figure 11: Annual growth of R&D intensity (R&D expenditure divided by total sales) and annual growth of openness from 2002 to 2012. Bubble size corresponds to total sales in the year 2012. The color indicates the annual growth in number of publications.
Figure 12 links on a firm-level the type of collaboration between pharmaceutical companies (such as
interfirm alliance, purely internal research – see Figure 10) and the efficiency of publication activity (the
total R&D expenditures in a year divided by the number of papers in a year). Two observations can be
made. First, companies that engage more frequently in interfirm-alliances are also more open in general.
Second, companies that are the least efficient in creating publications (Merck, GlaxoSmithKline,
AstraZeneca, EliLilly) tend to be located on the lower left (low openness, low share of inter-firm alliances)
while the most efficient firms in publications tend to be on the upper right (high openness, high share of
inter-firm alliances)
37
Figure 12: Share of inter-firm alliances and and openness from 2002 to 2012. An interfirm alliance is defined as a paper that is jointly published by two or more of the ten largest pharmaceutical companies. Bubble size corresponds to Total Sales in 2012. The color indicates the R&D expenditure per published paper.
Two main observations from the firm-level analysis of Value Creation can be drawn. First, increasing the
openness seems to increase the number of published papers, however with the exception of the notable
exceptions of the two largest firms in the sample. Second, the data indicates that engaging in inter-firm
alliances and displaying a high degree of openness increases the efficiency of publishing activity. The
next section will focus on value creation aspects across different disease areas.
4.1.3 Disease Area Patterns in Value Creation
Table 10 provides an overview of the publications across disease areas. We observe openness scores
from 58.5% (Immunology) to 70.5% (Gastroenterology & Hepatology). Biotechnology has the lowest
openness of 43%.
38
Total publications
Relative share Openness
Gastroenterology & Hepatology 1106 3.3% 70.5%
Cardiovascular System & Cardiology 2866 8.6% 69.5%
Table 10: Number of publications, relative share, and openness in publications of disease areas from 2002 to 2012. ENT = ear, nose, throat. Not assigned are records which could not be clearly attributed to one disease area.
The next section presents the patterns and finding from the analysis of patent data from 2002 to 2012.
Hence, the attention moves from value creation to value capture.
4.2 Value Capture in the Top10 Pharmaceutical Companies
While publications have increased both in total number and in openness, patents have decreased in
these two respects. Figure 13 indicates the decline in total number of patents and average openness of
patents which is contrary to the increase in publications (see Figure 9). Within 10 years the number of
patents has halved and the openness has been reduced by a factor of three.
Figure 14 sheds light on the different collaboration types that are present in publications and patents.
Almost 80% of the publications are co-authored by at least one external author. Conversely, 80% of the
patents are assigned to one company only. Co-issuance with other institutions such as universities
amounts to only 20%. Interfirm co-issuance in patents is almost non-existent.
39
Figure 13: Number of patents and average openness of patents of the top10 pharmaceutical companies from 2002 to 2012.
Figure 14: Type of collaboration in publications and patents from 2002 to 2012. Interfirm alliance describes a publication that is co-authored or a patent that is issued with another Top10 pharmaceutical company.
Figure 15 visualizes which disease areas have gained in importance relative to other disease areas and
how their openness has evolved in the last 10 years. One observes that some key disease areas such as
oncology and infectious diseases have gained in importance. More mature disease areas like urogenital
and hematology have lost significance. Interestingly, an increase in relative importance does not parallel
an increase in openness and a rather complex picture emerges. Please note that openness refers to
openness in publications since openness in patents is very spurious for patents (see Figure 14). The
company-level and disease area specific data on patent openness is given in appendix 7.7.
Figure 15: Annual change of patent share of different disease areas against the change in openness of publications. Bubble size corresponds to total number of patents. The color does not have any special meaning in this case and is for illustration purposes only.
