Joe Collier, Ike Iheanacho

 

 

Lancet 2002; 360: 1405-09

 

See Commentary

 

 

---

Medicines Policy Unit, Department of Pharmacology and Clinical Pharmacology, St George's Hospital Medical School, London SW17 0RE, UK (Prof J Collier FRCP); and Consumers' Association, London NW1 4DF (Prof J Collier, I Iheanacho MBBS)

 

---

Correspondence to: Prof Joe Collier, Medicines Policy Unit, Department of Pharmacology, St George's Hospital Medical School, London SW17 0RE, UK (e-mail:jcollier@sghms.ac.uk)

 

Information to steer the search for new products
Information required by regulators
Assessment of the validity of information submitted
Information published in the medical press
Generation of information and wider research implications
Impartial information provided by companies to users
Promotional information
When promotional information distorts
Conclusions
References

The pharmaceutical industry spends more time and resources on generation, collation, and dissemination of medical information than it does on production of medicines. This information is essential as a resource for development of medicines, but is also needed to satisfy licensing requirements, protect patents, promote sales, and advise patients, prescribers, and dispensers. Such information is of great commercial value, and most of it is confidential, protected by regulations about intellectual property rights. Through their generation and dissemination of information, transnational companies can greatly influence clinical practice. Sometimes, their commercially determined goals represent genuine advances in health-care provision, but most often they are implicated in excessive and costly production of information that is largely kept secret, often duplicated, and can risk undermining the best interests of patients and society.

Production of medicines is a data-rich business. Although the primary function of drug companies is to develop and market drugs, these companies spend more time and resources generating, gathering, and disseminating information.¹ These data are so important to the business interests of companies that most are regarded as commercially sensitive and remain confidential and protected by law. In this article, we consider the role of industry as informant, concentrating on the activities of transnational companies--ie, those with a global rather than a national or regional customer base--operating within the legislative frameworks of industrialised countries. However, such a focus offers a limited perspective of the global picture, partly because transnational companies are atypically large, with many having incomes greater than the gross national products of developing countries^2,3 (see paper in this series by Henry and Lexchin, to be published in The Lancet on Nov 16). Also, they represent only a few drug companies worldwide--of over 3000 companies affiliated with the European Federation of Pharmaceutical Industries and Associations,⁴ fewer than 100 are likely to be transnationals (Adrian Towse, personal communication). Moreover, only one in three countries has an effective legislative framework for the control of pharmaceutical promotion.⁵

A review of this nature has other limitations. For instance, little detailed material is published about the informational aspects of the industry, and what is available is rarely seen in the standard medical or scientific press. Instead, insights generally come from legal, governmental, or industrial reports, the lay media, or comments in scientific reviews, such sources that are difficult, if not impossible, to validate. Although recognising these caveats, here we review how the global industry generates and uses information to develop medicines, to convince regulators to allow its products to be marketed, to provide patients and healthcare professionals with impartial advice, and to promote sale of its products.

 

 

Transnational companies plan their research programmes for new drugs in accordance with predictions on future markets and the potential for creation of lucrative products. Such programmes, which industry presently estimates cost an average of around US$800 million for each drug⁶ (see paper in this series by Henry and Lexchin, to be published in The Lancet on Nov 16), typically take around 10-12 years to complete. Planning of the programmes relies crucially on having data about present pharmacological, pharmaceutical, molecular biological, technological, and therapeutic advances, global economic and epidemiological developments, and if possible, knowledge of competitor activity. All this background material has to be collated and the programme integrated into the company's particular area of expertise and its present (and projected) product portfolio. Most of these steering data remain confidential, a position defended in the European Union (EU), for example, by the courts.⁷

The programme proper comprises two phases, which together absorb most of a new drug's total development costs, leaving a relatively small proportion (less than 25%; Adrian Towse, personal communication) to cover the cost of production of the medicines themselves. The initial discovery phase involves investigative research in vitro and in animals. Again, very little of the information generated in this phase is made public, the main exceptions being material included in patent applications or disclosed during legal disputes. Moreover, to link any information that does reach the public domain with specific medicines could be difficult, not least because the product might be referred to by a code number rather than its subsequent brand or generic name (for example, propranolol, one of the original beta-blockers, was referred to in early publications as ICI 45,520).⁸ The discovery phase is followed by the development phase, in which research is undertaken in human beings, mainly in clinical trials. Most of the data generated during this phase again remain confidential, although some are included in published clinical reports.

