All together now: Improving cross-sector collaboration in the UK biomedical industry

8/7/2019 All together now: Improving cross-sector collaboration in the UK biomedical industry

1/45

Research report: March 2011

All together now:Improving cross-sector collaborationin the UK biomedical industry

Louise Marston


2/45

2

Executive summary

A strong history in pharmaceuticals andchemicals, combined with world-leading

university research, has created a signifcantcompetitive advantage or the UK biomedicalsciences industry. This industry is importantto the UK economy, accounting or 9 per cento exports and 28 per cent o business R&Dspending.1 It is an industry where the UK isobjectively one o the world leaders, but ourrole is not secure. As Pfzer announces theclosure o its UK R&D acility, the only oneoutside o the US, it is clear that our historicadvantages will not be enough to sustain theUKs position in the uture. As pharmaceutical

companies change their R&D strategies toa new, more outsourced model, and othercountries improve their scientifc inrastructure,we need to change to keep up. However, thereis an opportunity to raise our game by buildingbetter connections between the assets wealready have.

Many reports have examined this industry,but we believe that this is the frst attemptto examine its competitiveness in terms ocollaboration. The UK has signifcant assets inthe NHS, research universities, large medicalcharities, pharmaceutical companies, andclusters o smaller companies. This reportdemonstrates that these assets deliver more bybeing better connected. Collaboration allowsa better use o resources, avoiding duplicationand improving access to specialist acilitiesand expertise. Most importantly, collaborationimproves the capacity or innovation, whichis critical at a time when the industrys R&Dproductivity rates continue to all and pharmaincreasingly looks to external partners or itsdrug discovery. This report primarily examines

academic-industry collaborations, but alsolooks at the connections to smaller companies,clinicians and charities.

This report presents new data on the impact ocollaborative working, showing that biomedical

academic papers that are co-published withindustry have greater citation impact thanpurely academic papers. This fnding held trueacross all the countries we examined.

The UK starts in a very strong position: therelative citation impact o UK biomedicalresearch is high, around 1.5 times worldaverage in the academic sector and twice worldaverage in the corporate sector. More researchis conducted jointly with industry than inany other country except Finland. Academic

collaboration with the corporate sectorrepresents about 6 per cent o academic sectoroutput. This compares with just 3 per cent inGermany and Japan.

Industry currently unds around 10 per cento biomedical research in UK universities. Iall universities were to achieve 15 per cent oindustry income, this would mean an increasein industry unding or universities o 60 percent, an increase o more than 100 million.This fgure would represent just 8.4 per cent oextramural R&D spend by pharma companies,the same level spent in 2002.

However, although the UK perorms well onthis bibliometric output, the share has beenstatic where other countries have grown theirshare o academic-industry papers more rapidly Germany has seen 35 per cent growth overten years, compared to 14 per cent in the UK.Moreover, industrys contribution has beenalling as a share o their external researchunding. There is a danger that the UK willremain static while other countries grow and

overtake us as a competitive destination orresearch and investment. The potential is clear the university sector alone could increase

1. Ofce o National Statistics(2009) Expenditure on R&Dperormed in UK businesses:2009. Newport: ONS.Retrieved rom: http://www.statistics.gov.uk/StatBase/Product.asp?vlnk=8206&Pos=&ColRank=1&Rank=240


3/45

industry income by 60 per cent i it raisedperormance levels across all institutions.

Although the UK has maintained its shareo approved trials within Europe, those trialsail to recruit patients all too requently. Newdata presented in this report shows that the

result is a declining global share o patientsenrolled in clinical trials. The UK now enrolsless than 2 per cent o global patients inphase II trials, precisely the early stage oresearch where we should excel. The UK risksacquiring a reputation as a place where it is notworth attempting to run a clinical trial, withconsequent impacts on research investmentand on opportunities or healthcare.

Successul collaboration cannot be mandated.But it is possible to create the rightinrastructure, develop skilled people and

introduce the best processes to enable it tohappen. The ollowing measures can encouragecollaboration in the UK system:

Infrastructure

The UK should accelerate the

development of electronic patient

records to support medical research, and

aim to become the world leader. These

records have been successully introducedin Scotland, and provide a valuable researchresource. A revised approach where smallersystems are implemented or a local ordisease-specifc group can be an eectiveway to introduce records, while still beinguseul to researchers. The current pilot inNHS Ayrshire & Arran demonstrates onepossible model or this approach. Greaterpublic awareness is also needed, buildingupon existing support or the principle ousing electronic records to support healthresearch.

Funders and public institutions should

prioritise opportunities to share resources

and services. By reducing running costs throughgreater sharing o labs, imaging centres, dataand test results, and expertise, there couldbe signifcant savings that would allow moreresources to be devoted to biomedical research.Such co-operation will also encourage greatercollaboration. Funders can support thischange in behaviour in their unding policiesand assessment criteria. A group o major

unders has recently set out their requirementsto increase access to public health researchdata generated by work they und.

To support shared resources, the

VAT system should be reformed to

encourage rather than prevent research

collaboration. Universities and charitiesmust use their buildings at least 95 per cento the time or non-business charitablepurposes to retain a zero VAT rate on new

buildings. Given the apparent importanceo business collaboration to developinghigh quality research, this rule seemsinappropriate, discouraging partnershipswith industry, and with other charities, evenon early-stage research. The consultationo the VAT rules relating to shared services,promised by the Chancellor in the June 2010budget, should be completed urgently.

Funders and career structures should

increase their focus on research support

services to expand the capacity of

researchers. Greater development ospecialised areas o research support can reeup researchers or clinicians to ocus on theirspecialism. This can also provide industrywith access to the right researchers or aparticular project.

People

Industry needs to increase the number of

industry placements and secondments.Undergraduate placements, training schemesand secondments to industry help to buildcontacts and introduce researchers tothe realities o industry today. However,placements are declining, reducingopportunities to develop collaborativelinks between universities and companies.Knowledge Transer Partnerships allow smallercompanies to work with university researchers,and could be better used in this sector; theWellcome Trust Translation Medicine andTherapeutics scheme provides a good modelor involving larger companies in training.Industry can provide more opportunities, andpublic and private sector employers can placemore value on this type o experience.

Industry hiring and promotion criteria

should take greater account of cross-

sector experience. Those who work in bothindustry and universities are better placedto identiy opportunities or collaboration.Few are able to move rom industry intoacademia, and those who move in the

other direction oten lose their universityperspective. Public sector employers in thesector should consider industry experience in


4/45

4

hiring criteria, and greater use o sabbaticalsand secondments. Funders and researchassessments also need to recognise relevantindustry experience in awarding unding, oruniversities will not ollow suit.

Recognise the importance of co-located

facilities. Locating industry, charities anduniversities close to each other in clusterscould bring real benefts in shared knowledgeand discourse that can be invaluable insparking new ideas. Such clusters couldalso acilitate easy movement betweensectors without having to move home. Thesuccess o a cluster depends on having amixture o dierent organisations locatedtogether both public and private as wellas an environment that encourages opendiscussions o work.

Create organisational structures thatsupport applied work and collaboration.

Organising research around problemsrather than disciplines can help academicdepartments to engage with industry.Providing simple entry points or smallpharma companies engaging with universitiescould increase the degree o co-operationwith these frms. Having people dedicated toseeking out opportunities to solve problemsthrough collaboration can provide much-needed momentum, as it does at MIMIT in

Manchester.

Process

Introduce common standards for

researchers in industry and universities

to make collaboration easier. There aredierent approaches within commercialand academic laboratories, which canmake translating academic results into acommercial context difcult or impossible.Common standards and expectations couldmake collaboration easier.

Government research funding should

be reformed to remove penalties for

collaborative research. Its easier to getresearch council unding or an academicproject than an industry collaborationproject. Current unding allocations tendto avour blue-sky proposals over appliedwork, and despite some planned changesto the Research Excellence Framework,

this is reected in the excellence criteriaused to judge good research departments.Those academics who build collaborative

links should be recognised or this. Fundersshould prioritise the development oappropriate collaboration metrics. It shouldalso be possible to und research done incollaboration with companies in certaincircumstances.

University intellectual property policiesshould be more consistent to enable

successful collaborations. Universitiestoo oten overvalue intellectual property.They sometimes make it too hard toestablish collaborations, as IP negotiationsare disproportionate to the size o theresearch project. There should be muchgreater consistency and alignment betweenuniversities and the NHS across the UK onIP agreements, based on an objective odeveloping the research urther rather thanincome. Academic Health Science Centres

could provide a useul lead on this matter.

The NHS and universities should factor

in the full benets of industry co-

operation into its pricing policies so that

collaboration is not deterred. Universitiesactor in Full Economic Costs to assess whatthey charge industry or research work.Universities should share more o the risk andthe reward o shared work, and should actorin benefts gained by working with industryinto their pricing. Otherwise they are pricing

themselves out o the international market.Universities that wish to charge commercialrates or work should also be able to deliverhigh standards o project management.