Some pharmaceutical companies specialize in certain disease areas, hence they issue more patents
specific to this therapeutic field. Figure 16 indicates the focus of the ten largest pharmaceutical
companies. While some companies (GlaxoSmithKline, Pfizer) have a fairly equal distribution across all
disease areas, some companies focus on key areas (Novartis on oncology, Eli Lilly on neurosciences &
neurology).
41
Figure 16: Relative share of patents of each of the 14 disease areas. Color indicates the relative share of patents from 2002 to 2012 for the ten largest pharmaceutical companies. The companies are ordered in increasing order (measured in total sales) from top to bottom.
4.3 Interplay of Value Creation and Value Capture
One might expect that the number of publications and patents follow a linear relationship: The more
publications, the more patents. However, the data clearly suggest that this is not the case. Figure 17
plots the number of publications against number of patents of each disease area. Even though we see
that more publications generally also means more patents, certain areas have a relatively high numbers
of publications, but not that many patents (Hematology, Neurosciences & Neurology) while certain other
areas (Biotechnology & Applied Microbiology) have a low number of publications but many patents.
42
Figure 17: Number of publications and number of patents of all disease areas
Table 11 shows in which areas the ten largest pharmaceutical companies had the largest share of
publications and patents. We see that they focused mostly on Neurosciences, Oncology, and
Table 11: Focus areas of the ten largest pharmaceutical companies. Focus areas determined by taking the average of the publications and patents share. The two areas with the highest average are the two focal areas.
Figure 18 shows the evolution of the share of focal areas in the total publication and patent portfolio.
While the share of focal areas on publications has remained constant, we see a slight increase at around
2010 in the share of focal areas in the overall patent portfolio. However, at this time the total number of
patents has also drastically decreased, therefore it could be that this is a statistical artifact.
In addition, no significant differences in disease areas on which the large pharmaceutical companies
focus on can be observed. Both have increased by about the same amount.
44
Figure 18: Evolution of focal areas of ten largest pharmaceutical companies from 2002 to 2012. Focal areas means the highest relative share of disease areas (Average of patent share and publication share)
7.1 VBA Function for Assigning Company Affiliation
Public Function CompanyName(Addresses As String, Optional ReprintAuthor As String) As String Dim CombinedString As String CombinedString = Addresses & ReprintAuthor If InStr(1, CombinedString, "Novartis", vbTextCompare) > 0 Then CompanyName = CompanyName & "[Novartis]" End If If InStr(1, CombinedString, "Pfizer", vbTextCompare) > 0 Then CompanyName = CompanyName & "[Pfizer]" End If If InStr(1, CombinedString, "Sanofi", vbTextCompare) > 0 Or InStr(1, CombinedString, "Aventis", vbTextCompare) > 0 Then CompanyName = CompanyName & "[Sanofi]" End If If InStr(1, CombinedString, "Merck", vbTextCompare) > 0 Then CompanyName = CompanyName & "[Merck]" End If If InStr(1, CombinedString, "Glaxo", vbTextCompare) > 0 Or InStr(1, CombinedString, "SmithKline", vbTextCompare) > 0 Then CompanyName = CompanyName & "[GlaxoSmithKline]" End If If InStr(1, CombinedString, "Abbott", vbTextCompare) > 0 Then CompanyName = CompanyName & "[Abbott]" End If If InStr(1, CombinedString, "Astra", vbTextCompare) > 0 Or InStr(1, CombinedString, "Zeneca", vbTextCompare) > 0 Then CompanyName = CompanyName & "[Astrazeneca]" End If If InStr(1, CombinedString, "Roche", vbTextCompare) > 0 Then CompanyName = CompanyName & "[Roche]" End If If InStr(1, CombinedString, "Lilly", vbTextCompare) > 0 Or InStr(1, CombinedString, "ELIL", vbTextCompare) > 0 Then CompanyName = CompanyName & "[Eli Lilly]" End If If InStr(1, CombinedString, "Bristol Myers", vbTextCompare) > 0 Or InStr(1, CombinedString, "Bristol-Myers", vbTextCompare) > 0 Then CompanyName = CompanyName & "[Bristol Myers]" End If End Function
7.2 VBA Function to Assign Publications to Subject Areas
Public Function CategorizePublication(Category As String, SubjectCategories As String) As Boolean
56