Whatever the research programme, it must be sufficiently durable and flexible to respond to new challenges and insights; for example, sildenafil (Viagra) was noted to ameliorate erectile dysfunction only late in its research programme. After a product's launch into the marketplace, research continues but along different lines, for instance the company assesses how its product compares with other, new technologies and looks for new ways (indications) in which the product might be used in the future.

 

 

Regulatory authorities in industrialised countries set out in detail the information--and so the type and content of research--they require in the application for a product to be manufactured, sold, supplied, or promoted. Such requirements have in the past varied from country to country. Now, however, standardised positions have been agreed throughout the EU, and between the USA, the EU, and Japan, by development of the International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use (http://www.ich.org; see paper in this series by Abraham, to be published in The Lancet on Nov 9). In brief, authorities need detailed information on: the proposed use for the drug; in what dose and form it will be available; its quality (eg, consistency, dose, impurities, stability, contaminants, colourants, additives); details of its manufacture; its claimed therapeutic actions; and how it is handled by the body (eg, its absorption, metabolism, distribution, and excretion). Also needed are results of a battery of safety and toxicity tests which are designed, for example, to check for damage to organs, unwanted symptoms, development of cancers, damage to eggs or sperm, and the likelihood of damage to the fetus.

Although early work in vitro or in animals is mainly done in house or by contract research organisations (CROs), most research in human beings is commissioned out and done by independent clinicians and scientists working with patients or healthy volunteers in clinical trials. Whatever the form of the research, the company developing the product has a dominant role, because it (or a CRO working on its behalf) funds the study, designs the protocol, chooses the investigators, and in many instances (especially with multicentre trials), is involved in the collation, interpretation, and reporting of data. Throughout the research, companies are looking for any therapeutic features that can be used later as unique selling points for their products. Because the eventual promotional activity is so crucial, the company's marketing division may have a role in study protocol design.

It is the responsibility of the manufacturer (or its agent) to submit the licence application and supply the necessary supporting data. It must also propose what the medicine should be used for in practice (its indications) and draft the contents of impartial product information about it. Finally, the manufacturer suggests how the medicine should be made available--ie, whether it should be supplied prescription-only or sold over the counter. Ultimately, an application to market the product will be rejected if the authority judges that the supporting information is inadequate, untrue, or misleading. As a result of these requirements, vast amounts of data are submitted to each regulatory authority.

A new-medicine application to the UK authorities, for instance, will be presented in a dossier of about 250, 1-1·5 kg, A4-sized files, with the UK Medicines Control Agency's largest submission covering in total around 100 000 pages. Indeed, so much information is accumulated by transnational companies (the data submitted are only a portion of that generated) that their scientists are often unaware of, or cannot trace, earlier work on file, and as a result find themselves duplicating research (Rick Fuller, personal communication).

 

 

Assessing the basis for licensing decisions is difficult, not least because the full marketing submission is only seen by the applicant and the relevant regulatory authority or those working on behalf of the authority. Only a limited amount of data are made public--eg, in published reports of clinical trials, as a product's summary of characteristics or Patients' Information Leaflet, as the European Public Assessment Report for medicines licensed centrally by the EU, and as various expert US Food and Drugs Administration (FDA) reports. Also, the work of regulatory authorities is rarely amenable to scrutiny by independent experts, or their decision-making processes subject to independent (and public) audit (see paper in this series by Abraham, to be published in The Lancet on Nov 9). Although some information is available in the USA, it is not usually clear, for instance, how complete the submitted data are, or to what extent regulatory authorities refer to, or can use, data other than those submitted in the application. Each regulatory authority decides alone on the adequacy and reliability of the information, whether marketing of the product is compatible with public-health interests, and, if so, whether and how the product should be made available.

Some insights into quality of information submitted to regulators are available from internal reviews or from material published in journals, with around 30% of submitted clinical trial data published at the time of licensing, and 50% within 5 years (Bjorn Beerman, personal communication). For example, a review of the UK Medicines Control Agency has questioned the adequacy of checks made on the validity of information received from companies,⁹ and in the experience of the UK Drug and Therapeutics Bulletin, few clinical trials of new medicines are without important shortcomings, especially with respect to trial design and data interpretation (see below).

