A portrait of commissioned literature: connecting the dots between medical education and communication companies, authors, funders and publishers Authors: Maud Bernisson Published: 2026 ABSTRACT BackgroundMedical education and communication companies (MECCs) are key actors in the production of commissioned publications. Analyzing publications that acknowledge medical writing offers a broad overview of their role, which is scarce in the literature. The reason is that activities related to industry-funded research are highly opaque, which makes access to this information particularly hard, and related analyses scarce. This article offers a novel approach to collecting, and thus, analyzing data. It maps how MECCs, funders, authors, and publishers organize to produce this literature and identifies risks of research integrity breaches.MethodsThe database contains the metadata of 29 911 commissioned papers collected from Web of Science (WoS). All articles involving medical writing were collected if they mentioned “medical writing” in WoS, if they were published by MECCs websites, and if they were listed in publication trackers (documents that list the publications of companies, found on the Industry Documents Library (IDL)). The metadata were collected semi-automatically (manually and with R). Reference lists were built to extract MECCs from acknowledgments and harmonize organizations’ names. The analysis maps the relationships among MECCs, authors, funders, and publishers and shows how they connect.ResultsIn contrast with pharmaceutical companies and research institutes, only 33.6% of MECCs are listed in the bylines and even less (18%) are acknowledged in the publications found on their websites and on the Industry Documents Library. Even though they provide writing, MECCs are usually not considered as authors.Medical writing is a flourishing business and MECCs are key actors. Only a few sponsors (5.5%) outsource medical writing to many MECCs, and a few MECCs (7.7%) have a high number of clients. The medical writing market is thus very competitive and can benefit sponsors. MECCs sometimes publish in journals that are owned by their parent or sister companies. The business model of the publishers largely benefits from medical writing by publishing their product, and sometimes providing it as a service.ConclusionsThe database presented in this article offers new routes to explore the key role of MECCs in industry-funded research and to further assess the commissioned literature and its impacts on biomedical science and research integrity. Introduction When a medical writer is the author but not acknowledged as such in a publication, ghostwriting is often considered the core issue. Medical writers can appear in acknowledgments but rarely achieve the status of byline authors in publications. Acknowledgments fail to adequately report the role of each actor involved in the production of scientific communications, but somewhat succeed in reporting the conflicts of interest of researchers as required by most publication guidelines. In its broadly accepted definition, a conflict of interest solely concerns individuals. It occurs when a person “(a) is in a relationship with another requiring the exercise of judgment on the other's behalf and (b) has a (‘secondary,’ ‘unusual,’ or ‘adverse’) interest tending to interfere with the proper exercise of such judgment” [1]. To date, the most effective tools to tackle conflicts of interest revolve around transparency, which is largely insufficient to address risks of integrity breaches stemming from industry-funded research [2, 3]. The organization of industry-funded research enables “institutional practice[s] that ha[ve] epistemic consequences” [4]. Conflicts of interest do not concern just an individual but most actors involved in the production of industry-funded research. A concrete example is the ghost management of medicine [5], which describes how pharmaceutical companies can supervise a broad set of practices and actors involved in the many tasks required for research. Focusing on medical writing provides an opportunity to shed light on these practices. Pharmaceutical companies commission papers either in-house or outsource them to medical education and communication companies (MECCs). MECCs are paid to design and implement scientific communication strategies, ranging from continuing medical education to publication plans. Publication planning involves framing arguments and communicating them through “scientific platforms within the biomedical literature,” which thus includes scientific articles written by medical writers [5]. Clients oversee MECCs, which manage publication plans and researchers who author, legitimize, and bear accountability for the publication. In exchange, researchers benefit from advantages offered by pharmaceutical companies. As a result, the role of the MECC described in the final publication is confined to medical writing, in accordance with the definition of the pharmaceutical and medical writing industries: a medical writer “supports the work of individual publications” [6]. Based on empirical evidence or personal experience, researchers found that medical writing