Select Case Category Case "Allergy, Respiratory Systems, ENT" If InStr(1, SubjectCategories, "Respiratory System", vbBinaryCompare) > 0 Or InStr(1, SubjectCategories, "Otorhinolaryngology", vbBinaryCompare) > 0 _ Or InStr(1, SubjectCategories, "Allergy", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Biotechnology & Applied Microbiology" If InStr(1, SubjectCategories, "Biotechnology & Applied Microbiology", vbBinaryCompare) > 0 Or InStr(1, SubjectCategories, "Genetics", vbBinaryCompare) > 0 _ Or InStr(1, SubjectCategories, "Mycology", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Cardiovascular System & Cardiology" If InStr(1, SubjectCategories, "Cardiovascular System & Cardiology", vbBinaryCompare) > 0 Then CategorizePublication = True End If If InStr(1, SubjectCategories, "Hematology", vbBinaryCompare) > 0 Then CategorizePublication = False End If Case "Dermatology" If InStr(1, SubjectCategories, "Dermatology", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Endocrinology & Metabolism" If InStr(1, SubjectCategories, "Endocrinology & Metabolism", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Gastroenterology & Hepatology" If InStr(1, SubjectCategories, "Gastroenterology & Hepatology", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Hematology" If InStr(1, SubjectCategories, "Hematology", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Immunology" If InStr(1, SubjectCategories, "Immunology", vbBinaryCompare) > 0 Or InStr(1, SubjectCategories, "Transplantation", vbBinaryCompare) > 0 Then CategorizePublication = True End If If InStr(1, SubjectCategories, "Respiratory System", vbBinaryCompare) > 0 Or InStr(1, SubjectCategories, "Otorhinolaryngology", vbBinaryCompare) > 0 _ Or InStr(1, SubjectCategories, "Allergy", vbBinaryCompare) > 0 Or InStr(1, SubjectCategories, "Infectious Diseases", vbBinaryCompare) > 0 _ Or InStr(1, SubjectCategories, "Parasitology", vbBinaryCompare) > 0 Or InStr(1, SubjectCategories, "Virology", vbBinaryCompare) > 0 Then CategorizePublication = False End If
57
Case "Infectious Diseases, Virology, Parasitology" If InStr(1, SubjectCategories, "Infectious Diseases", vbBinaryCompare) > 0 Or InStr(1, SubjectCategories, "Parasitology", vbBinaryCompare) > 0 _ Or InStr(1, SubjectCategories, "Virology", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Neurosciences & Neurology" If InStr(1, SubjectCategories, "Neurosciences & Neurology", vbBinaryCompare) > 0 Or InStr(1, SubjectCategories, "Psychiatry", vbBinaryCompare) > 0 _ Or InStr(1, SubjectCategories, "Substance Abuse", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Oncology" If InStr(1, SubjectCategories, "Oncology", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Ophthalmology" If InStr(1, SubjectCategories, "Ophthalmology", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Rheumatology" If InStr(1, SubjectCategories, "Rheumatology", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case "Urogenital & Pregnancy" If InStr(1, SubjectCategories, "Urology & Nephrology", vbBinaryCompare) > 0 Or InStr(1, SubjectCategories, "Reproductive Biology", vbBinaryCompare) > 0 _ Or InStr(1, SubjectCategories, "Obstetrics & Gynecology", vbBinaryCompare) > 0 Then CategorizePublication = True End If Case Else CategorizePublication = False End Select End Function
7.3 VBA Function to Assign Patent Codes to Subject Areas
Public Function CategorizePatent(Category As String, IPCcodes As String) As Boolean CategorizePatent = False Select Case Category Case "Allergy, Respiratory Systems, ENT" If InStr(1, IPCcodes, "A61P-037/08", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-011/06", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/16", vbBinaryCompare) > 0 Then CategorizePatent = True End If '' Remove Dermatology If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then
58
CategorizePatent = False End If '' Remove Opthalmology If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If Case "Biotechnology & Applied Microbiology" If InStr(1, IPCcodes, "C12N", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "C12P", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "C12Q", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61K-048", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A01H-001/06", vbBinaryCompare) > 0 Then CategorizePatent = True ''' Remove immunology If InStr(1, IPCcodes, "A61P-037/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-037/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-037/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "G01N-033/53", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "G01N-033/54", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "G01N-033/55", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "G01N-033/56", vbBinaryCompare) > 0 Then CategorizePatent = False End If End If Case "Cardiovascular System & Cardiology" If InStr(1, IPCcodes, "A61P-009", vbBinaryCompare) > 0 Then CategorizePatent = True '' Remove allergy If InStr(1, IPCcodes, "A61P-037/08", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-011/06", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/16", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Dermatology If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Gastroenterology If InStr(1, IPCcodes, "A61P-001", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Hematology
