Having worked within the pharmaceutical sector for more than a decade, Matheson writes from the perspective of an industry insider. Matheson’s article “Corporate Science and the Husbandry of Scientific and Medical Knowledge by the Pharmaceutical Industry” can be taken as an auto-ethnographic study informed primarily by his own experience and evidenced by cited studies and examples. In his paper, he argues that pharma shapes medical knowledge, not only by generating drugs, data, and devices, but also by contributing to medical-scientific discourse (Matheson 359).
Pulling back the curtain, Matheson tours us through the structural organization of the pharmaceutical sector, revealing a complex system of internal and external bodies “configured to the generation of revenue on a drug-by-drug basis” (359). The system is constrained implicitly by the values and norms of medicine, the “core maxim of science”, and explicitly through government regulation. Marketing and medical experts, including qualified doctors and scientists, navigate new drugs through this system. Together they sculpt the drug into a successful medical product by marshaling the data and discourse surrounding it.
Matheson is concerned primarily with the ways pharmaceutical companies understand their data and “regulate the understanding of clinicians and scientists,” with the ultimate aim of crafting a public disease-drug discourse or ‘drug narrative’ that serves market considerations (360). This discourse initially takes shape as a ‘product canon’, which he describes as a market-tested “set of propositions” outlining the company’s understanding of the drug, its connected biology, and its anticipated medical implementation. Importantly, Matheson describes how the ‘product canon’ informs the clinical development program, so that trials can be designed to confirm it. This feedback loop values trials that (unsurprisingly) support the ‘product canon’, so that “the company’s vision of its product becomes increasingly self-confirming” (360). Closer to market release, the ‘product canon’ becomes reified as a series of ‘key messages’ conveying patient needs, patient benefits, and superiority over competition – all of which comprise the core of the ‘drug narrative’.
Crafting a drug narrative involves identifying a medical ‘puzzle’ and positioning the emerging drug as its solution. Matheson writes that creating perceptions of an unmet need, medicalizing pre- and post-disease conditions, and reformulating conceptions of disease are tactics commonly deployed by medical and marketing experts to generate medical ‘puzzles’; he cites many such ‘puzzles’ that have arisen in the past decade (363).
Once the drug narrative is determined internally, it is expertly handled into the public discourse “dominated by academic science and clinical medicine,” who finally determine “the adoption of drugs by regulators, purchasers and prescribing clinicians” (363). Matheson claims that pharma influences both the content and structure of public discourse, and he provides a comprehensive chart outlining various techniques of persuasion. Contracting specialized publication planning agencies to secure numerous journal publications is a critical step toward shaping content, while procuring endorsement from KOLs, often through unrestricted educational grants, that allow pharma some control over the channels through which discourse proceeds.
Matheson adapts Kuhn’s seminal notion of normal science to summarize the effects of profit-oriented drug narratives. Where Kuhn identified a tension between innovation and tradition, Matheson finds a tension between innovation, tradition, and profitability. Matheson contends that all participants in the dispositif of medical and scientific knowledge practice a meta-science, which produces scientific knowledge that is overtly or implicitly directed toward some higher-order, external end. Yet Matheson suggests that pharma’s unique, corporate motivations result in a “compromising of standards of scientific truthfulness to a greater degree than in traditional academic science” (376). In this sense, the pharmaceutical industry conducts normative science, producing a mode of knowledge geared toward corporate objectives.
Hence Matheson proposes the creation of an International Standard of Integrity in Science. This organization, comprised of journals, professional societies and teaching institutions, would act as the global wardens of medical-scientific knowledge. Commercial interests will continue to shape medical-scientific discourse, Matheson concludes, but we should strive to make that discourse commercially transparent.
Matheson’s auto-ethnographic study stressed that publication planning agencies are critical in the development of new drugs, since these marketing/medical experts efficiently secure publication in medical journals. Sergio Sismondo thoroughly explores this area of the pharmaceutical industry in his paper “Ghosts in the Machine: Publication Planning in the Medical Sciences”. Herein, Sismondo reports his experience as an attendee at the third annual meeting of the International Society of Medical Planning Professionals, providing an ethnographic account of this microcosm of publication planning and supplementing his observations with many statistics, examples, and other studies.
Sismondo presents publication planning as the ‘ghost management’ of medical research and publication. He argues that clinical research has recently been directed towards marketing interests, claiming that pharmaceutical companies “have increased efforts to systematically treat research as a resource that needs to be carefully developed to affect the opinions of researchers and practitioners” (171). Indeed, Sismondo provides statistics indicating that the funding for research data has shifted from being mainly produced by academic research to being purchased by contract research organizations (176). This structural shift allots pharma greater control over data, which is often analyzed by in-house statisticians, and managed by a publication plan.