Conclusion

The diversity o perspectives and organisationalsystems is vital to innovation i everyoneworked in the same way, these collaborationswould be much less valuable. But the barrierso mindset and incentives can be modifed toembrace collaboration. The speed with whichwe develop new cures, drugs and approachesto patient care is dependent on the extent towhich we allow biomedical innovations andtrials to take place.

The pharmaceutical industry is undergoing aseismic shit, transitioning rom large internalR&D acilities to more external collaborationsand partnerships. Recognising the importanceo collaboration to the industry and to

innovation will mean changing the unding,incentives and career structures o the UKbiomedical sector.


5/45

This report was written by Louise Marston. The author would like to thank ShanthaShanmugalingam or his many contributions to the project. Stian Westlake and Conor Ryan alsoprovided very valuable comments.

Data or this report was provided by the Higher Education Statistics Agency, Thomson Reuters,Evidence, CMR International, the Ofce or National Statistics, the National Science Foundation,and the Association o University Technology Managers.

The ollowing people have contributed to the research in this report.

Association o the British Pharmaceutical Industry (ABPI) Sarah Jones and Susanna Hellden

Academy o Medical Sciences Rachel Quinn, Robert Frost and Laurie Smith

Association o Medical Research Charities (AMRC) Becky Purvis

AstraZeneca John Stageman and Mark Robertson

Association o University Technology Managers (AUTM) Ashley Stevens

BioIndustry Association (BIA) Clive Dix, Nigel Gaymond, Joseph Wildy and Francetta CarrBrunel University Health Economics Research Group Martin Buxton

Canadian Institutes o Health Laura McAuley and Peggy Borbey

Cancer Research UK Alasdair Rankin and Aoie Regan

Chie Scientists Ofce Scotland Proessor Sir John Savill

Foreign & Commonwealth Ofce Sam Myers and Jessica Wright

GlaxoSmithKline Sue Middleton, Pauline Williams, Elaine Irving, Mark Bodmer, Malcolm Skingle,Jon Ellis and Je Kipling

Health Tech and Medicines KTN Malcolm Rhodes and Mark Bustard

Icon plc Cyril Clarke, VP Translational Medicine

Imperial College Proessor Stephen Smith and Kerry Clough

Lein Diagnostics Dr Daniel Daly

MIMIT Jackie Oldham, Director

MRC Sir John Chisholm and Wendy Ewart

MRC Drug Saety Science Centre Hugh Laverty

MRC Protein Phosphorylation Unit Proessor Sir Philip Cohen

MRC Technology Dave Tapolczay

National Science Foundation Raymond Wole

NIHR Dr Louise Wood, Dr Clare Morgan, Matthew Hallsworth and Glenn Wells

Ofce or Health Economics Jon Sussex and Phil ONeill

P1vital Gerry Dawson

Pfzer Virginia Acha and Ruth McKernan

Pfzer Scotland Diane Thomson, Steve Moore, Mark Beggs and Alistair Strachan

Stevenage Science Park Alan Baxter

The Wellcome Trust David Lynn, Annie Colgan, John Hemming and Liz Allen

Thomson Reuters Jonathan Adams, Karen Gurney, Neal Dunkinson, Amy Mann andJulien Debeauvais

University College London Dr Simcha Jong

University o Edinburgh Proessor Jonathan Seckl

Acknowledgements


6/45

6

Contents

Improving cross-sector collaboration in the UK biomedical industry

Part 1: Introduction 7

1.1 Biomedical research is an important UK industry investing heavily in R&D 7

1.2 This is an industry dependent on high quality research 7

1.3 We underuse the resources we have 8

1.4 Collaboration mechanisms can improve the industry 10

1.5 Motivations or change 11

Part 2: How much improvement could we see in UK collaboration? 13

2.1 About the data 13

2.2 Benefts o collaboration 14

2.3 Co-publication 16

2.4 University research income 17

2.5 Clinical trials 23

2.6 How big is the opportunity? 27

Part 3: How can we improve collaboration? 28

3.1 Create the right inrastructure 283.2 Support people who move between sectors 34

3.3 Get the right processes in place 38

Part 4: Conclusion and further work 42

4.1 Further work 42

Acronym Soup 43


7/45

Part 1: Introduction

1.1 Biomedical research is an importantUK industry investing heavily in R&D

The biomedical science sector is a rapidlygrowing sector, and a source o economicgrowth or the UK, employing 143,000 peopleand generating over 30 billion in turnover.2Encompassing pharmaceuticals, medicaltechnology and medical biotechnology, itis a high-tech and highly innovative sector,closely linked to the research excellence o UKuniversities.

UK pharmaceutical industry R&D expenditure

is third in the world ater the US and Japan3

4.4 billion in 2009.4 Biomedical sciencealso represents a sector that is economicallyimportant or the UK with more than 9 percent o UK exports worth 21 billion and28 per cent o all UK business Research andDevelopment (R&D) spending.5

An ageing Western population and increasingaccess to healthcare in the BRIC countriespositions this industry as one that is likelyto grow strongly in the 21st century. Manyother countries have recognised this andinvested in their research capabilities. There isa growing awareness that the UK is vulnerableto competition, and that we cannot becomplacent about our position. Between 1990and 2008, R&D investment in the United Statesgrew 5.6 times, whilst in Europe it only grew3.5 times.6 Singapore has invested over 6.25billion to develop the biomedical sciencessector since 1996,7 and is preparing to investa urther 7.5 billion between 2011 and 2015.Other countries have used recent economicstimulus spending to support the biomedical

industry the US National Institutes o Healthreceiving an additional $10.4 billion (7billion) over two years, and Frances Grand

Emprunt (big loan) injecting 22 billion (20billion) into research labs and universities.

These investments come at a time whenUK biomedical research budgets have beenmaintained in real terms, but not increased.

1.2 This is an industry dependent onhigh quality research

Pharmaceutical companies say that researchexcellence is their main driver or locatingresearch and trials in the UK. These companies

look or the best researchers, and the mostinteresting and relevant work; that is whatthey fnd in the UK. They also invest heavilyin R&D themselves: six o the top 25 R&Dinvestors in the UK are pharma or biotechcompanies.8 However, pharma companiesalso acknowledge that their R&D model ischanging. GlaxoSmithKline now does 50 percent o its drug discovery externally. Thevulnerability o this position has been madeclear as three UK R&D acilities have beenclosed by pharmaceutical companies since2010, including Pfzers recent announcemento 2,400 job losses in Sandwich. That fgureis equivalent to two-thirds o sta at theMRC. These closures and job losses atGlaxoSmithKlines Harlow plant mean the losso up to 4,100 jobs.9

The impact and importance o UK researchcan be measured using international citations.The relative citation impact o UK biomedicalresearch is high, at 1.5 times the world averageor academic research, and twice the averageor industry publications. This is greater than

the impact o US research in this feld which is1.37 and 1.68 respectively.

2. Department or BusinessInnovation and Skills,Department o Health, andUK Trade and Investment(2009) Strength andOpportunity: The landscapeo the medical technology,medical biotechnology andindustrial biotechnologyenterprises in the UK.London: BIS, DoH, UKTI.Available at: http://www.berr.gov.uk/fles/fle53947.pd

3. OECD (2011) Main Scienceand Technology Indicators.Volume 2010/1. Paris: OECD.

4. Ofce or National Statistics(2009) Business Expenditureon R&D (BERD): Expenditureon R&D perormed in UKbusinesses, 2009 dataset.Newport: ONS.

5. Ofce o National Statistics(2009) Expenditure on R&Dperormed in UK businesses:2009. Newport: ONS.Retrieved rom: http://www.statistics.gov.uk/StatBase/Product.asp?vlnk=8206&Pos=&ColRank=1&Rank=240

6. European Federation oPharmaceutical Industriesand Associations. See www.epia.org.

7. Science and InnovationNetwork Technology sector

report, Singapore, September2009.

8. BIS (2010) R&D Scoreboard2010. London: BIS. Retrievedrom: http://www.innovation.gov.uk/rd_scoreboard/?p=22on 20 December 2010

9. GlaxoSmithKline announcedin February 2010 that itwould lose 380 jobs at itsHarlow plant. AstraZenecaannounced its intention toclose the Charnwood site inMarch 2010, with the loss oup to 1,200 jobs. Merck thenannounced the closure o theNewhouse acility in July,with the potential loss o 250

jobs. Pfzer then announcedin January 2011 that it wouldbe closing its only R&D siteoutside the US in Sandwich,Kent, with the loss o up to2,400 jobs.