shares the criteria of ghostwriting [7, 8], which is only the tip of the iceberg [5]. Industry-funded research has a structural impact on science, and consequently, its results [9]. For example, industry-funded research is subject to a publication bias, that is, the tendency to publish positive results more often than negative results; in addition, industry-funded research reaches more positive conclusions about the new drug it has developed than independent research [10], a phenomenon also called the “funding effect” [9]. Based on empirical evidence, often discovered in internal documents of tobacco, chemical, food, fossil fuel, and pharmaceutical companies released following US litigation, additional studies have found evidence of several types of “corporate manipulation of research,” which involved medical writing [11, 12]. Analyses of similar types of evidence have revealed practices like sugar-coating drugs in scientific narratives and the scientification of marketing arguments in scientific communication [13–16]. They also illustrate how publication planning intertwines medical writing with marketing [5, 17]. These practices are not only individual but also stem from the organization of the pharmaceutical and medical writing industries to produce scientific communication. A variety of companies, including publishers or contract research organizations (CROs), have started offering scientific communication services [5], driven by and contributing to the boom in this sector. When a publisher owns a MECC, risks of research integrity breaches increase. For example, a 2009 court case revealed that the Australasian Journal of Bone and Joint Medicine, a peer-reviewed journal owned by Elsevier, contained articles that were essentially disguised advertisements for Vioxx (Merck) [18]. Elsevier’s MECC, Excerpta Medica, was found to be involved, which led the publisher to sell it in 2010 [5]. Excerpta is now part of the Adelphi Group, a major scientific communication group. While this is an extreme example, the ownership of a MECC by a publisher, whose business model partly depends on high publication rates, raises concerns about research integrity breaches. The medical writing industry is a prominent actor in industry-funded research and is thus expected to produce thousands of articles every year. The opacity surrounding its activities makes it difficult, if not impossible, to provide a hard number that represents the extent of the literature this industry produces. A few studies, however, estimate the share of the ghostwritten literature. For example, Gøtzsche et al. [19] selected 44 industry-initiated trials approved between 1994 and 1995 by Danish ethics committees and compared the authors of the research protocols, the individuals who performed the analysis, and the authors of the publications. They found discrepancies, and thus ghost authorship, in 75% of their dataset; Flanagin et al. [20] surveyed authors about their role in writing publications they authored, and found that 11% of the papers published by three main and three smaller biomedical journals in 1996 were ghostwritten. Based on an internal document (a list of publications by a MECC) made available during a lawsuit involving Pfizer, Sismondo [21] found that between 18 and 40% of articles about Zoloft (sertraline) were managed by MECCs, hired by Pfizer. These results are old and vary due to differences in methods and in the scope of the studies, hence the importance of drawing a more contemporary portrait of this type of literature. This paper offers a new method for investigating sponsored-research literature. It includes the construction of a database of 29 911 references of commissioned papers [22]. Its analysis describes how medical writing configures the relationships of the actors involved in the scientific publication industry. MECCs constitute the core of the analysis, which highlights their relationships with other actors involved in scientific publishing and identifies the risks of research integrity breaches stemming from these relationships. The first inquiry focuses on authorship. The second traces the collaborations between funders and MECCs. The last inquiry features the relationships between MECCs and publishers. Methods The database of commissioned papers encompasses the DOIs of references from online resources: 229 articles from publication trackers (documents listing commissioned publications produced for a company) discovered in the Industry Documents Library1 (IDL database); 4 555 publications made available on the websites of MECCs and CROs (MECCs database); and 26 858 articles from Web of Science that mention “medical writ*” in the acknowledgments (WoS database). DOIs of these publications from Web of Science were then collected, if they existed and were indexed, with semi-automatic methods (see Table 1).Table 1Number of DOIs found in the Industry Documents Library and on the MECCs websites by company(a) DOIs found with publication trackers of companies in the Industry Documents LibraryCompanies#References found#DOIs found on WoSMcKinsey (Purdue)5137TEVA12515JUUL245Mallinckrodt388146Philip Morris13826Total726229(b) DOIs found on the websites of