59
If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Opthalmology If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Urogenital & Pregnancy If InStr(1, IPCcodes, "A61P-013", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-015", vbBinaryCompare) > 0 Then CategorizePatent = False End If End If Case "Dermatology" If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then CategorizePatent = True End If '' Remove Opthalmology If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If Case "Endocrinology & Metabolism" If InStr(1, IPCcodes, "A61P-003", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-005", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-021/06", vbBinaryCompare) > 0 Then CategorizePatent = True '' Remove allergy If InStr(1, IPCcodes, "A61P-037/08", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-011/06", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/16", vbBinaryCompare) > 0 Then CategorizePatent = False End If ''' Remove Cardiovascular If InStr(1, IPCcodes, "A61P-009", vbBinaryCompare) > 0 Then CategorizePatent = False
60
End If '' Remove Dermatology If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Gastroenterology If InStr(1, IPCcodes, "A61P-001", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Hematology If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Opthalmology If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Urogenital & Pregnancy If InStr(1, IPCcodes, "A61P-013", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-015", vbBinaryCompare) > 0 Then CategorizePatent = False End If End If Case "Gastroenterology & Hepatology" If InStr(1, IPCcodes, "A61P-001", vbBinaryCompare) > 0 Then CategorizePatent = True '' Remove allergy If InStr(1, IPCcodes, "A61P-037/08", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-011/06", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/16", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Dermatology If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Hematology If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Opthalmology
61
If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Urogenital & Pregnancy If InStr(1, IPCcodes, "A61P-013", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-015", vbBinaryCompare) > 0 Then CategorizePatent = False End If End If Case "Hematology" If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = True End If Case "Immunology" If InStr(1, IPCcodes, "A61P-037/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-037/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-037/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "G01N-033/53", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "G01N-033/54", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "G01N-033/55", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "G01N-033/56", vbBinaryCompare) > 0 Then CategorizePatent = True '' Remove allergy If InStr(1, IPCcodes, "A61P-037/08", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-011/06", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/16", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Dermatology If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Gastroenterology If InStr(1, IPCcodes, "A61P-001", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Hematology If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Opthalmology
62
If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Urogenital & Pregnancy If InStr(1, IPCcodes, "A61P-013", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-015", vbBinaryCompare) > 0 Then CategorizePatent = False End If End If Case "Infectious Diseases, Virology, Parasitology" If InStr(1, IPCcodes, "A61P-031", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-033", vbBinaryCompare) > 0 Then CategorizePatent = True '' Remove allergy If InStr(1, IPCcodes, "A61P-037/08", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-011/06", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/16", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Dermatology If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Gastroenterology If InStr(1, IPCcodes, "A61P-001", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Hematology If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Opthalmology If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If