8/45

The UK has a strong reputation or world-leading science in the feld o oncology (cancerresearch). UK spending on cancer research isestimated at 13.18 per capita, compared to17.61 in the US, and well above the Europeanaverage o 3.42.10 In part, this reputationis built on the unusually strong links ormed

between academics and clinicians in theUK cancer research community, supportedby 20 years o relationship development.This expertise is critically dependent onstrong relationships between undingbodies, researchers, clinicians and industryunders,such as the collaboration betweencancer research unders and industry throughthe National Cancer Research Institute (NCRI).The NCRI was the frst national umbrellaorganisation to bring together government,charity and industry unders. The NationalCancer Research Network served as a modelor the National Clinical research networks thatollowed.

1.3 We underuse the resources we have

The UK has a number o characteristics thatgive us this international advantage. Theseinclude: our world-leading universities; anestablished cohort o clinical academics,who pursue a clinical career as well as doing

research; and a large and diverse population,with a good standard o medical care, makingthe UK a suitable site or clinical trials; and

a diverse system o public research undingthrough the National Institute o HealthResearch within the NHS, and the ResearchCouncils.

However, there are signifcant assets thatdierentiate the UK, and which could be better

exploited to support the industry. At a timewhen the industry is moving to a new modelo research, with much greater use o externalpartners, the system also needs to adapt tonew ways o working. We dont yet have ajoined-up system to do this. It took a longtime to build the relationships and networksthat make the oncology research system workwell, so the difculty o this task shouldnt beunderestimated. These underexploited researchassets include:

The NHS

Research excellence within universities andresearch institutes

Large medical charities

Large pharmaceutical companies

Communities o smaller companies

The NHS

The interaction between the biomedical

industry and the NHS should be a productiveone or both parties. However the NHS is otenperceived as a reluctant partner in healthcare

88

10. ECRM (2007) Investmentsand outputs o cancerresearch: rom the publicsector to industry. EuropeanCancer Research FundingSurvey 2007. London:ECRM.

Box 1: Developing specialist centres with a disease focus Cancer ResearchUK Centres

Cancer Research UK has established anational network o cancer research centresin a major strategic initiative. Cancer

Research UK Centres are partnershipsworking on a local level with universities,NHS Trusts, cancer networks and othercharities, and on a national level withgovernment and industry.

The Manchester Cancer ResearchCentre (MCRC) is a multi-million poundpartnership ounded by The University oManchester (which includes the PatersonInstitute or Cancer Research), The ChristieNHS Foundation Trust and Cancer ResearchUK. The MCRC is a world-class centre o

excellence or research, with ambitions

to more than double the level o cancerresearch activity in Manchester by 2015.

Additional partners, including industry,have been attracted to the concentration oexpertise and inrastructure in Manchesterthat is enabling cancer research to ourish.Extensive collaboration with industryinvolves many dierent pharmaceutical andbioscience companies. AstraZeneca hasa long-standing and successul strategicpartnership with Manchester, providingsignifcant investment against areas omutual scientifc interest. Areas o interestinclude biomarker identifcation andvalidation, biobanking, imaging, radiation

combinations and clinical research training.


9/45

research. Research is seen as a peripheralactivity in most parts o the health service,approvals or clinical trials are slow to obtainand little thought is given to the researchservices and resources the NHS could oer.

The system is moving in the right direction.

A considerable investment in new researchinrastructure has been made since the BestResearch or Best Health report in 2006and the creation o the National Institute orHealth Research (NIHR).11 This recognises theimportance o health research to both theeconomy and the nations health. Investmenthas been put into research centres andacilities, new clinical academic career pathsand reorms to make it easier or researchers togain permission to work with NHS patients.

This commitment has been maintained by the

coalition government, who have preservedNIHR and MRC budgets, amidst cuts in otherresearch areas.

The change goes beyond the NHS Trustsassociated with large research universities. AllTrusts are now required to report the numbero patients recruited into clinical trials in theirquality accounts, putting the issue on the desko every Trust Chie Executive.12 What eectthis will have remains to be seen.

Research excellence within universities andresearch institutes

The UK has an international reputation orresearch excellence the key driver o industrywhen seeking out research partners. The UKhas our universities in the global top ten orClinical, Pre-clinical and Health research13and is second in the world in its output oclinical and health-related research papersand citations.14 Hal o the MRCs spendinggoes to its research institutes, such as theinternationally recognised Laboratory oMolecular Biology in Cambridge.

Clinical academics are an important parto the UK research system. They make upbetween 5 and 10 per cent o the UK medicalworkorce.15 Most are employed by a university,contributing to undergraduate teaching andcurriculum, post-graduate medical training, andpractising medicine or around hal the week.The UK has around 3,000 clinical academics, afgure that has been increasing since 2006.

Connecting the worlds o research science

and clinical practice is critical or innovation.Biomedical innovation is not a linear process,with basic science being refned through trials

until it reaches clinical practice. Rather, itrelies on a range o eedback loops to ensurethat research is connected to clinical needand patient experience. Clinical academics canprovide this eedback.

Bridges between research and clinical practice

can also be built at the institutional level. In2008, fve Academic Health Science Centres(AHSCs) were created in the UK, to bringtogether leading research universities with NHSTrusts. These have been modelled on exampleso similar institutions in the United States asinstitutions to integrate research, education,and clinical hospital practice.

Large medical charities

The UK is a major location or charitableinvestment in biomedical research. Charitiescontribute approximately one third o all public

expenditure on medical research in the UK,over 1 billion in 2009/10.

Clinical Medicine research in universities is evenmore reliant on charities it is unded 46 percent by charities, 41 per cent by government orresearch councils, and 9 per cent by industry.Charities provided 834 million o universityresearch unding in 2008/09.16

The Wellcome Trust is an international exemplaror charitable unding o basic research. With

a science budget o around 590 million in2008/09, it has huge inuence on the sector,and oten leads the way in new models osupporting research. There are many otherlarge medical charities in the UK, such asCancer Research UK, The British HeartFoundation and Arthritis Research UK, whohave a combined research budget o over 400million.17 They provide important perspectiveson the prioritisation o research or patientbeneft, but can also connect research topatient groups more directly, involving themwith trial design and implementation. Thisstrength is recognised by the UK public whoare more generous in their donations tomedical research charities than people in mostother countries. Seventy per cent o the UKpublic say they have donated money to medicalresearch campaigns, a fgure that is second inEurope, and well above the EU27 average o 39per cent.18

Research by the Ofce or Health Economicsdescribes the complementary role thatpublic and charitable research unding has in

stimulating private sector R&D investment. Thereport states: A 1 increase in UK governmentor charity spending on medical research could

11. NIHR (2006) Best Researchor Best Health. London:NIHR. Available at: http://www.dh.gov.uk/en/Publicationsandstatistics/Publications/PublicationsPolicyAndGuidance/DH_4127127

12. NIHR has more inormationon the Quality Accounts

requirements or research.See http://www.nihr.ac.uk/Pages/QualityAccounts.aspx

13. Times Higher EducationWorld University Rankings2010. Available at: http://www.timeshighereducation.co.uk/world-university-rankings/2010-2011/top-200.html

14. Evidence or the Departmentor Business, Innovation andSkills. 2009. Internationalcomparative perormance othe UK Research Base.

15. Fitzpatrick, S. (2010) ASurvey o Stafng Levels oMedical Clinical Academicsin UK Medical Schools asat 31 July 2009. London:Medical Schools Council.Available at: http://www.medschools.ac.uk/Publications/Pages/Stafng-survey-2009.aspx.

16. HESA data rom universityfnancial returns.

17. AMRC Member Database.Retrieved rom: http://www.amrc.org.uk/our-members_member-profles

18. Eurobarometer (2010)Special Survey: 340Science and Technology.Brussels: EC. Availableat: http://ec.europa.eu/public_opinion/archives/eb_special_en.htm


10/45

lead to an increase in private research spending

from the pharmaceutical industry of between

2.20 and 5.10.19

Large pharmaceutical companies

Two o the top ten pharmaceutical companiesby market capitalisation are based in the UK

GlaxoSmithKline and AstraZeneca. The pharmaindustry is unusual among high-tech industriesin having large active companies based inthe UK. The electronics industry requentlycomplains about a shortage o large companiesto invest in acquisitions, training anddevelopment. It also means that when smallercompanies are acquired, there is a good chancethat they can remain in the UK. However,historically, UK stock markets have been a lessavourable exit route or small biotechnologycompanies, aecting venture unding or thesecompanies at the early stages. The UK also

attracts large amounts o inward investmentin pharmaceutical R&D, 1.6 billion in 2009.Unortunately, the vulnerability o thisinvestment has been underlined by Pfzersdecision to close its Sandwich site, its only R&Dacility outside the US. This and other R&Dsite closures have underlined the global natureo these companies and the mobility o theirresearch activities. A UK headquarters is notenough to guarantee UK R&D activity or thesecompanies, who must compare the UK withuniversities, workorces, tax regimes and health

institutions worldwide. Moreover, large R&Dacilities are less relevant to the new model orpharma, which relies more heavily on externalpartnerships and collaborations. As AndrewWitty, CEO o GlaxoSmithKline, said o thevacant Pfzer site: The last thing we need is abig pile of bricks with air-conditioning.20

Small company communities

In addition to the large pharmaceuticalcompanies, there is a strong community osmaller companies in the UK, those developingpharmaceutical compounds, as well asbiotechnology companies and those developingdevices and diagnostic technologies. A 2009report21 highlighted the importance o thewider community o biotechnology and medicaltechnology companies. These companies arealmost all SMEs, have a combined turnovero 14.8 billion compared to pharmaceuticalcompanies turnover o 15.6 billion peryear, and employ over 75,000 people. Aspharmaceutical companies increasinglylook to external sources or new ideas andtechnologies, a healthy small company

community is important or the health o thesector.