the MECCs that make their publications publicly availableRTIHS86122709Open health678557IQVIA3599519COS719361Cellworks15690GKM14980Kstrategic4847RBCCons11647KJT5748Alcedis4332MC Analytics3817Costello18017Boston2016Gouya Insights1915Csur2112Total14 4554567 Table 1Number of DOIs found in the Industry Documents Library and on the MECCs websites by company(a) DOIs found with publication trackers of companies in the Industry Documents LibraryCompanies#References found#DOIs found on WoSMcKinsey (Purdue)5137TEVA12515JUUL245Mallinckrodt388146Philip Morris13826Total726229(b) DOIs found on the websites of the MECCs that make their publications publicly availableRTIHS86122709Open health678557IQVIA3599519COS719361Cellworks15690GKM14980Kstrategic4847RBCCons11647KJT5748Alcedis4332MC Analytics3817Costello18017Boston2016Gouya Insights1915Csur2112Total14 4554567 Number of DOIs found in the Industry Documents Library and on the MECCs websites by company I downloaded the BibTeX files of the final list of DOIs on Web of Science, and converted these files into a dataframe with the R library Bibliometrix, which resulted in data loss (12 DOIs from the MECCs database). The final database contains 31 642 items, including duplicates, or 29 911 unique DOIs.2 25 423 of the 29 911 unique publications acknowledge medical writing. The earliest paper was published in 1972 and the latest in 2025 (see Figure 6 in Appendix 1). BibTeX files do not provide a list of MECCs but of acknowledgments. Bibliometrix does not identify MECCs either. I thus built and used a reference list of MECCs to extract the companies mentioned in the acknowledgments of the papers with the R library stringr. The database needed further cleaning, and I built additional reference lists to harmonize the names of publishers, affiliations, and funders in R (see Appendix 2). In addition to Bibliometrix, I used dplyr and tidyr for the analysis, and ggraph and ggplot2 to visualize the results (Appendix 1 describes the method in greater details). Results Authorship The first line of inquiry investigates authorship through authors’ affiliations of the papers in connection with their funders and MECCs to explore transparency practices and the distribution of responsibilities among the actors involved in the production of papers. Top funders are pharmaceutical companies. Pfizer sponsored 9.9% of all unique publications (2 975 of 29 911), followed by Novartis, which funded 6.6% publications (n = 1 971), and AstraZeneca, which sponsored 6.5% publications (n = 1 949). It can either indicate that these companies implement better transparency practices than others, or that they subcontract MECCs more often to produce publications. Pharmaceutical companies often play a larger role than mere sponsors, as suggested by Fig. 1, which visualizes the relations between MECCs, affiliations of (co-)authors and funders in the 25 423 publications acknowledging MECCs and funders. Fig. 1Top collaborations among MECCs (if acknowledged in the publications), researchers (through their affiliations) and funders (threeFieldsPlot() from Bibliometrix 4.3.2) Top collaborations among MECCs (if acknowledged in the publications), researchers (through their affiliations) and funders (threeFieldsPlot() from Bibliometrix 4.3.2) The figure suggests that pharmaceutical companies’ employees often (co-)author publications sponsored or co-sponsored by their employers, even though MECCs are involved to provide medical writing. The figure also suggests strong collaborations between MECCs and sponsors. MECC employees are clearly under-represented in the affiliations (only Research Triangle Institute stands out), indicating a lack of transparency on the role of MECCs, which mostly appear in the acknowledgments. Excluding the publications found on the WoS from the analysis shows a different dynamic. Only 18% (868 of 4 784) of the publications found in the IDL and on the websites of MECCs mention a MECC in the acknowledgments (see Appendix 1, p9). In addition, MECCs appear in the byline of 33.6% (n = 1 609) of these publications, and especially the Research Triangle Institute, which appears in the byline of 1 117 of its 2 709 publications. The affiliations displayed in the articles of the database highlight and question the role played by research institutes, through their researchers, in contributing to the collaborations between MECCs and sponsors. In Fig. 2, the 83 nodes represent authors’ affiliations that appear more than 4 500 times in co-authored publications (23 753 publications are co-authored). The network includes 10 pharmaceutical companies (“Samsung” refers to “Samsung Bioepis,” specialized in biosimilars) and 73 research institutes, which constitute 88% of the nodes. The most opaque nodes represent the highest numbers of collaborations. In other words, Pfizer’s employees are the most collaborative authors among pharmaceutical companies, while authors affiliated with Harvard University, the University of California System, the University of Texas System, and the University of London, are the highest collaborators among research institutes. The opacity of edges also represents the highest numbers. Edges are distributed evenly