 

 

Drug companies recognise the enormous value of publishing clinical trial reports about their products in medical journals, especially when the journal is prestigious.¹⁰ Such publications are important because they raise awareness of products. However, to improve sales, it is also crucial that the published report shows the company's product in a favourable light. Publication is especially helpful if the article is published around the time of the product's launch. Echoing these aspirations, trials with negative results tend to be published much later than those with more positive conclusions.¹¹ Moreover, the conclusions of trials sponsored by drug companies, rather than by other sources, tend to be more favourable to the sponsor's product.^12-14

No clear explanation is available for the preponderance of positive company-sponsored studies. Possibilities include inherent biases in trial design,¹⁵ for instance, by use of inappropriate comparator drugs,^15,16 drug doses,^16,17 or methods to assess outcomes.^18,19 Occasionally, in attempts to ensure a positive bias, companies have threatened legal action to stop nominally independent researchers from publishing negative material.^20-23 Moreover, researchers who communicate negative results have faced intimidation, efforts to discredit them professionally, and threats of legal action to recover the value of lost sales.²⁴ Often, the potential for such company intervention is written into the researcher's contract. For example, in a sample of US research centres, around 30% of contracts with researchers allowed the sponsoring company to delete information from the report and to delay publication.²² Also, participation in an academic-industry relationship and commercialisation of university research are both associated with delays in publication.²⁵

One initiative to address such sources of bias has been introduced by journals that are members of the International Committee of Medical Journal Editors, including the BMJ and The Lancet. These journals now require contributors to disclose details of their own and their funders' roles in studies.^26,27 Such contributors have to declare that they accept full responsibility for how the study was conducted, had access to the data, and controlled the decision to publish.

 

The predominance of positive results in published reports might also be associated with the income journals receive from drug-company advertising. As a result, some editors might find it difficult not to comply with industry's wishes to publish material, no matter how poor the study might be, or to change the negative wording of trial reports submitted by independent researchers. Whatever their origin, trials with positive results are more likely than others to be: selected for presentation at scientific meetings; reported in print; published promptly; published as full reports; published in widely-read journals; published in English; published in more than one report; and cited in related reports.²⁸

 

 

The pharmaceutical industry is the single largest sponsor of medical research--and thus the largest generator of related information--in, for example, Canada,²⁹ the USA,³⁰ and the UK,³¹ and in some countries it is the only realistic source of such support. Industry-funded research is overwhelmingly drug-oriented, so it could distort a country's research effort by, for instance, diverting resources and personnel away from non-drug interventions or other aspects of health care. Moreover, as increasing numbers of medical researchers are drawn to the industry, alternative voices and opinions can become muted, and novel avenues of research might be overlooked.

Is is widely accepted that 90% of the world's health problems (mainly those of developing countries) attract only 10% of global health resources (the so-called 10/90 health-research gap),^32,33 and industry's investment pattern might contribute to this discrepancy. Transnational companies have traditionally ignored work on treatments for diseases that occur mainly in developing countries, because they know even if appropriate treatments were produced, they would not bring adequate returns.³⁴ A second potential distorting factor resulting from patterns of research is so-called evidence capture, whereby health-care policies blindly follow published evidence irrespective of whether the information provided is reliable or appropriate to the needs of society at large. Industry's research interests and activity dominate research outputs so much--eg, trials on the value of antiarthritic drugs far outnumber those on walking sticks--that the sheer weight of new evidence, rather than its quality, might lead to inappropriate shifts in therapeutic practice away from tried, familiar, and usually cheaper approaches, to novel, unfamiliar, and generally more expensive alternatives that offer no real clinical advantage.

 

 

In many industrialised countries, it is standard for drug companies to provide impartial information about their products to health-care professionals and patients (see section on information required by regulators). Such information includes summaries of product characteristics, packaging and labelling details, and Patients' Information Leaflets. This material, which is drafted, published, paid for, and distributed by the company, is usually required by law, has legal status, and is mainly controlled by regulatory authorities. In some countries, the summary of product characteristics might be the only impartial information available to those prescribing or taking medicines, and so it is of great value. However, information provided in this way still has inevitable limitations, since it can only refer in detail to the accompanying product and so cannot advise on its value compared with other medicines or non-drug treatments. Moreover, summaries of product characteristics can fail in their primary public-health objective to inform clinicians and patients, by offering advice that is unclear or impractical.³⁵ Part of the responsibility for such failure lies with the regulatory authority itself.