1.4 Collaboration mechanisms canimprove the industry

The question o how the NHS works with thepharmaceutical industry has been addressed bya number o reports in recent years, includingthose by Sir David Cooksey,22 the Ofce or Lie

Sciences,

23

the Academy o Medical Sciences,

24

the National Institute or Health Research25and the Royal College o Physicians.26Recommendations emerging rom these reportscan be grouped into a number o high-levelthemes:

Institutions: Recommendations aroundthe budgets and goals o key vehicles orchange such as NIHR, UK Clinical ResearchCollaboration (UKCRC), the Ofce orStrategic Coordination o Health Research(OSCHR), the Medical Research Council and

the Super Cluster.

Regulation: changes in the regulation oclinical trials and pricing regulation.

Funding: improving access to unding orbiotech and pharma companies; preservingpublic unding or areas o need.

Skills: ensure the supply o skilledindividuals, and make it easier or doctorsand other healthcare proessionals to acquire

research skills.

Collaborative mechanisms: ora, incentivesand metrics that promote interactionsbetween the various players.

While the other our areas are important,this report seeks to examine the potentialo collaborative mechanisms or improvingthe UKs biomedical industry. We aim toprovide practical recommendations onhow collaboration can be encouraged, as acost-eective way to make the most o ourbiomedical science assets.

The 2010 Academy o Medical SciencesReaping the Rewards report27 highlighted theneed to:

Encourage alliances between the NHS,universities and industry to share the risk andreward associated with generating more cost-eective and novel therapeutics, diagnosticsand devices.

10

19. Alzheimers ResearchTrust & Ofce o HealthEconomics (2009) ForwardTogether. Cambridge:Alzheimers Research.Available at: http://www.alzheimers-research.org.uk/news/article.php?type=News&id=477.

20. Witty, A. (2011) Drugs:Supply running low.Financial Times. 9February 2011. See:http://www.t.com/cms/s/0/46d4a950-348e-11e0-9ebc-00144eabdc0.html?tcamp=crm/email/2011210/nbe/ScienceEnvironment/

product#axzz1DXn9zmx21. Department or Business

Innovation and Skills,Department o Health, andUK Trade and Investment(2009) Strength andOpportunity: The landscapeo the medical technology,medical biotechnology andindustrial biotechnologyenterprises in the UK.London: BIS, DoH, UKTI.Available at: http://www.berr.gov.uk/fles/fle53947.pd

22. Bioscience Innovationand Growth Team (2003)Bioscience 2015: Improving

Natonal Health, IncreasingNational Wealth. London:IB-IGT.

23. Ofce or Lie Sciences(2009) Lie SciencesBlueprint. London: OLS;also Ofce or Lie Sciences(2010) Delivering theBlueprint. London: OLS.

24. Academy o MedicalSciences (2010) Reapingthe Rewards: a vision or UKmedical science. London:Academy o MedicalSciences.

25. National Institute or HealthResearch (2006) BestResearch or Best Health.London: NIHR; also NIHR(2008) Best Research:implementation review.London: NIHR.

26. Royal College o Physicians(2009) Innovating orHealth. London: RoyalCollege o Physicians.

27. Academy o MedicalSciences (2010) Reapingthe Rewards: a visionor UK medical science.London: Academyo Medical Sciences.Available at: http://www.acmedsci.ac.uk/index.php?pid=99&puid=172


11/45

Similarly, the Royal College o Physicians28recommended that they would:

Work with the DH [Department o Health] todevise incentives or NHS bodies and doctorsto take part in research.

These are valuable goals. We aim to show howthese goals can be achieved through specifcincentives and by removing or mitigatingblockages to such alliances.

1.5 Motivations for change

This report comes at a critical time or theindustry. All parts o the system must changethe way they work to survive in the uture.Public budgets are being squeezed in bothhealthcare and research. At the same time, anageing population will increase demand or

healthcare as the latest research equipmentand healthcare treatments become morespecialised and more expensive.

Pharma has seen a decline in researchproductivity in recent years it costs more andmore to get a drug to launch, as the numbero potential compounds alling at each trialstage increases. New compounds must alsoprove they exceed the perormance o existingtherapies, as many diseases already havelicensed treatments which would be displacedby new ones. This reects a more competitivemarketplace and more stringent regulatoryrequirements. To improve productivity, pharmacompanies are now sourcing ideas and researchcapacity rom external sources, oten reducingin-house R&D capacity as a result. PatrickVallance o GlaxoSmithKline says that theynow do 50 per cent o their drug discoveryexternally a signifcant change rom just aew years ago. This pool o outsourced R&Dpresents a signifcant growth opportunity bothor universities and small companies in the UK.

The UK biomedical research system

Figure 1 is not intended to be a comprehensiveview o the entire UK system, but rather anindicator o the key elements. Funding bodies

1

OSCHR

The Office for Strategic

Coordination of Health Research

NIHR

National Institutefor Health Research

Clinical Research

Clinical Research Facilities

Biomedical Research CentresBiomedical Research Units

Experimental Cancer Research Centres

Strategic

Health

Authorities

(SHAs)

NHSHospital

Trusts

Primary Care

Trusts (PCTs)

MRC

890m 704m

Universities &

Medical Schools

Research

Institutes

834m200m

966m

BBSRC

EPSRCVenture

CapitalInvestorsFUNDING

PUBLIC FUNDERS

RESEARCH

PRACTICE

Clinical Research

Organisations (CROs)

MHRA

Medicines and Healthcare

Products Regulatory Agency

NICE / SMC

Life Sciences Industry

Pharma, Biotech, Devices

ABPI BIA ABHI

HEFCE

Wellcome

Trust

CHARITIES

CancerResearch UK

British Heart

Foundation

480m

334m

79m

966m

Figure 1: The UK biomedical research system

28. Royal College o Physicians(2009) Innovating orHealth. London: RoyalCollege o Physicians.


12/45

are arranged at the top o the diagram, withresearch and clinical practice arrayed lowerdown. Some bodies span both areas medicalschools being an example o an institutionconducting research as well as clinical practice.

The fgures indicate some o the unding owsreerred to later in the report. These fgures arebased on 2008/09 fgures or consistency.

For more detail on some o the acronyms inthis report, see the Acronym Soup appendix onpage 44.

12

Box 2: Structure of the UK biomedical system and some terminology

There is a defned series o developmentstages involved in drug development.Although it is not a purely linear process,there is a series o stages associated with

approval milestones that will be reerred toin this report.

Basic biological research leads to appliedresearch, done on human cells or cultureswith a clinical need or disease targetin mind. Translational or experimentalmedicine are terms applied to the boundarybetween research and clinical application.This oten takes the orm o very earlyexperiments on patients, oten to establisha biomarker or indicator that thetreatment is operating in the expected way.

Clinical trials are numbered rom I to IV,with phase I frst-in-human trials usuallydone on healthy volunteers. Experimentalmedicine experiments on patients, using

micro doses o the treatments, may betermed phase 0. Phase II seeks to establishthe eective dose (phase IIa) and measureefcacy (phase IIb), so are conducted onsmall groups o patients.

Phase III trials are large-scale patient trialsto assess efcacy with as much certainty aspossible. Phase IV trials are post-approvaland provide surveillance o usage, lookingor long-term eects and rare side eectsor interactions.


13/45

Part 2: How much improvement could we see in UK

collaboration?

In this chapter, we set out to quantiy theUKs current perormance in biomedical

collaboration, and compare it with othercountries. Existing data on collaborationis limited, so the conclusions below arenecessarily incomplete; we hope that utureresearch will improve the available data. Wehave also examined the size o the prize thepotential opportunity or the UK that couldbe unlocked. Later chapters examine ways toaccess this opportunity.