between pharmaceutical companies and research institutes, except for a few that tend to collaborate closely, such as Harvard University with Harvard Medical School. However, affiliations in this figure can also belong to the same researcher. The even distribution shows, like Fig. 1, a common use of researchers affiliated with research institutes to author these papers. In addition, no MECC appears in this network of co-authored papers, which, again, questions the potential misalignment between the role performed by a MECC and the role attributed to this MECC in the final paper. Fig. 2Network of co-affiliations (ggraph 2.2.1) Network of co-affiliations (ggraph 2.2.1) This configuration made of the acknowledgments of MECCs in publications, affiliations of independent researchers and employees of pharmaceutical companies, and pharmaceutical companies as sponsors probably indicates publication planning. MECCs and funders Among the 25 423 publications acknowledging MECCs, one publication can be funded by one or by several sponsors and it can acknowledge the work of one or several MECCs. A methodological approach can thus consider the relationships between individual funders and individual MECCs or the relationships between several funders and several MECCs to produce a publication. The first approach provides insights into the recurrence of collaborations between one funder and one acknowledged MECC across publications. In the database, 1 046 MECCs received funding from an average of 14.4 funders, but with a median of 3. In fact, 81 MECCs (7.7%) received funding from more than 50 different funders (Figure 15 in Appendix 1 visualizes the quantiles of this dataset). Regarding funders, 1 123 of them funded publications acknowledging an average of 13.4 different MECCs, with a median of 3. Only 62 funders (5.5%) funded more than 50 different MECCs acknowledged in publications (Figure 16 in Appendix 1 visualizes the quantiles of this dataset). To summarize, a small group of pharmaceutical companies subcontracted a small group of MECCs. This approach fails to represent the rate of production of publications by funders and MECCs because it counts every publication by unique funder and by unique MECC. Thus, if several funders or several MECCs have collaborated, the same publication will be counted several times. For example, Sanofi alone funded 155 publications that acknowledge Excerpta, and Regeneron alone funded 138 publications that acknowledge Excerpta. However, 119 of these publications, which acknowledge Excerpta, are the result of a collaboration between Sanofi and Regeneron. Considering co-funders and MECCs collaborating to produce a publication can overcome this issue. Figure 3 shows the most recurrent co-funding collaborations, such as the collaboration between Sanofi, Regeneron, and Excerpta, and captures rates of publications. Figure 3 represents 2 852 publications (9.5% of the 29 911 unique publications in the database) and displays 47 collaborations between funders and MECCs, which led to the production of more than 30 publications. The MECC InScience Communications, part of Springer Nature Group, has the highest number of clients (4). Pfizer is the company funding the most MECCs in this graph (7). These numbers provide insights on the business models of the MECCs, which seem to depend on a few clients. These findings also suggest a recurrence of collaborations between major pharmaceutical companies and MECCs, which also probably indicates publication planning.Fig. 3Number of publications by funder(s) and by MECC(s) acknowledged (ggplot2 3.5.1) Number of publications by funder(s) and by MECC(s) acknowledged (ggplot2 3.5.1) MECCs and publishers Mapping the MECCs acknowledged in publications together with the publishers of those publications shows which one publishes commissioned papers the most. MECCs primarily target major and legitimate publishing avenues: Wiley, Elsevier, Springer, and Taylor and Francis are the main publishers of this literature, and some of these publishers belong to the same parent company (see Figure 19 in Appendix 1). For example, Springer Nature acquired BioMed Central (BMC) in 2008. Taylor and Francis acquired Dove Medical Press in 2017 and Future Medicine, part of Future Science Group, in 2023. Figure 4 represents 2 941 of the 25 423 publications acknowledging MECCs. The most acknowledged MECC in this literature is InScience Communications (844), which belongs to Springer Nature (Figure 13 in Appendix 1). Figure 4 shows that it published at least 119 papers in journals owned by its parent company. Collaborations between InScience Communications and Wolters Kluwer led to 29 additional papers published by Adis, Springer. In the same vein, Frontline Medical Communications published 23 publications that acknowledged Global Academy, with which it is currently affiliated. Some MECCs (or larger businesses providing MECC services) also own publishers. For example, Wolters Kluwer currently owns Lippincott Williams & Wilkins.Fig. 4Number of publications by publishers and by MECCs (n > 20, ggplot2 3.5.1) Number of publications by publishers and by MECCs (n > 20, ggplot2 3.5.1) Limitations Analyses involving MECCs