63
'' Remove Urogenital & Pregnancy If InStr(1, IPCcodes, "A61P-013", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-015", vbBinaryCompare) > 0 Then CategorizePatent = False End If End If Case "Neurosciences & Neurology" If InStr(1, IPCcodes, "A61P-025", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-021/02", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-021/04", vbBinaryCompare) > 0 Then CategorizePatent = True '' Remove allergy If InStr(1, IPCcodes, "A61P-037/08", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-011/06", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/16", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Dermatology If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Gastroenterology If InStr(1, IPCcodes, "A61P-001", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Hematology If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Opthalmology If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Urogenital & Pregnancy If InStr(1, IPCcodes, "A61P-013", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-015", vbBinaryCompare) > 0 Then CategorizePatent = False End If ''' Remove Rhemuatology If InStr(1, IPCcodes, "A61P-029", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-019", vbBinaryCompare) > 0 Then
64
CategorizePatent = False End If End If Case "Oncology" If InStr(1, IPCcodes, "A61P-035", vbBinaryCompare) > 0 Then CategorizePatent = True '' Remove allergy If InStr(1, IPCcodes, "A61P-037/08", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-011/06", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/16", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Dermatology If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Gastroenterology If InStr(1, IPCcodes, "A61P-001", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Hematology If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Opthalmology If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Urogenital & Pregnancy If InStr(1, IPCcodes, "A61P-013", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-015", vbBinaryCompare) > 0 Then CategorizePatent = False End If End If Case "Ophthalmology" If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then
65
CategorizePatent = True '' Remove Hematology If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = False End If End If Case "Rheumatology" If InStr(1, IPCcodes, "A61P-029", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-019", vbBinaryCompare) > 0 Then CategorizePatent = True '' Remove allergy If InStr(1, IPCcodes, "A61P-037/08", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-011/06", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/16", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Dermatology If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Gastroenterology If InStr(1, IPCcodes, "A61P-001", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Hematology If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Opthalmology If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Urogenital & Pregnancy If InStr(1, IPCcodes, "A61P-013", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-015", vbBinaryCompare) > 0 Then CategorizePatent = False End If End If Case "Urogenital & Pregnancy"
66
If InStr(1, IPCcodes, "A61P-013", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-015", vbBinaryCompare) > 0 Then CategorizePatent = True '' Remove allergy If InStr(1, IPCcodes, "A61P-037/08", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-011/06", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/16", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Dermatology If InStr(1, IPCcodes, "A61P-017", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Hematology If InStr(1, IPCcodes, "A61P-007", vbBinaryCompare) > 0 Then CategorizePatent = False End If '' Remove Opthalmology If InStr(1, IPCcodes, "A61P-027/02", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/04", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/06", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/08", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/10", vbBinaryCompare) > 0 Or InStr(1, IPCcodes, "A61P-027/12", vbBinaryCompare) > 0 _ Or InStr(1, IPCcodes, "A61P-027/14", vbBinaryCompare) > 0 Then CategorizePatent = False End If End If End Select End Function
7.4 VBA Function to Determine Openness