2.1 About the data

We have worked with Thomson Reuters andthe Higher Education Statistics Agency (HESA)to compile data on the UKs biomedicalcollaboration perormance. The countries weprofle include the largest pharma players aswell as smaller nations such as Singapore andFinland that are growing in importance. Figuresthroughout this report can be accessed alongwith the underlying data at http://www.nesta.org.uk

3

Box 3: Denition of biomedical research

For data sourced rom Thomson Reuters(Scientifc) Inc., the journal categoriesrepresenting biomedical research wereagreed ater consultation with NESTAand include all those in the ollowingThomson Reuters Essential ScienceIndicators research felds. These researchfelds encompass 71 journal categoriesrom the 251 used in the Web o Scienceclassifcation scheme:

Biology & Biochemistry

Clinical Medicine

Immunology

Microbiology

Molecular Biology & Genetics

Neuroscience & Behaviour

Pharmacology & Toxicology

For the Higher Education data, subjectareas were selected or analysis rom thecategories used or university fnancialreturns to HEFCE:

Anatomy & Physiology

Biosciences

Clinical Dentistry

Clinical Medicine

Pharmacy & Pharmacology

Veterinary Science


14/45

Figure 2 illustrates the output volume, citationimpact and growth in output o biomedicalpapers or the countries within our analysis. The

US obviously dominates publication volumes.The UK, Germany and Japan come next withbroadly similar levels o output, but with verydierent levels o citation impact rom thework. Growth in the number o publications isbroadly similar or the US, UK and Germany,with Canada and Singapore growing aster,and Finland and Japan showing slower growth.The chart illustrates the point that the UK isa signifcant source o biomedical research,producing highly cited work, but with growthslower than countries like Canada, Germanyand the US, that position is vulnerable.

2.2 Benets of collaboration

Beore looking at the status and the utureopportunities or the industry, we need toexamine why collaboration is important, andthe benefts that might be expected romincreasing it.

Collaboration produces more advanced

innovationsInnovation is brought about by therecombination o knowledge and resources

rom dierent parts o the system. Connectedsystems, where this knowledge ows to whereit can be applied, trump disconnected systems.

We know that frms that introduce new tomarket innovations are more likely to co-operate with external partners or innovationthan those who introduce more incrementalnew to frm products or services.30 Thosewho introduce these more advanced orms oinnovation are twice as likely to use universitiesas innovation partners, in addition to suppliersand customers, compared to those pursuingincremental innovations.31

New and advanced technologies require manytypes o expertise, which are hard to assemblein one frm or organisation, so externalpartnerships are necessary to bring in the rangeo knowledge needed or a new to marketinnovation. Innovation results rom the clashbetween different modes of behaviour and

habits of thought.32

Collaborative research has greater impact

A new analysis o UK biomedical publicationsshows that our impact in biomedical sciencesis high relative to other countries, and that

collaborative papers have a greater citationimpact than purely academic papers.

14

Figure 2: Biomedical publishing in the countries we analysed

Source: Thomson Reuters (Scientifc) Inc.29

Note: Bubble size represents the total number o biomedical papers published 2000 to 2009.

29. Citation impact is the

weighted average citationimpact or academic,corporate and academic-corporate papers. Growthis the percentage increasein total numbers o papersbetween 2000 and 2009.Total volume o papers 2000to 2009 determines thebubble area.

30. Tether, B.S. (2002) Whoco-operates or innovation,and why: An empiricalanalysis. Research Policy.31, No. 6 (August 2002),pp.947-967. Available at:http://www.sciencedirect.com/science/article/B6V77-45KYWP7-6/2/c475d83da9069bbe902a082a729b89

31. Kaumann, A. and Tdtling,F. (2001) Science-industryinteraction in the process oinnovation: the importanceo boundary-crossingbetween systems. ResearchPolicy. 3o, pp.791-804.Available at: http://dx.doi.org/10.1016/S0048-7333(00)00118-9.

32. Lundvall, B.-. (2010)National Systems oInnovation: Toward a Theoryo Innovation and InteractiveLearning. London: AnthemPress. p.36. Available at:http://books.google.com/books?id=iDXGwacw-4oC&pgis=1.

0.6

0.4

0.8

0.2

2

1.8

1.6

1.4

1.2

1

0

Citation

impact

0% 50% 100%-200% 250%

Growth in papers 2000-2009

UK

Canada Singapore

Germany

Japan

FinlandUS


15/45

Science has some well-established metrics orquality o research, oten ocusing on howrequently a paper is cited by other authors.This is not a perect measure o a worksimportance, but it is widely used by undingbodies and research institutions.

To review whether collaborations betweenacademic and industry researchers are moreimpactul than when working alone, weexamined the citation impact o all biomedicalpapers published in these three groups.Citation rates vary between research felds andwith time, consequently, analyses must takeboth feld and year into account. In addition,the type o publication will inuence thecitation count. For this reason, only citationcounts rom reviews and articles are used incalculations o citation impact. The standardnormalisation actor is the world average

citations per paper or the year and journalcategory in which the paper was published,which is rebased to 1. The citation impact inthis analysis is the overall average or the ten-year period.

The evidence rom this data is frstly, that UKcitation impact in biomedical sciences is highrelative to other countries, but papers withboth academic and industry authors are evenmore impactul than either purely academic orpurely industry-based publications. An analysisconducted by Pfzer o their co-published work

confrms that the average impact o researchco-published with European authors has risenabove the average impact o research withNorth American co-authors (although theEuropean volumes are smaller).

This analysis supports other work that hasdemonstrated a link between interdisciplinaryand inter-institution collaboration andincreased citation impact.33,34 It is not possibleto completely rule out the inuence o sel-citation on these fgures (where authors citetheir own work in later papers), but equally

there is no reason to suppose that the levelo sel-citation is greater in one part o thedata (either by sector or by country) thanin another. Other research investigating theimpact o this eect has ound it to be small,

5

2.5

2

1.5

1

0.5

0

UnitedKingdom

United States Canada Finland Singapore Germany Japan

Ac adem ic Ac ade mi c- co rp or ate C or por at e

Figure 3: Relative citation impact o biomedical papers across seven countries

Source: Thomson Reuters (Scientifc) Inc.

33. Hsu, J. and Huang, D.(2010) Correlation betweenimpact and collaboration.Scientometrics. 86,pp.317324. Available at:http://www.springerlink.com/index/10.1007/s11192-010-0265-x.

34. Katz, J.S. and Hicks,D. (1997) How muchis a collaboration

worth? A calibratedbibliometric model.Scientometrics.,40(3),pp.541-554.


16/45

and that it does not aect the conclusion thatcollaborative papers have greater impact.35The consistent dierence between sectorsin dierent countries also reinorces thisconclusion.

Although it is hard to measure the degree o

genuine collaboration that lies behind theseco-published papers, these data seem toreute any perception that research done withindustry is o lower quality, damages academicstatus or can be considered as less serious. Italso demonstrates that increased collaborationcould beneft not only income levels but alsothe citation impact o research.

Further benets of collaboration: better use

of resources, better insights

Collaboration brings together those withcomplementary skills to produce more

eective teams and greater insights. A studyo academic organisational structures in theuniversities o Northern Caliornia describeshow those with a problem-ocused structureand many dierent disciplines in the same teamwere better at generating spin-out companiesand industry links.36 These case studies suggestthat university spino networks can providevaluable expertise back to universities, andindustry links can highlight their currentproblems as areas or new research. Suchlinks bring together basic and applied

researchers; create connections to industryto ensure that research links to problemsrelevant to industry; and provide space ororganisational experimentation and rewardingentrepreneurship.

Collaboration also allows researchersgreater access to resources, whether thoseare proprietary compounds rom pharmacompanies or imaging acilities or industryresearchers.

Measuring collaboration

We have seen the advantages o collaboration.So how does the UK compare internationally?We use co-publication rates, research incomeand clinical trials as three metrics to assess this.Data on collaborative research and innovationis scarce, especially within sectors. A truecollaboration requires elements o sharedrisk and reward, as well as parallel (ratherthan sequential) eorts by the collaborators.These are not actors generally measuredin innovation surveys or fnancial accounts.The indicators or collaboration we have

used attempt to provide a proxy or truecollaboration rates, and are used becausethey can be compared internationally. These

data are new and we think provide importantnew insights, however, or those reasons,the conclusions should be treated with somecaution. We would welcome eorts by those inthe sector to expand and improve upon metricsor collaboration.

2.3 Co-publication

Co-publication is simply the creation o ascientifc paper with multiple authors, usuallyrom dierent institutions. Academic-industryco-publication, where at least one author isrom a university, and one rom a corporation,is a metric o collaboration that allowsinternational comparisons to be made easily,due to the availability o data on scientifcpublications. It also seems to be a good

indicator o a genuine scientifc collaboration,where the authors named are likely to haveworked together.

We chose seven countries or this analysis,including the top our locations or businessexpenditure on pharmaceutical R&D (US,Japan, UK and Germany), and smaller butgrowing locations o Singapore, Finland andCanada.

The UKs publication o biomedical papers (see

Box 3) makes up around 40 per cent o thenational research base, a similar proportion tothe US and Germany. The UK publishes 8.5 percent o the world output, and all the countrieswe studied have grown their biomedicalpublication output in the past decade, the UKslightly less than Germany or the U.S.

