Profile Animal- Based Research Reported in the UK National Press
Total Page：16
File Type：pdf, Size：1020Kb
Open access Original research BMJ Open Science: first published as 10.1136/bmjos-2019-100039 on 20 October 2020. Downloaded from Clinical impact of high- profile animal- based research reported in the UK national press Jarrod Bailey ,1 Michael Balls2 This article has received OSF ABSTRACT Strengths and limitations of this study badges for Open data. Objectives We evaluated animal- based biomedical ‘breakthroughs’ reported in the UK national press in 1995 To cite: Bailey J, Balls M. This study investigates exaggeration in the media (25 years prior to the conclusion of this study). Based ► Clinical impact of high- profile of the significance and human relevance of animal on evidence of overspeculative reporting of biomedical animal- based research research. reported in the UK national research in other areas (eg, press releases and scientific ► The study focuses on articles in the UK national press. BMJ Open Science papers), we specifically examined animal research in press in a particular year, and specifically follows up 2020;4:e100039. doi:10.1136/ the media, asking, ‘In a given year, what proportion of the fate of forecasted ‘breakthroughs’, to see if they bmjos-2019-100039 animal research “breakthroughs”’ published in the UK had resulted in human benefit >20 years later. national press had translated, more than 20 years later, to ► Prepublication history for ► This study was comprehensive, objective and approved interventions?’ this paper is available online. To detailed. view these files, please visit the Methods We searched the Nexis media database ( ► Significant research was conducted for each media- journal online (http:// dx. doi. org/ LexisNexis. com) for animal- based biomedical reports in reported breakthrough, and all its findings have 10. 1136/ bmjos- 2019- 100039). the UK national press. The only restrictions were that the been made available in this report. intervention should be specific, such as a named drug, Received 24 July 2019 ► One limitation is that the focus was on one calendar gene, biomedical pathway, to facilitate follow- up, and that Revised 21 July 2020 year (1995). there should be claims of some clinical promise. copyright. Accepted 17 August 2020 ► However, there is no reason to believe that analyses Main outcome measures Were any interventions of other years would lead to significantly different approved for human use? If so, when and by which conclusions, with regard to the overspeculation and agency? If not, why, and how far did development overstatement of potential human benefits from proceed? Were any other, directly related interventions animal- based research in the media. approved? Did any of the reports overstate human relevance? Results Overspeculation and exaggeration of human relevance was evident in all the articles examined. Of 27 uncertain relevance to human health and http://openscience.bmj.com/ unique published ‘breakthroughs’, only one had clearly do not provide key facts or acknowledge resulted in human benefit. Twenty were classified as important limitations’. Of these, 90% lacked failures, three were inconclusive and three were partially caveats about extrapolating animal/labora- successful. tory studies to people, while explicitly making Conclusions The results of animal- based preclinical claims about relevance to human health, and research studies are commonly overstated in media 29% exaggerated the importance of the find- reports, to prematurely imply often-imminent ings they described. Notably, this was much ‘breakthroughs’ relevant to human medicine. more common for animal studies: 41% were exaggerated in this way, compared with 18% on September 25, 2021 by guest. Protected INTRODUCTION of human studies. Sumner et al7 examined 462 Animal experiments remain controver- press releases produced by the UK’s leading © Author(s) (or their sial, with issues including the welfare of 20 (Russell Group) universities, along with employer(s)) 2020. Re- use the animals involved, and the question- the associated scientific papers and print/ permitted under CC BY. able human relevance of animal data.1–4 online news stories, and concluded that 36% Published by BMJ. 1 Despite increasing evidence of the latter (see ‘contained exaggerated inference to humans Cruelty Free International, 5 London, UK Bailey for a review), overstatement of the from animal research’. 