only consider publications acknowledging MECCs (25 423), which constitutes a smaller dataset than the total number of unique publications (29 911). The frequent merging, rebranding, and selling of branches of pharmaceutical companies and MECCs limit the possibilities to identify them accurately over time. Even though they play a key role in industry-funded research, key opinion leaders are absent from this study because they are difficult to study across different topics since they specialize in one domain of research. Other limitations include the lack of transparency about medical writing in scientific articles, even though current transparency practices partly permitted this study to be conducted. Analysis and discussion This study provides new insights on the organization of industry-funded research and suggests strong relationships between major MECCs, research institutes, funders, and publishers. Transparency and research independence Collaborations among MECCs, sponsors, and well-known research institutes reconfigure authorship. While MECCs are increasingly acknowledged in publications, they typically do not appear in the bylines, even though previous research indicates that they contribute a substantial portion of the writing, if not all of it [5, 7]. On paper, byline authors are responsible for the content, not medical writers or pharmaceutical companies. The ICMJE’s definition of authorship aligns with industry’s definitions of medical writing, such as the Good Publication Practice (GPP) Guidelines, which state that a medical writer only “contributes” to a manuscript and should be acknowledged as such. This definition frames the problem of medical writing as a lack of transparency, and overshadows key issues related to publication planning such as corporate responsibility [23], conflicts of interest of KOLs, medical writers and company employees [4, 24], and marketing [7, 25, 26]. The overview of the most common affiliations suggests strong collaborations between pharmaceutical companies and research institutes, raising questions about research independence. Representing prominent research institutes, the authors participate in a network made of MECCs and sponsors, which configures the nature of their relationships, whether financial or of another nature. Recurring collaborations among researchers affiliated with research institutes, sponsors, and MECCs have been previously described in isolated cases [7, 8]. The cumulative evidence from these cases [5, 17, 25], combined with these findings, suggests that this is rather common practice. Within this organization, the independence of those researchers is in jeopardy. The organization of actors to publish scientific communication also questions the efficacy of conflicts of interest and transparency requirements, which mainly rely on self-regulation [2, 3]. They fail not only to disclose the extent of the conflicts of interest of those researchers, but also to describe the precise role of the sponsors and the MECCs. Financial relationships The medical writing market benefits pharmaceutical companies, as a few pharmaceutical companies subcontract a select number of MECCs to produce commissioned papers. The market seems to be highly competitive for MECCs. Findings also suggest that the main pharmaceutical companies resort more to MECCs for commissioned papers and might have better transparency practices than smaller companies. However, such conclusions cannot be confirmed unless those companies make their financial relationships and contractual obligations publicly accessible. Publication trackers and Fig. 3 suggest that the production of articles by batches, or publication planning, is common practice. The publication trackers found in the IDL are a result per se as they trace publications and indicate, in some cases, the product supported by several, if not many, publications (eg, Mallinckrodt’s publication tracker in Appendix 3). Healy & Cattell [27] used a similar document made available during a litigation and authored by a MECC that worked for Pfizer to promote its drug Zoloft (sertraline). They found that the commissioned literature had 50% more impact on the therapeutic domain than the non-commissioned literature on sertraline. The authors underlined that sponsors, or pharmaceutical companies, tend to avoid publishing negative results, and sponsorship has become “a demonstrable predictor of the findings of that study,” a clear example of the “funding effect.” This can also be a sign of disguising marketing as scientific communication. Litigation cases and previous studies found that many companies, including Purdue, [13, 16] Mallinckrodt [28], and Philip Morris [11, 12, 14, 15], have promoted misleading information in scientific publications about the health impacts of their products. Out of 229 of their publications (see Table 1), none were retracted nor received an expression of concern. Further inquiry is necessary to understand these low numbers and the potential impact of these publications on prescribers, researchers, and policymakers. Relations among sponsors are illustrated through thousands of co-funded publications in the database. The majority of these