Public Function JointnessInstitutional(ByVal Addresses As String, CompanyName As String) As Single Dim i As Integer Dim NumberOfInstitutions As Integer Dim SemiColonLocation As Integer Dim BracketLocation As Integer Dim TempAddresses As String Dim ExternalInstitutionCount As Integer ''' Before 2006 no individual author affiliation If Len(Addresses) - Len(Replace(Addresses, "]", "")) = 0 Then NumberOfInstitutions = 1 + Len(Addresses) - Len(Replace(Addresses, ";", "")) If NumberOfInstitutions <> 1 Then For i = 1 To NumberOfInstitutions - 1 SemiColonLocation = InStr(1, Addresses, ";", vbBinaryCompare) TempAddresses = Left(Addresses, InStr(1, Addresses, ";", vbBinaryCompare)) If CompanyName = "Sanofi" Then
67
If InStr(1, TempAddresses, "Sanofi", vbTextCompare) = 0 And InStr(1, TempAddresses, "Aventis", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "GlaxoSmithKline" Then If InStr(1, TempAddresses, "Glaxo", vbTextCompare) = 0 And InStr(1, TempAddresses, "SmithKline", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "Eli Lilly" Then If InStr(1, TempAddresses, "Lilly", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "Astrazeneca" Then If InStr(1, TempAddresses, "Astra", vbTextCompare) = 0 And InStr(1, TempAddresses, "Zeneca", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If Else If InStr(1, TempAddresses, CompanyName, vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If End If Addresses = Right(Addresses, Len(Addresses) - SemiColonLocation) Next i End If ''' For last address If CompanyName = "Sanofi" Then If InStr(1, Addresses, "Sanofi", vbTextCompare) = 0 And InStr(1, Addresses, "Aventis", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "GlaxoSmithKline" Then If InStr(1, Addresses, "Glaxo", vbTextCompare) = 0 And InStr(1, Addresses, "SmithKline", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "Eli Lilly" Then If InStr(1, Addresses, "Lilly", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "Astrazeneca" Then If InStr(1, Addresses, "Astra", vbTextCompare) = 0 And InStr(1, Addresses, "Zeneca", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If Else If InStr(1, Addresses, CompanyName, vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1
68
End If End If Else ''' After 2006 Individual Author affiliation NumberOfInstitutions = Len(Addresses) - Len(Replace(Addresses, "]", "")) If NumberOfInstitutions <> 1 Then For i = 1 To NumberOfInstitutions - 1 BracketLocation = InStr(1, Addresses, "]", vbBinaryCompare) SemiColonLocation = InStr(BracketLocation, Addresses, ";", vbBinaryCompare) TempAddresses = Mid(Addresses, BracketLocation + 1, SemiColonLocation - BracketLocation) If CompanyName = "Sanofi" Then If InStr(1, TempAddresses, "Sanofi", vbTextCompare) = 0 And InStr(1, TempAddresses, "Aventis", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "GlaxoSmithKline" Then If InStr(1, TempAddresses, "Glaxo", vbTextCompare) = 0 And InStr(1, TempAddresses, "SmithKline", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "Eli Lilly" Then If InStr(1, TempAddresses, "Lilly", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "Astrazeneca" Then If InStr(1, TempAddresses, "Astra", vbTextCompare) = 0 And InStr(1, TempAddresses, "Zeneca", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If Else If InStr(1, TempAddresses, CompanyName, vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If End If Addresses = Right(Addresses, Len(Addresses) - SemiColonLocation) Next i End If ''' For last address If CompanyName = "Sanofi" Then If InStr(1, Addresses, "Sanofi", vbTextCompare) = 0 And InStr(1, Addresses, "Aventis", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "GlaxoSmithKline" Then If InStr(1, Addresses, "Glaxo", vbTextCompare) = 0 And InStr(1, Addresses, "SmithKline", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "Eli Lilly" Then
69
If InStr(1, Addresses, "Lilly", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If ElseIf CompanyName = "Astrazeneca" Then If InStr(1, Addresses, "Astra", vbTextCompare) = 0 And InStr(1, Addresses, "Zeneca", vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If Else If InStr(1, Addresses, CompanyName, vbTextCompare) = 0 Then ExternalInstitutionCount = ExternalInstitutionCount + 1 End If End If End If JointnessInstitutional = ExternalInstitutionCount / NumberOfInstitutions End Function
7.5 IPC Codes Linked to Subject Areas
IPC Code Description Assigned WoS
Categories
A61P 1/00 Drugs for disorders of the alimentary tract or the digestive
system [7]
Gastroenterology & Hepatology
A61P 1/02 · Stomatological preparations, e.g. drugs for caries, aphtae,
periodontitis [7] Gastroenterology & Hepatology
A61P 1/04 · for ulcers, gastritis or reflux esophagitis, e.g. antacids,
inhibitors of acid secretion, mucosal protectants [7] Gastroenterology & Hepatology
A61P 1/06 · Anti-spasmodics, e.g. drugs for colics, esophagic dyskinesia
[7] Gastroenterology & Hepatology
A61P 1/08 · for nausea, cinetosis or vertigo; Antiemetics [7] Gastroenterology & Hepatology