2University of Nottingham human benefits of animal research is wide- Exaggeration of animal-based findings has Faculty of Medicine and Health spread, and occurs throughout the whole also been noted in online and other media. Sciences, Nottingham, UK research process, from institutional press Haneef et al8 examined the health section of releases through to reports in the media. For Google News for ‘spin’, and concluded that Correspondence to 6 Dr Jarrod Bailey; example, Woloshin et al reported that press almost half (48%) of the reports they exam- jarrod. bailey@ releases from academic medical centres in ined that involved animal studies ‘implied contemporarysciences. org the USA ‘often promote research that has overgeneralization/misleading extrapolation Bailey J, Balls M. BMJ Open Science 2020;4:e100039. doi:10.1136/bmjos-2019-100039 1 Open access BMJ Open Science: first published as 10.1136/bmjos-2019-100039 on 20 October 2020. Downloaded from from animals to humans’. The UK’s ‘Leveson Inquiry into content, provided by the international company, LexisNexis the culture, practices and ethics of the press’ concluded ( lexisnexis. com). Media sources were selected to include that ‘overselling the results of non-human studies as a ‘UK national newspapers’, in the calendar year 1995. The promised cure potentially confuses readers and might search strategy involved selecting the ‘Medical research’ contribute to disillusionment with science’.9 The website index term, then adding the following animal terms to HealthNewsReview. org published an article in July 2018 identify news items based on animal research: ‘animal OR about the exaggeration of the applicability and rele- mouse OR mice OR rodent OR rat OR dog OR cat OR vance of animal data to humans, based on many of its monkey OR primate OR guinea pig OR rabbit’. Articles 6000 posts.10 ‘Vigilance’ was advised for both patients and that did not describe a clear, direct clinical promise, or that physicians when interpreting health claims that are often described a non- clinical application (eg, agricultural or exaggerated and/or unfounded, specifically for Parkin- veterinary), or that described only mechanisms of action, son’s disease and other movement disorders, and which pathophysiology or diagnosis, or in which the intervention included ‘unfulfilled promises of animal models’.11 was not of a specific named procedure or compound, or Exaggeration is also evident in scientific publica- was not speculated to be associated with a specific gene/ tions. Contopoulos- Ioannidis et al12 examined 101 scien- molecule/pathway, were excluded. For each report, the tific publications published in top scientific journals associated academic publication(s) were obtained, where (including, but not limited to animal research) and found available, and as much of the following data that were avail- that basic research rarely impacted clinical practice, even able were extracted: title, news media source, publishing when it was considered ‘highly promising’: 20 years later, journal, date of publication, author name(s), PubMed ID only five drugs were licensed for clinical use as a result, and links, animal species and numbers used, intervention, and only one was used extensively for the licensed indi- preventive/therapeutic in nature, expected clinical benefit cations. Lindl et al13 concluded that 17 animal research and years to expected benefit, relevant text and summary programmes licensed in Germany in the early 1990s, of findings, disease in question, institution, funding body, which promised new therapies, or at least direct clin- harms to animals, any salient quotes from authors, any ical impact, had resulted in ‘no clinical relevance’ 17 obvious related material, etc. years later.14 Hackam conducted a systematic review to To investigate whether clinical benefit transpired within see how often highly cited animal studies from the top 20- plus years, the following websites and sources were seven science journals translated to human success, and consulted: PubMed, the European Medicines Agency, the copyright. concluded that caution should be applied when extrapo- UK Medicines and Healthcare Regulatory Agency, the lating the findings of prominent animal research to the US Food and Drug Administration (FDA), clinicaltrials. care of human disease.15 Of 76 qualifying animal studies, gov, Medscape. com, the National Institute for Health and 28 had positive outcomes in human trials; but only 8 led Care Excellence, the British National Formulary, Centers to therapies approved for clinical use.15 for Disease Control and Prevention, the WHO and the US In summary, in the past approximately 15 years, various National Library of Medicine’s TOXNET. Data obtained efforts have been made to assess the outcomes and from thorough searches of these sources were collated, http://openscience.bmj.com/ human benefits of scientific breakthroughs, and how and used to determine the outcome of each ‘break- accurately and speculatively these were reported. Over- through’ with regard to any further studies that were statement, overspeculation and exaggeration were highly conducted; whether these were human, animal or both;