O the total UK academic output, 30 percent are co-authored with external partners.Academics collaborate not only with industry,but also with other academics, and externalorganisations when doing research, and this isreected in those with which they co-publisharticles. Around 18 per cent o the academicoutput is published with a health organisationand 6 per cent with industry researchers.Roughly hal o UK industry papers have anacademic co-author.

The UK has high rates of academic-industry

collaboration, but other countries are

improving

The UK produces the greatest number oacademic-corporate papers ater the USA.

The UK also has the highest proportion oacademic-industry co-publication, relative toacademic output, o the countries we looked

16

35. Wuchty, S., Jones, B.and Uzzi, B. (2007) TheIncreasing Dominance oTeams in Production oKnowledge. Science.316(5827), pp.1036-1039. Available at:http://www.sciencemag.org/cgi/content/abstract/316/5827/1036

36. Jong, S. (2006) How

organizational structuresin science shape spin-ofrms: the biochemistrydepartments o Berkeley,Stanord, and UCSF and thebirth o the biotech industry.Industrial and CorporateChange. 15(2), pp.251-283. Available at: http://icc.oxordjournals.org/cgi/doi/10.1093/icc/dtj014


17/45

at: 6 per cent o all academic publications wereco-authored with industry over ten years. Thiscompares with just 3 per cent in Germany andJapan.

In all countries except Japan, the number oacademic-corporate collaborative papers hasincreased over the period, Germanys recentincrease being particularly rapid.

The UKs academic-industry co-publicationrate, as a percentage o academic output, hasremained airly steady over the last ten years,while other countries have increased their ratetowards the UK level. This is a common themein the data the UK does well, but as othercountries invest and improve, our historicstrengths are no guarantee o uture success.

There is no reason to believe that 6 percent is a natural limit to this activity. Aspharmaceutical companies continue to becomemore comortable with publications, andas budgets are squeezed, there is scope toincrease the share o biomedical work involvingindustry researchers even urther.

2.4 University research income

The second source o data we have used isthe research income received by universities,

recorded on fnance returns sent to the HigherEducation Statistics Agency.

Sponsorship o university research bycompanies is just one orm o academic-

industry collaboration, but the availability ouniversity fnance data allows this measurementto be examined in some detail, at least in theUK. Universities and companies interact inmany ways, including consultancy, contractresearch, licensing o intellectual property,provision o training, industry placements andcollaborative research. Thereore, these fgureswill not represent the ull extent o industryspending with higher education institutions,and in-kind industry contributions such asaccess to equipment or compounds are notincluded. It also excludes income rom researchinstitutions such as the MRC institutes.

UK universities received a total o 2 billionin research unding or biomedical disciplinesin 2008-09. Most o this money came romcharities and rom government, via the researchcouncils, central government spending andlocal authorities; 9.7 per cent, or 200 millionwas rom UK and international companies.

O the 200 million spent by industry onacademic research unding, 129 million was

rom UK industry, and the remainder, arounda third, rom overseas sources. The majorityo this was or unding in clinical medicine

7

3,000

4,000

2,000

1,000

10,000

9,000

8,000

7,000

6,000

5,000

0

As percentage of total output

2009

academic-

corporate

papers

5.6% 5.7% 3.7% 4.6% 2.3% 6.3% 5.3%

UnitedStates

UnitedKingdom

Germany Canada Japan Finland Singapore

Figure 4: Academic-corporate co-published papers, 2009



18/45

18

3%

4%

2%

1%

7%

6%

5%

0%

United Kingdom

Canada

Finland

Germany

Japan

Singapore

United States

2000 2001 2002 2003 2004 2005 2006 2007 20092008

2,500

2,000

1,500

1,000

500

0

millions

1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09

Other Government and research councils Charities Industry and commercial sources

Figure 5: Co-publication trends over time

Figure 6: Sources o university biomedical research income


Source: Higher Education Statistics Agency (HESA)37

37. It should be noted that achange in categorisationin 2007/08 allowed thecontribution o EU andinternational companiesto be captured or the frsttime. Beore that date, thisincome is shown on thechart above as Other.


19/45

disciplines 138 million (this is also the

subject area that receives the majority ogovernment research unding).

It is worth noting that UK industry incomeor university research has grown more slowlythan other sources over the past ten years 4per cent per annum compared to 10 per centgrowth in government unding, and 7 per centby UK charities. Some have speculated that theimpact o ull economic cost pricing may beresponsible. This costing model was introducedin 2006 to ensure that universities calculatethe ull costs o each piece o work, includingoverheads. It has sometimes meant that the ulleconomic cost is then charged to commercialunders, making the UK a less competitiveplace or research, when compared to costs oequivalent work in Europe or the US (whichmay not incorporate ull overheads into theirpricing).

This metric can also be examined rom theother side, as expenditure on R&D by industry.An international comparison shows thatbusiness enterprise expenditure on all orms o

R&D in the pharmaceutical industry increasedby 5.3 per cent per year between 2002 and2006 in the UK, by 28.7 per cent in the US

and 12.6 per cent in Germany. These rapid

increases in pharmas R&D spending, as well asgovernments investment in biomedical researchin the UK, mask some more worrying trends.

Industry-unded research has been allingrelative to government and charity unding,although this could equally reect a rapidincrease in government unding in this area.The comprehensive spending review has largelymaintained unding to the sector, throughthe science budget and the NIHR allocationrom the Department o Health, althoughuniversities will have other unding cuts thatmay impact their ability or willingness tocollaborate. However, university unding hasalso allen as a share o industry spendingon external R&D. Where industry once spentaround 9 per cent o their external R&D budgetwith UK universities, they now spend closerto 5 per cent. This suggests that althoughindustry is spending more on external R&D,they are looking less and less to UK universitiesas a research partner.

International comparisons of university

income from industryUS universities do not collect data on incomeby subject area or department, so it is

9

UK industry

Europe industry

International industry

01 Clinical Medicine

02 Clinical Dentistry

03 Veterinary Science

04 Anatomy & Physiology

10 Biosciences

08 Pharmacy & Pharmacology

Figure 7: Industry unding or university research = 200 million

Source: HESA


20/45

20

impossible to produce a comparable breakdowno the respective contributions made bygovernment, charities and industries or thoseinstitutions. We can, however, make somecomparisons by looking at pharmaceuticalcompany R&D spending with universities over asimilar period.

While directly comparable data on USbiomedical research income is not available, thedata on pharmaceutical company spending inTable 1 suggests that the UK gains a greaterproportion o university income than the US

rom industry. The UK average industry incomeo 14 per cent o total university researchunding, compared to 9 per cent or the USinstitutions. Restricting the US dataset only tothose institutions with at least 50 per cent otheir invention disclosures in therapeutics (i.e.with a ocus on biomedical) gives an average o11 per cent.

Table 2 compares institutions, grouped bytheir total research income or the subject.For clinical medicine subjects, where themajority o industry spending goes, there is

38. UK university income isprovided by subject notby source o unds. Thisfgure is or industry-undedclinical medicine researchincome, on the assumptionthat this is wholly unded bythe pharmaceutical industry.Across all biomedicalsubjects, industry income is129 million.

3%

4%

2%

1%

10%

9%

8%

7%

6%

5%

0%

2001 2002 2003 2004 2005 2006 2007 2008

Industry spending as percentage of total university income

Industry spending with universities as percentage of pharma intramural R&D spending

Industry spending with universities as percentage of pharma extramural R&D spending

Figure 8: Ratio o industry R&D spending to university income

Source: HESA, Ofce or National Statistics

Table 1: Pharmaceutical industry R&D unding, US and UK

Source: OECD, NSF, ONS, HESA or 2008/09

Pharma industry R&D Extramural R&D spend Higher education Ratio of higher

spend (BERD) by pharma industry income from pharma education spend to

industry all extramural R&D

UK $6.18 billion (OECD) 2.48 billion (ONS) 84 million38 (HESA) 3.4%

US $69.5 billion (NSF) $17.1 billion (NSF) $469 million (NSF) 2.8%


21/45

no obvious link between the ability o high-income institutions (those in the top quartile)to raise industry unding and those with lower

incomes until you reach the lowest quartile.This relationship does not hold as true orbiosciences, although the lower quartiles aredominated by institutions with very smallresearch income levels, and where a ewinstitutions have 100 per cent industry unding,presumably or a one-o project.

Finally, we have compared the proportion oresearch income that is industry-unded withthe corresponding RAE assessment or thatsubject area (Figure 9).39

Figure 9 confrms that there is little correlationbetween research quality ratings andpercentage o industry income. In other words,the ability to work with industry is independento research quality or unding quantities,above a certain threshold. This implies thatother actors play a greater role, and wecan speculate that these might include theamount o outreach work the university does,or the organisational structure o the researchdepartments, and that some universities havesignifcant capacity or improvement. It is also

possible that the balance o basic and appliedresearch work is a determining actor, butwe have not been able to examine that with

the available data. By using just the clinicalmedicine subject area, we have attempted tominimise the inuence o this actor on the

analysis.