 

 

By announcing a medicine's availability and its potential use, advertising can be valuable to prescribers and patients. It is also an important component of drug-company business, with, for instance, much more money spent on promotion than on research and development,³⁶ or on provision of impartial information. In the UK (one of the few places where advertising spend is capped), adherence to the Pharmaceutical Price Regulation Scheme (PPRS) means that for every pound spent by the industry on impartial information, around five pounds will be spent on promotional activity.

Work on promotion starts at the trial-design stage, when the company plans how to obtain information that might point to unique selling points for the new product. Once these therapeutic advantages have been defined and the company feels the product will obtain marketing authorisation, the campaign proper begins. In transnational companies, such campaigns will be carefully orchestrated so that messages remain consistent when presented to different audiences over time, and from country to country--eg, in media outlets and printed promotion, as relayed by drug-company representatives or during organised symposia.

The main stages in any promotional campaign for a new product concentrate on dissemination of information rather than its generation. Usually, the first step is to increase the target audience's awareness of a disorder and to highlight present deficits in its treatment. For example, one announcement coupled the observation that "psychologists estimate that one in 15 people suffer from acute social awkwardness" with the fact that "a new drug that relieves shyness could be available within a year".³⁷ Having persuaded people of the treatment need, the next stage is to introduce the new medicine, telling of its unique advantages. The ultimate aim is to induce clinicians to prescribe the product and for familiarised members of the public to ask for it. Ultimately, a successful campaign will develop, secure, and reinforce brand loyalty, objectives similar to promotional campaigns for longer established products.

Traditional channels for promotion of prescription-only products includes advertisements, mailings, and visits to clinicians by drug-company representatives--New Zealand and the USA also allow promotion direct to the public. Furthermore, the company's marketing department--working with outside agents such as public-relations companies and medical education agencies--will try to ensure that supportive articles--ideally editorials--are placed in relevant specialist journals; products are endorsed by opinion-leaders and journalists;^38,39 product-labelled gifts (eg, pens, notepads, calendars) are displayed by health-care professionals and distributed at events (eg, symposia, postgraduate educational meetings); the lay press publishes disguised promotional material;⁴⁰ and products are championed by consumer (patients') support groups.

These overtly promotional methods are bolstered by more indirect leverage. For example, drug company-sponsored science-teaching programmes in schools (now commonplace in the UK) can induce loyalty to industry at an early age and encourage the idea that medicines are necessarily good solutions to ill health. Funding of prizes, scholarships, or travel bursaries to students, scientists, and clinicians also inevitably engenders company loyalty.

Regulatory authorities are ultimately responsible for ensuring that promotional activity is at least accurate, truthful, balanced, up-to-date, unlikely to mislead,⁴¹ and consistent with the summary of product characteristics or its equivalent. However, in many countries, regulatory authorities are either absent or ineffective, and in industrialised countries, they typically devolve much of the policing to the industry itself. Perhaps, as a result, inappropriate promotion is commonplace. However strict and well-policed the controls in individual countries, cross-border promotion offers a particular challenge. Regulators can find it difficult to deal with promotional material from another country, where the relevant legislation is different to that in their country or where promotional claims have no traceable or verifiable source, as can arise with advertising on the internet.

 

 

There is no doubt that medicines offer enormous health benefits; however, a crucial issue for patients and society at large is that treatment choices are made rationally, with patients receiving the best available medicines given in such a way that fullest advantage is ensured. These goals are unlikely to be met if the reasons for prescription are distorted, and end up tipping the balance away from patients' interests and towards those of the pharmaceutical industry. The more effective the advertising and any accompanying promotional devices, the greater the potential for such distortion. Rational prescribing is inevitably threatened when, for example, opinion-leaders are briefed, promoted, cultured, and supported by manufacturers;³⁹ when patients' advocacy groups are funded and supported by the industry (eg, in the UK, Pfizer, manufacturer of sildenafil, supports the Impotence Association [http://www.impotence.org.uk] and the Men's Health Forum [http://www. menshealthforum.org.uk]); when, as in many countries, there are few, independent, non-commercial sources of information, leaving prescribers heavily reliant on drug-company representatives for their information;⁵ when controls of promotion are weak and promotional excesses greatest;⁴ when claims in drug advertising are not supported by the accompanying bibliographical references;⁴² or when most young doctors have gifts endorsing particular companies or their products.⁴³

