BMJ Open Is Committed to Open Peer Review. As Part of This Commitment We Make the Peer Review History of Every Article We Publish Publicly Available

BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from

BMJ Open is committed to open peer review. As part of this commitment we make the peer review history of every article we publish publicly available.

When an article is published we post the peer reviewers’ comments and the authors’ responses online. We also post the versions of the paper that were used during peer review. These are the versions that the peer review comments apply to.

The versions of the paper that follow are the versions that were submitted during the peer review process. They are not the versions of record or the final published versions. They should not be cited or distributed as the published version of this manuscript.

BMJ Open is an open access journal and the full, final, typeset and author-corrected version of record of the manuscript is available on our site with no access controls, subscription charges or pay-per-view fees (http://bmjopen.bmj.com).

If you have any questions on BMJ Open’s open peer review process please email [email protected] http://bmjopen.bmj.com/ on September 23, 2021 by guest. Protected copyright. BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from

Adverse events - Hazards of communicating medical science in the age of disinformation

Journal: BMJ Open ManuscriptFor ID peerbmjopen-2019-035626 review only Article Type: Original research

Date Submitted by the 08-Nov-2019 Author:

Complete List of Authors: Grimes, David; University of Oxford, & Queens University Belfast O'Connor, Robert; Irish Cancer Society

PUBLIC HEALTH, INFECTIOUS DISEASES, ONCOLOGY, MEDICAL Keywords: JOURNALISM, MEDICAL EDUCATION & TRAINING

http://bmjopen.bmj.com/ on September 23, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 21 BMJ Open

1 2 3

4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 8 9 I, the Submitting Author has the right to grant and does grant on behalf of all authors of the Work (as defined 10 in the below author licence), an exclusive licence and/or a non-exclusive licence for contributions from authors 11 who are: i) UK Crown employees; ii) where BMJ has agreed a CC-BY licence shall apply, and/or iii) in accordance 12 with the terms applicable for US Federal Government officers or employees acting as part of their official 13 duties; on a worldwide, perpetual, irrevocable, royalty-free basis to BMJ Publishing Group Ltd (“BMJ”) its 14 licensees and where the relevant Journal is co-owned by BMJ to the co-owners of the Journal, to publish the 15 Work in this journal and any other BMJ products and to exploit all rights, as set out in our licence. 16 17 The Submitting Author accepts and understands that any supply made under these terms is made by BMJ to 18 the Submitting Author Forunless you peer are acting as review an employee on behalf only of your employer or a postgraduate 19 student of an affiliated institution which is paying any applicable article publishing charge (“APC”) for Open 20 Access articles. Where the Submitting Author wishes to make the Work available on an Open Access basis (and 21 intends to pay the relevant APC), the terms of reuse of such Open Access shall be governed by a Creative 22 Commons licence – details of these licences and which Creative Commons licence will apply to this Work are set 23 out in our licence referred to above. 24 25 Other than as permitted in any relevant BMJ Author’s Self Archiving Policies, I confirm this Work has not been 26 accepted for publication elsewhere, is not being considered for publication elsewhere and does not duplicate 27 material already published. I confirm all authors consent to publication of this Work and authorise the granting 28 of this licence. 29 30 31 32 33 34 35

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 21 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 1 Adverse events - Hazards of communicating medical 2 3 science in the age of disinformation 4 1,2* 3 5 David Robert Grimes and Robert O’Connor 6

7 1. School of physical sciences, Dublin City University 8 2. Department of Oncology, University of Oxford 9 3. Irish Cancer Society, Dublin 10 11 Correspondence: [email protected] / [email protected] 12 13 Abstract 14 15 Objectives: Disinformation on medical matters has become an increasing public health concern. Public engagement by scientists, 16 clinicians, and patient advocatesFor can contribute peer towards review public understanding onlyof medicine. However, depth of feeling on many issues 17 18 (notably vaccination and cancer) can lead to adverse reactions for those communicating medical science, including vexatious 19 interactions and targeted campaigns. Our objective in this work is to establish a taxonomy of common negative experiences 20 encountered by those communicating medical science, and suggest guidelines so that they may be circumvented. 21 22 23 Design: We establish a taxonomy of the common negative experiences reported by those communicating medical science, informed 24 by surveying medical science communicators with public platforms. 25 26 27 Participants: 142 prominent medical science communicators (defined as having >1000 Twitter followers and experience 28 communicating medical science on social and traditional media platforms) were invited to take part in a survey, with 101 responses. 29 30 31 Results: 101 responses were analysed. Most participants experienced abusive behaviour (91.9%), including persistent harassment 32 http://bmjopen.bmj.com/ (69.3%) and physical violence and intimidation (5.9%). A substantial number (38.6%) received vexatious complaints to their 33 34 employers, professional bodies, or legal intimidation. The majority (62.4%) reported negative mental health sequelae due to public 35 outreach, including depression, anxiety, and stress. A significant minority (19.8%) were obligated to seek police advice or legal 36 counsel due to actions associated with their outreach work. While the majority targeted with vexatious complaints felt supported by 37 38 their employer / professional body, 32.4% reported neutral, poor or non-existent support. 39