Publication plans consist of comprehensive strategies of information dissemination, and are designed to “extract more scientific and commercial value out of data and analyses, sometimes by designing studies with that value in mind, and always by carefully constructing papers that establish consistent profiles for drugs” (172). Plans begin as analyses that identify their audience, outline scientific and clinical ‘communication points’, and evaluate competitor publications, among other things. Plans orchestrate publication releases for maximum impact, recommending a publication pyramid where papers reporting on later stages of trials can reference and build upon earlier publications. The virtues of a publication plan are speed, foresight, efficiency; Sismondo suggests that they seem to promise “a world without uncertainty, of papers written and published on schedule” (176).
Yet Sismondo observes, however, that there are obstacles. Journal editors often inconsistently apply the Ingelfinger rule, which rejects manuscripts posting previously published results. Publications must also meet the authorship requirements that their papers – ghostwritten by a team of medical writers, but ‘authored’ by authority-conferring KOLS – can rarely meet, so pharma reacts by making the publication process less transparent. This discrepancy between appearance and reality characterizes publication planning generally.
Sismondo writes: “Theirs is the job of persuading without appearing to persuade” (193). Throughout his experience at the conference, Sismondo found publication planners were keen to enforce scientific – and, implicitly, ethical – standards and produce proper scientific data, but he notes that their plan’s analyzes, for instance, “are parts of an apparatus of interested intervention, not disinterested diffusion of results” (179). Journal editors, too, are part of this problem: Sismondo writes that they continue to treat the science mechanically, yet both editors and planners are aware that the science demands extensive decision-making.
In closing, Sismondo remarks that the ghost-management carried out by publication planners “might be seen to undermine the regulatory objectivity that the new mode of biomedical science produces” (192). Yet he argues that the knowledge produced should not be dismissed, but rather understood as “science done in a new, corporate mode” (193). He also contends that all medical research, analysis, and writing involve decisions in clinical trial design, interpretations of cultivated data, and decisions in the rhetoric used to communicate knowledge. While pharma-sponsored decision-making in medical science may be driven by commercial interests, it nonetheless produces knowledge – knowledge realized through sophisticated research supported by vast resources, and guided by planners who can fulfill their marketing objectives only by adhering to scientific norms.
Sismondo’s second piece, “How pharmaceutical industry funding affects trial outcomes: Causal structures and responses,” corroborates the two previous articles with many statistics that signal pharma’s influences over clinical research. In this piece, Sismondo analyzes the strong correlation between pharma-funded clinical trials and industry-favouring results.
He begins by explaining how sponsorship bias results from the direct and indirect actions of researchers. Publication planners, the primary direct channel through which pharma shapes medicine, manage the multitude of randomized, controlled clinical trials and craft research data into reports. CROs, who carry out much of this research, produce on pharma’s terms, and Sismondo argues that competing academic researchers feel pressure to do the same. Funding affects physicians indirectly through implicit obligations generated by gift giving. Sismondo claims that sponsored researchers likely have other relationships with their funding company, and these relationships change “habits and attitudes of thought toward the company and its products” (1911). FDA advisory committees are similarly influenced by these sorts of relationships, and “tend to support the interests of firms that support their work” (1911).
Sismondo discusses five causes explaining the association of funding and results. Industry-sponsored trials, while of high quality, are often biased through refined, context-specific design choices. When a drug trial is successful, it is often repeated to produce favourable, predictable results; he points to the contrast between the hundreds of publications on high-revenue drug, and the meager writings on those with lesser returns (1911). Thirdly, a survey reported CROs “prematurely terminating studies, changing protocols while studies were in progress, and writing first drafts of reports,” resulting in scientific misconduct (1912). Sponsored publications reporting on data collected by CROs then employ “interpretive and rhetorical choices to reflect their interests” (1912). Finally, he reports that journals earn revenue by selling reprints for drug promotion, resulting in “positive data being over-reported relative to negative data” (1912).
Sismondo proceeds to outline aimed at correcting these biases, He maintains all solutions are merely patch jobs; a proper solution, he says, would be to eliminate bias. Since bias results from the close-ties between the industry and researchers, Sismondo proposes they be isolated from each other. He concludes by acknowledging this separation is a challenging task, but a necessary one if we want to prevent pharm interests from influencing clinical research knowledge.