I all institutions that already receive someamount o industry unding or research wereto reach the benchmark set by the betterperorming institutions o 15 per cent industryunding (rom both UK and internationalsources), there would be an additional 118million o research unding available to theacademic community, an increase o 60 percent over current levels, and a return to2002 levels o industry unding, as a shareo pharmas external R&D spend. Additionalincome is by no means the most importantbeneft o collaborative research, and industryrepresentatives are at pains to be regarded asresearch partners and not as simple undingsources, but the fgure is illustrative o thepotential or increased collaboration. Thiscould provide signifcant resources at a timewhen government research unding, thoughprotected in real terms, is unlikely to increaseduring the UK Governments Spending Reviewperiod through to 2014-2015.

Learning from ScotlandScotland represents an interesting casestudy or industry collaboration. In 2006 the

1

39. Note: 4* rated researchwas calculated rom theweighted average by stanumbers o 4* ratings across5 Units o Assessment in the2008 RAE: Cancer Studies,Cardiovascular Medicine,Other Hospital Based ClinicalSubjects, Other LaboratoryBased Clinical Subjects andInection and Immunology.

This was compared to theindustry unding or ClinicalMedicine subjects obtainedrom HESA.

Table 2: UK universities comparison by quartile o research income

Quartiles by total clinical medicine Average research Average percentage Average industry

funding (n=39) income of industry funding research funding

Top quartile 238,915 10.3% 24,334

2nd quartile 83,357 9.1% 7,262

3rd quartile 28,264 12.3% 3,580

Bottom quartile 2,248 3.6% 119

Quartiles by bioscience funding Average research Average percentage Average industry

(n=99) income of industry funding research funding

Top quartile 48,224 5.5% 2,436

2nd quartile 9,334 10.0% 988

3rd

quartile 1,566 14.5% 230Bottom quartile 97 22.7% 30

Source: HESA


22/45

22

40. Percentage o research thatis 4* rated is a weightedaverage (by sta submitted)across the medicalsubjects o cancer studies,cardiovascular medicine,inection and immunology,other hospital basedclinical subjects and otherlaboratory based clinicalsubjects. This is based on

publicly available 2008data rom the ResearchAssessment Exercise.

20%

25%

15%

10%

5%

40%

35%

30%

0%

0% 5% 10% 15% 20% 25% 30% 35% 45%40%

Percentage

of industry

funding

Percentage of research that is 4* rated

60

50

40

30

20

10

0

millions

1999/00 2000/01 2001/02 2002/03 2003/04 2004/05 2005/06 2006/07 2007/08 2008/09

England Scotland

Figure 9: Link between research quality ratings and industry income

Figure 10: English and Scottish industry income

Source: HESA, HEFCE (RAE)40

Source: HESA


23/45


24/45

24

4%

2%

16%

8%

14%

12%

10%

6%

0%

25%

20%

10%

5%

15%

50%

45%

35%

40%

30%

0%

United Kingdom Canada Finland Germany

Japan Singapore

United States

2000 2001 2002 2003 2004 2005 2006 2007 20092008

United

States

4%

2%

18%

16%

8%

14%

12%

10%

6%

0%

30%

20%

10%

80%

70%

50%

60%

40%

0%

United Kingdom Canada Finland Germany

Japan Singapore United States

2000 2001 2002 2003 2004 2005 2006 2007 20092008

United

States

United States

Figure 11: Clinical trials approved per year, as a percentage o global total

Figure 12: Clinical trial sites approved, as a percentage o global total




25/45

5

4%

2%

12%

10%

6%

8%

0%

25%

20%

10%

5%

15%

45%

35%

40%

30%

0%

United Kingdom Canada Finland

Germany Japan United States

2004-2005 2005-2006 2006-2007 2007-2008

US share

of global

patients

Share of

global

patients

United States

3%

4%

2%

1%

10%

6%

9%

8%

7%

5%

0%

25%

20%

10%

5%

15%

50%

45%

35%

40%

30%

0%

United Kingdom Canada Finland

Germany Japan United States

2004-2005 2005-2006 2006-2007 2007-2008

Share of

global

patients

US share

of global

patients

United States

Figure 13: Share o global patients enrolled in phase II trials

Figure 14: Share o global patients enrolled in phase III trials

Source: CMR International a Thomson Reuters business

Source: CMR International a Thomson Reuters business


26/45

26

45. NIHR Clinical Research

Network Industry Metrics Q12010/2011.

46. NIHR Clinical ResearchNetwork Industry Metrics Q12010/2011.

47. Academy o MedicalSciences (2011) A newpathway or the regulationand governance o healthresearch, 2011. London:Academy o MedicalSciences. Retrievedrom: http://www.acmedsci.ac.uk/index.php?pid=47&prid=88 2011-01-13

48. See http://www.crncc.nihr.ac.uk/

Lie+sciences+industry/nwe/nwe_metrics

Recruited

to target

Targetpercentage

Over

recruited 200%

0% 200%

0%Under

recruited

New closed studies Previously closed studies

Closed early Ran to agreed timelines

Time percentage

Extendedrecruitmentwindow

Figure 15: Recruitment perormance o all closed commercial studies with NIHR Networksupport

Source: Reproduced rom NIHR Clinical Research Network Industry Metrics Q1 2010/2011

seems unlikely. It may also be that changesmade since the creation o NIHR have nothad any signifcant eect in the time periodcovered by this data, and that there has beenimprovement since 2008. However, NIHR dataon recruitment patterns indicates that this isan area o signifcant underperormance or

the UK, and one that needs to be monitored.Figures or 2010 indicate that more than500,000 participants took part in NIHR ClinicalResearch Network Portolio studies, which wasa 200,000 increase on the previous year.

The Clinical Research Networks within NIHRare monitoring recruitment to time and totarget or the studies within the Networks, andrecognise that too many studies are still ailingto recruit to target (although comparablefgures or other countries are hard to obtain).Just 50 per cent o industry trials which

closed in 2009/10 achieved their recruitmenttarget (this includes those studies that are notsupported by the Clinical Research Networks).45

Figure 15 represents the perormance o allclosed commercial studies run with NIHRClinical Research Network support to date. Itshould be noted that the studies represented

will have sites that are ully supported by theNetworks but may also have some sites withoutany Network support, and a number o studieswere adopted onto the portolio beore thesupport Networks were ully established.46

Despite the apparent increase in trials

approved, companies that want to do trials inthe UK say they fnd the process rustrating.The Academy o Medical Sciences publisheda detailed review o clinical trial regulationin January 2011, recommending a moreproportionate and symmetrical regulatorysystem to support health research.47 Thetime taken to process a set o approvals andpermissions is currently unpredictable. Delaysin approval mean that recruitment or a multi-centre trial is oten well underway in othercountries by the time UK permission to recruitis received. The UK sites thereore start out

behind, and i other centres reach and exceedtheir recruitment targets, the UK site may beclosed down beore the planned date, as it isno longer needed. Progress is being made, withdata rom the NIHR Clinical Research NetworkExemplar programme highlighting that seven othe 20 industry studies involved enrolled theirfrst global patient in the UK.48


27/45

7

Research that requires clinical subjectsnecessitates collaboration with hospitals andhealth centres. One indicator o the extento such collaboration could be an analysis olevels o trial activity in each trust. However,NHS Trusts have not previously had to publishtheir industry income levels or trial activity.

There is a new requirement introduced in2010 or NHS Trusts to publish the numbero patients recruited into clinical trials in theirquality accounts. This will hopeully increasethe attention that this orm o collaborationreceives.

2.6 How big is the opportunity?

I the collaborations between thepharmaceutical industry and universities and

between pharma companies and start-upcompanies could be improved, this would havehealthcare benefts as well as boosting boththe private and public research sectors.

This improvement would come rom twosources:

Private sector company growth improvedreturn on investment, opportunities or smallcompanies, greater research capacity.

Public sector income to universities, publicresearch institutes and the NHS.

I all universities improved to the level o thetop quartile, in terms o the percentage oresearch income rom industry, a 60 per centincrease in industry unding could be available.

As the pharmaceutical industry continuesto grow external R&D spend, there is hugepotential or growth by small companies aswell as universities. At current growth rates oalmost 15 per cent per annum, by 2015, therecould be an additional 3 billion in externalR&D unding available.

NHS participation in trials

UK patients enrolled in phase II and III clinicaltrials, as a percentage o worldwide patientsand in total fgures, have allen since 2004.I we recovered the UK share to 3 per centor both phase II and III, this would representapproximately an additional 3,000 patients peryear with access to trials. Kinapse estimatedin 2008 that the pharmaceutical industry

spent between 1 and 2 billion per year onclinical development in the UK. Comparingthe rates o trial approval with the decline in

patients indicates that there is vast scope orimprovement here. Given the National Instituteo Health Research budget o just under abillion, any change in this fgure could makea material dierence to research and trialsunding in the UK.