40 on September 23, 2021 by guest. Protected copyright. Conclusions: Those engaging in public outreach of medical science are vulnerable to negative repercussions, and we suggest 41 42 guidelines for professional bodies and organisations to remedy some of these impacts on front-line members. 43 44 45 Strengths and limitations 46  Establishes a taxonomy of adverse reactions those communicating medical science can encounter. 47  Suggests guidelines for professional bodies, research institutes, and medical centres to circumvent the worst impacts on 48 49 members engaged in public outreach on medical science. 50  The taxonomy of adverse impacts was informed by surveying prominent communicators of medical science. 51 52  Survey results should only be taken as indicative of the scope of the issue at this juncture. 53  Much further research is needed to ascertain how the medical community can best act to counter the rise of medical 54 disinformation whilst protecting practitioners. 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 21 BMJ Open

Grimes and O’Connor Page 2 of 10 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 1 2 3 4 Introduction 5 6 Despite being fundamental to societal well-being, many aspects of medical science remain 7 poorly understood and frequently distrusted. Disinformation undermining health science and 8 evidence-based medicine has increased markedly in the era of social media, and dangerous 9 misconceptions abound, from perceived cancer risks and ostensible cures [1] to dangerous 10 falsehoods about vaccination [2]. Improving public awareness and understanding of science 11 12 and medicine is imperative if we are to maintain continued progress in research endeav-ours, 13 and scientists, physicians, and science communicators have a crucial role to play in shaping 14 public perceptions. Medical science is largely publicly funded, and direct communication of 15 research with the wider public can be extraordinarily beneficial on a societal level. Accordingly, 16 public engagementFor has peer become a prerequisite review for many funding only bodies. Informed engagement 17 by patient advocates and media gures too can have marked impact on public understanding of 18 19 medicine, empowering the public with facts with which to make important health decisions. 20 21 Improving public understanding of medical science is vital, as there many scenarios 22 where public perception (or a vocal subset of that) is starkly at odds with scientific 23 consensus. Frequently, medical science contradicts a narrative strongly held by particular 24 25 groups within the wider public. For our purposes, we define a ‘narrative’ as a world-view 26 or mindset shared by a particular subgroup, which unifies that grouping. Narratives are 27 often articles of faith, empowerment or comfort, frequently unsupported by available 28 evidence or at odds with scientific consensus. For clarity, we concentrate herein on 29 situations where there is no reputable evidence for a narrative, or where overwhelming 30 scientific consensus is firmly against that viewpoint. 31

32 http://bmjopen.bmj.com/ 33 Misguided narratives can be supremely damaging, and the anti-vaccine movement is 34 perhaps the most obvious example of this. Despite their life-saving efficacy of vaccination, 35 opposition has existed since the time of Jenner [3]. The rise of social media has seen 36 significant propagation of anti-vaccine narratives [4, 5, 6], driving uptake rates down and 37 causing serious harm worldwide [7, 8]. In 2018, Europe saw the highest number of cases 38 39 of measles cases in 20 years, numbering over 82,525 cases with at least 72 death - over