funders are pharmaceutical companies while government organizations are a minority. Financial involvement of several sponsors to produce research, such as the collaboration between Sanofi and Regeneron to develop a new drug, illustrates how increasingly common collaborations among sponsors have become. However, the difficulty in accessing information about these collaborations limits further analyses [29]. The medical writing market highly benefits the business model of for-profit publishers. Authors affiliated with a research institute and who benefit from the financial support of pharmaceutical companies have more chances to publish quickly in costly avenues. Commissioning a manuscript approximately cost $13 375 to $18 000 in the late 1990s [30]. A 2015 proposal submitted by Peloton Advantage to Teva Pharmaceuticals offered “Primary Manuscript development and submission” for $37 924 [31]. In twenty years, prices may have at least doubled. On top of medical writing, authors have to pay Articles Processing Charges (APCs). APCs of high-impact journals often start at $2 000 (median for gold open access in 2023) and can reach $12 290 (Nature portfolio of hybrid journals in 2024) [32]. Publishers are developing additional services, which make publishing even more costly and less accessible for independent researchers. For example, big publishers, such as Taylor & Francis (T&F), Future Medicine Ltd. (now T&F), and Elsevier, have developed paid fast-track or “accelerated” publishing services for a selection of journals, so that a publication can be reviewed faster than traditional peer review. In 2022, T&F offered this service for almost double the price of the standard cost of publication [33]. In an article published by the T&F Journal “Current Medical Research and Opinion” (see Figure 18 in Appendix 1), employees of Pfizer, AstraZeneca, Merck, and the MECC MedThink SciCom described how “fast-track review” could make the publication process more efficient [34]. These services and increasing costs of APCs for high impact journals may be aligning with the financial capacities of sponsored researchers, but not with those of independent researchers. Additional financial relationships between pharmaceutical companies and publishers include advertising, database subscriptions, and the sale of reprints. The medical writing market and the publishing market also overlap. Some companies own both MECCs and publishers, which are indicators of vertical integration and a shift towards an oligopoly. This configuration, as for example the ownership of both a MECC and a publisher, such as InScience Communications and Springer Nature, increases the risks of integrity breaches. Findings show that MECCs sometimes publish in journals owned by a publisher that is either a sister or a parent company. This configuration can jeopardize the independence of journals and editors, as demonstrated by the case involving Elsevier, its previously owned MECC Excerpta, and Merck [18]. Another example involves Informa, which owns Routledge and Taylor & Francis. In a 2015 email, the publisher offered discounts for pre-prints and perpetual open access to the opioid manufacturer Mallinckrodt in exchange for the publication of 15 manuscripts in five of its journals [35]. Among them were listed Postgraduate Medicine, Current Medical Research & Opinion, and the Journal of Medical Economics, which are also among the top publishers of the commissioned literature that is constitutive of the database (Figure 18 in Appendix 1). Conclusion The nature, often financial, of the relationships between publishers, pharmaceutical companies, research institutes, and MECCs alerts about the risks of integrity breaches that can stem from their collaborations. The findings in this article highlight these issues and offer new routes to explore the key roles of MECCs, even though such inquiries are inevitably limited due to the opacity surrounding their activities. They confirm that conflict of interest disclosure is not enough to address the systemic problems of corporate-sponsored research. This study also supports the claim that publication planning is common [5]. The relationships among actors hired by pharmaceutical companies to produce and circulate scientific information create a nexus of conflicts of interest conceptualized as “institutional corruption” [4] or the “ghost management” of medicine [5]. Ultimately, readers bear the burden of identifying potential conflicts of interest and biases in peer-reviewed articles. Descriptions of systemic problems also tend to disregard publishers. As discussed, the medical writing industry and the publishers have developed services partly or fully for pharmaceutical companies. Their business model thus depends or thrives on the needs of their clients. More stringent regulations and transparency measures about the relationships between those actors are crucial. The database made available through this study seeks to be part of this solution by increasing transparency on their relationships, and offering possibilities for further assessing the literature produced by these collaborations and their impacts on biomedical science and research integrity.