 

Although any company could, in theory, exert influence by these means, the transnational companies have the resources to do so in an organised manner, and it is they who tend to set the norms. Moreover, these companies can develop and promote their products worldwide, with campaigns to recruit and influence opinion-leaders and journalists.⁴⁰ Finally, large companies often have enough influence to give them direct access to government policy-making. For instance, it is difficult to see how a government could resist pressure from companies with revenues that exceed their country's gross national product. Presently in the UK, representatives of the pharmaceutical industry and ministers from five governmental departments meet regularly as part of a Pharmaceutical Industry Competitiveness Task Force, to develop policies designed to strengthen the competitiveness of the UK business environment for the innovative pharmaceutical industry.⁴⁴

 

 

The pharmaceutical industry--and in particular large transnational companies--generates and collates vast amounts of information. Much of this material remains secret or is shared exclusively with regulatory authorities. A small proportion is publicly available, mainly as published clinical trials, promotional material, and impartial treatment advice directed to prescribers and patients. Through their investment in research, transnational companies have an important effect on the direction of medical research generally; via their promotional and educational activity, they are probably the biggest individual influence on prescribing practice. For the pharmaceutical industry, investment in information is time and money well spent. However, the huge scale of work involved, lack of openness, accompanying duplication, and distortion of the overall research effort and resulting messages make the business of information-generation inefficient and threatens patients' interests.

 

 

Joe Collier and Ike Iheanacho are Editor and Deputy Editor, respectively, of Drug and Therapeutics Bulletin.

 

 

1 Danzon PM. Trade and price differentials for pharmaceuticals: policy options. London: Office of Health Economics, 1997.

2 Laing R. Pharmaceutical company reports. http://dcc2.bumc.bu.edu/richardl/ARV_Drug_Prices/sld008.htm (accessed Sept 23, 2002).

3 O'Reilly B. Death of a continent.  Fortune 2000; 142: 258-76. [PubMed]

4 European Federation of Pharmaceutical Industries and Associations. 2001/2002: the year in review. Brussels: EFPIA, 2001.

5 Mintzes B. Blurring the boundaries: new trends in drug promotion. http://www.haiweb.org/pubs/blurring/blurring.intro.html (accessed Sept 23, 2002).

6 Anon. Cost of development soars to $802 million per drug. Scrip 2001; 2072: 15.

7 McKee M, Mossialos E, Baeten R. The impact of EU law on health care systems: work and society, vol 39. Brussels: Peter Lang (in press).

8 Drug and therapeutics bulletin and the national prescribing centre. Medicines and the NHS: a guide for directors, book 2--how medicines get to market. London: Which, 1997.

9 Chief medical officer. Withdrawal of an oral polio vaccine: analysis of events and implications. London: Department of Health, 2002.

10 Practical guide number 1, practical guide to medical publishing. Pharmaceutical Marketing, November, 2000.

11 Stern JM, Simes RJ. Publication bias: evidence of delayed publication in a cohort study of clinical research projects.  BMJ 1997; 315: 640-45. [PubMed]

12 Davidson RA. Source of funding and outcome of clinical trials.  J Gen Intern Med 1986; 1: 155-58.  [PubMed]

13 Kemmeren JM, Algra A, Grobbee DE. Third generation oral contraceptives and risk of venous thrombosis: meta-analysis. BMJ 2002; 323: 131.