28/45

Part 3: How can we improve collaboration?

Collaboration is both important and necessaryor biomedical science. We have shown it can

improve the impact o research. The size o theprize or the UK is big, and will aect not justresearch income, but the quality o treatmentand opportunities or patients.

But those at the sharp end have identifedsome real barriers to such collaboration.The UKCRN Directors Forum held a meetingbetween representatives o the NHS, academicsand industry in 200949 where perceptions oworking together were discussed.

Academics and NHS participants expressedtheir need or true collaboration withindustry; access to industry knowledge onclinical pharmacology, regulation and projectmanagement; and partnership rather thanservice provision. Industrys concerns wereor aster contracting and better delivery byacademics, along with metrics to assess these;more accurate easibility assessments; andaccess to knowledge and patients.

These issues arent just about money orintellectual property ownership, which arecommonly cited as barriers to collaboration they are rustrations o process, ando understanding each others priorities.Interventions which make it easier or thoseinvolved in the process o biomedical discoveryto appreciate other perspectives should help toaddress these areas.

This chapter addresses the specifc incentivesthat could address these concerns, successullycreating alliances between the NHS,universities, charities and industry at the levels

o departments and individuals. It also looksat barriers that currently limit collaborationin these areas. These represent both specifc

areas or improvement and general principles osuccessul collaborations.

Successul collaboration cannot be mandated.It works best when created rom the ground up,with willing participants; imposing them romabove can be counter-productive. That said itis possible to create conditions that make iteasier, encouraging the frst conversations totake place. Mutual respect or other skills andexpertise is critical. Relatively simple measuresthat improve understanding can help thesepeer relationships to be created.

The recommendations are divided into threeareas: inrastructure, people and processes.

3.1 Create the right infrastructure

NHS research inrastructure has changedsignifcantly in the last fve years. The NationalInstitute or Health Research (NIHR) wascreated in 2006, and has brought in morecompetitive unding structures and signifcantinvestment in new acilities such as BiomedicalResearch Units and Biomedical ResearchCentres based in research hospitals. ClinicalResearch Networks have also been establishedto co-ordinate access to the researchinrastructure. Those in the industry seemunited in supporting the direction in whichNIHR has travelled.

However, the NIHR is the frst to acknowledgethat it needs to join up these pieces oinrastructure, to make better use o existingresources and make them more accessible to

industry. It has established its Ofce or ClinicalResearch Inrastructure (NOCRI) to ensurecoordination o NIHR-unded inrastructure

28

49. Experimental MedicineDirectors Forum (8 July2009) Exploring JointWorking between academiaand industry within theexperimental medicinearena. Retrieved rom:http://www.ukcrn.org.uk/index/clinical/experimental/d_orum.html


29/45

and to provide a navigation and collaborationservice or potential research customersincluding industry. There is also a vastinvestment that has been made by the researchcouncils, major medical charities and byindustry in UK research inrastructure. Linkingthese pieces up into a coherent system is much

more challenging, but could make the systemas a whole much more eective.

Beyond this major challenge, there are urthersteps that could improve the national researchinrastructure.

Deploy electronic patient records to support

research

The UK should exploit its potential to be aworld-leader in the use o electronic patientrecords to support medical research. So ar,research has been an aterthought in the

design o English electronic patient records.This is missing a major opportunity to improvethe competitiveness o England or research,and to encourage public participation andawareness. Scotland has made much moreprogress than England in this area (see Box 5).

Electronic medical records make aspects othe research process much easier and moreaccessible, and create new opportunities orinnovation. Electronic patient records havebecome a highly political and controversial

issue, at least in England and Wales. A recentreview o public engagement by the NewEconomics Foundation50 shows that the publicwant to be better inormed about the systemand its potential benefts. The huge cost o theNational Programme or IT (NPIT), along withuncertainty about which data would be stored,has made many people wary, i not actively

hostile, to the idea. However, the researchopportunities represented by a national recordssystem are signifcant, and the public aregenerally supportive o other eorts to improvecare through research. For example, the UKBiobank has now recorded data and clinicalsamples or hal a million volunteers, as a data

resource or long-term research.By restricting the Summary Care Record to avery minimal set o inormation, reinorcingthe right to opt-out and requiring opt-inbeore additional inormation is stored, thecurrent government hopes to draw a line underthe controversy. These changes are likely tomake the records less useul or research.Nevertheless, it will be important to inormpatients o the possible research benefts,and provide simple ways or them to allowtheir data to be used or research. In the New

Economic Foundation research, 74 per cent oadults supported the use o electronic patientrecords or research purposes, although amajority also believed that specifc consentshould be given.51

An electronics records system can contributeto research and patient care, both in improvingrecruitment or clinical trials, and as a researchtool in its own right. When designing a clinicaltrial, especially an early-stage trial, a researcherwill have a target number o recruits to

ensure that the results are statistically valid.I these patients cannot be recruited in thetime needed, the whole trial will become anexpensive ailure. Knowing which patients withthe right profle are available in a certain areais crucial to designing the trial and choosingwhere to run it.

9

50. new economics oundation(2010) Who sees what:Exploring public viewson personal electronichealth records. London:ne. Available at: http://www.neweconomics.org/publications/who-sees-what

51. Ibid.

52. Kullo, I.J., Fan, J., Pathak,J., Savova, G.K., Ali, Z.and Chute, C.G. (2010)Leveraging inormatics orgenetic studies: use o theelectronic medical recordto enable a genome-wide association study operipheral arterial disease.Journal o the AmericanMedical InormaticsAssociation. 17 (5),p.568, DOI: 10.1136/jamia.2010.004366; see also:http://www.sciencedaily.com/releases/2010/09/100908094803.htm

Box 4: Electronic medical records as a research resource the Mayo Clinic

A group o researchers rom the MayoClinic used the electronic medical recordsaccumulated by the clinic to conduct anassociation study on those diagnosed withperipheral arterial disease. The authorsobserved that the data provided them witha scalable solution for clinical research,

providing comparable and consistent data

that can be employed in comparativeeffectiveness studies, outcomes research or

translational research. With patient consent,the health records were used to confrmcases o the disease, in this case PeripheralArterial Disease, and extracted demographicdata and laboratory values. Inormationsuch as smoking status was derived romnatural language processing o the clinicalnotes. These approaches simplifed logistics,

costs and timelines relative to moretraditional genomic studies.52


30/45

With a large record set and detailed data,genetic association studies can be perormed,comparing links between symptoms, treatmentsand genetic results, just by analysing therecords (see Box 4).

The UK could lead the world in electronic

patient records. As a country with cradle-to-grave NHS care, simply recording all o apatients interactions with the NHS eectivelyprovides a complete medical history. One othe main requirements or a successul nationalrecords system, a unique patient identifer,already exists as the NHS number althoughit is not widely used, so at this stage cannot beeectively used to connect multiple episodeso treatment without the paper records beingsent back to the GP.

In contrast, other countries ace signifcant

difculties in implementing electronic patientrecords. The US does not have a nationalpatient identifer, which hinders the use oelectronic data or research.53 Some privatehospital groups such as the Mayo Clinicand Kaiser Permanente have implementedelectronic records very successully, but the

systems are not necessarily interoperable.A survey in 2009 ound that only 17 percent o US physicians were using either abasic or comprehensive electronic recordssystem.54 The closest the US has come isthe Veterans healthcare system, which usesa widely praised electronic records system

called VistA. The system provides electronicrecords, and order entry or the US VeteransHealth Administrations eight million patientsacross 153 hospitals and many more clinics.55The open-source sotware has been creditedwith some o the dramatic improvements inhealthcare quality at the VHA at a time whencosts have been kept down. The sotware hasbeen adopted by many other institutions acrossthe world.

In the international market or trials, thisis an area where the UK could show a real

competitive advantage. A ew countries havesuccessully implemented a national system,among them Denmark, Finland and Sweden,but they do not have the genetic diversityo the UK, or the same breadth o universityresearch which makes the UK such an attractivelocation or clinical trials.

30

53. See http://www.iti.org/fles/2009-leadership-healthit.pd

54. Jha, A.K. et al. (2009)Use o Electronic HealthRecords in U.S. Hospitals.New England Journalo Medicine. 360(16),pp.1628-1638. Available at:http://dx.doi.org/10.1056/NEJMsa0900592.

55. See http://www1.va.gov/VETDATA/Pocket-Card/4X6_summer10_sharepoint.pd

Box 5: Electronic patient records in Scotland

Scotland has an electronic patient recordssystem, where everyone registering with aGP is given a unique n

All together now: Improving cross-sector collaboration in the UK biomedical industry

Documents