14 Kjaegard LL, Als-Nielsen B. Association between competing interests and authors' conclusions: epidemiological study of randomised clinical trials published in the BMJ.  BMJ 2002; 325: 249-52. [PubMed]

15 Montaner JSG, O'Shaughnessy MV, Schechter MT. Industry-sponsored clinical research: a double-edged sword.  Lancet 2001; 358: 1893-95. [Text]

16 Ray WA, Griffin MR, Avorn J. Evaluating drugs after their approval for clinical use.  N Engl J Med 1993; 329: 2029-32. [PubMed]

17 Rochon PA, Gurwitz JH, Simms RW, et al. A study of manufacturer-supported trials of nonsteroidal anti-inflammatory drugs in the treatment of arthritis.  Arch Intern Med 1994; 154: 157-63. [PubMed]

18 Altman DG, Schulz KF, Moher D, et al. The revised CONSORT statement for reporting randomized trials: explanation and elaboration.  Ann Intern Med 2001; 134: 663-94. [PubMed]

19 Hollis S, Campbell F. What is meant by intention to treat analysis? Survey of published randomised controlled trials.  BMJ 1999; 319: 670-74. [PubMed]

20 Spurgeon D. Report clears researcher who broke drug company agreement. BMJ 2001; 323: 1085.

21 Nathan DG, Weatherall DJ. Academia and industry: lessons from the unfortunate events in Toronto.  Lancet 1999; 353: 771-72.  [Text]

22 Rennie D. Thyroid storm.  JAMA 1997; 277: 1238-43.  [PubMed]

23 Editorial. The tightening grip of big pharma. Lancet 2001; 357: 1141.

24 Quick J. Maintaining the integrity of the clinical evidence base. Bull World Health Organ 2001; 79: 12.

25 Blumenthal D, Campbell EG, Anderson MS, Causino N, Louis KS. Withholding research results in academic life science: evidence from a national survey of faculty.  JAMA 1997; 277: 1224-28. [PubMed]

26 Smith R. Maintaining the integrity of the scientific record. BMJ 2001; 323: 588.

27 Davidoff F, DeAngelis CD, Drazen JM, et al. Sponsorship, authorship, and accountability.  Lancet  2001; 358: 854-56. [Text]

28 Chalmers I. National Institute for Clinical Excellence: second report of session 2001-2002, volume II. London: Stationery Office, 2002.

29 Canada's research-based pharmaceutical companies. Our commitment to innovation in Canada: over $7 billion since 1987. http://www.canadapharma.org/Media_Centre/News_Releases/2002/Feb12-02_e.html (accessed Sept 23, 2002).

30 Anon. How Government and the Rx industry cooperate for the benefit of patients. Washington: Pharmaceutical Research and Manufacturers of America, 2002. http://www.phrma.org/publications/documents/backgrounders//2002-02-12.343.phtml (accessed Aug 8, 2002).

31 ABPI. Pharmaceutical industry issues: research and the science base. http://www.abpi.org.uk/publications/publication_details/pharIdustryIssues/pii9.asp (accessed Sept 23, 2002).

32 Ramsay S. No closure in sight for the 10/90 health-research gap. Lancet 2001; 358: 1348.

33 Global Forum for Health Research. The 10/90 report on health research 2000. Geneva: Global Forum for Health Research, 2001.

34 Schieppati A, Remuzzi G, Garattini S. Modulating the profit motive to meet needs of the less-developed world.  Lancet 2001; 358: 1638-41. [Text]

35 Anon. Failings in treatment advice, SPCs and black triangles.  DTB  2001; 39: 25-27. [PubMed]

36 Anon. The potential of DTC advertising. Pharm J 2001; 267: 361. http://www.pharmj.com/Editorial/20010915/forum/advertising.html (accessed Sept 23, 2002).

37 Don't be shy. Pharmaceutical Marketing. January, 2002: 1311.

38 Jackson T. Are you being duped? BMJ 2001; 322: 1312.

39 Practical guide number 6, practical guide to medical education. Pharmaceutical Marketing, May, 2001.

40 Sweet M. Sponsored journalism award shocks Australian media. BMJ 2001; 323: 1258.

41 World Health Organization. Ethical criteria for medicinal drug promotion. Geneva: WHO, 1985.

42 Villanueva P, Peiró S, Librero J, Pereiró I. Accuracy of pharmaceutical advertisements in medical journals. Lancet (in press).

43 Sigworth SK, Nettleman MD, Cohen GM. Pharmaceutical branding of resident physicians.  JAMA 2001; 286: 1024-25 [PubMed]

44 Pharmaceutical Industry Competitiveness Task Force. Terms of reference. http://www.doh.gov.uk/pharmaceuticalsindustry/index.htm (accessed Oct 16, 2002).