40 fifteen-fold the figures from 2016 [9]. Such is the extent of the problem that in 2019 the on September 23, 2021 by guest. Protected copyright. 41 World Health Organisation described vaccine hesitancy as a “Top ten threat to global 42 health" [10]. Exposure to anti-vaccine conspiracy to anti-vaccine conspiracy theory is a 43 leading factor in parental intention to vaccinate [8], and evidence to dates suggests that 44 the deluge of vaccine disinformation across social media is extremely damaging to public 45 46 understanding and health. 47 48 Other strongly-held narratives which clash with the weight of available scientific evidence 49 include the claims propounded by the anti-fluoride movement [11, 12], the beliefs of the 50 electromagnetic hypersensitivity movement [13], and the narratives of alternative medicine [14, 51 52 15]. Cancer patients are especially vulnerable to misinformation, and frequently targeted by 53 charlatans and the misguided [16]. Consequences of this can be severe, with patients 54 sometimes delaying or refusing conventional treatment. The net result of this is diminished 55 survival statistics for those who engage with cancer pseudoscience, due to delayed treatment 56 and sometimes rejection of conventional medicine. In some instances, subscribing to un- 57 proven or disproven modalities could approximately half survival time [17]. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 21

Grimes and O’Connor Page 3 of 10 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 1 2 3 4 5 While health falsehoods has always existed, the social media age has created new avenues 6 for misinformation (misinformed advice) and disinformation (deliberate falsehoods) to 7 propagate [16], rapidly bringing discredited ideas and dangerous pseudoscience to vast new 8 audiences. Scaremongering stories from dubious outlets propagate more readily than reliable 9 fact-based information from reputable sources [18, 19]. In 2016, over half of all cancer stories 10 shared on Facebook were medically unsound. Some have harnessed pseudoscience to sell 11 12 questionable diets, supplements, and books, to the detriment of public understanding. Internet 13 health gure Joseph Mercola, for example, made over $7 million in 2010 alone, proffering highly 14 dubious treatments and advice [20], including denigration of conventional therapies for cancer. 15 Mercola is far from unique, and such proclamations have huge potential for patient harm. 16 For peer review only 17 To counter this, public outreach by scientists, physicians, and evidence-based health advo- 18 19 cates must be a crucial element to counter damaging ctions, and strive for evidence-based 20 medicine. A physician’s recommendation, for example, is central to parental decisions to vac- 21 cinate [21]. Addressing patient concerns improves public health, and personal engagement by 22 researchers and physicians can have a positive impact on public perception. Patient 23 advocates and media gures have substantial ability to shift public perception; after Ireland saw 24 HPV vaccine uptake drop from 87% to 51%, an alliance of healthcare professionals, 25 26 researchers, and patient advocates were instrumental in countering the dominant falsehoods, 27 and Ireland has seen a dramatic recovery in vaccine uptake rates [22]. To make inroads 28 against the deluge of dubious health claims to which we’re subjected, it’s vital that scientists, 29 clinicians, and patient groups must be on the vanguard of efforts to counter-act misinformation. 30 31

32 Those engaging in public outreach, however, often encounter enmity for publicly http://bmjopen.bmj.com/ 33 advocating scientific evidence. Scientific consensus often runs contrary to deeply-held 34 beliefs, leading to certain groups attempting to undermine legitimate scientific 35 communication. Motivations for this are multi-faceted, often depending on very specific 36 circumstances. Conspiratorial thinking underpins many narratives, and those attempting 37 to communication science are often vilified as “shills", or agents of a nebulous “Big 38 39 Pharma". The phenomenon of identity protective cognition is also commonly encountered

40 [23] and narrative believers frequently attacking those who cast doubt on their beliefs. on September 23, 2021 by guest. Protected copyright. 41 Even when handled with sympathy and compassion, professional and patient advocates 42 who challenge misconceptions can become targets for certain individuals and groups. 43 44 45 These negative responses can range from verbal abuse to coordinated harassment 46 campaigns, and even violence. Aside from being deeply unsettling, such responses can cause 47 professional and personal problems for those targeted. With the increasing emphasis on public 48 engagement and ubiquity of the internet, this subject warrants urgent consideration, as there 49 are currently few clear guidelines for researchers, clinicians, or patient advocates engaging in 50 outreach work. Nor indeed is there a unified understanding of how adverse effects can 51 52 manifest, and institutions and professional bodies are typically ill-equipped or muted in their 53 support. This leaves those in the public eye or studying contentious topics vulnerable to 54 vexatious complaints and even physical harm. Without awareness of this reality, a less than 55 ideal situation where professional bodies can be weaponized against researchers can too 56 easily emerge. Accordingly, the aim of this work is to ascertain the typical experiences of those 57 communicating medical science and identify how these might be counteracted. 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 21 BMJ Open

Grimes and O’Connor Page 4 of 10 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 For peer review only 17 Figure 1 A non-exhaustive taxonomy of negative experiences encountered by individuals engaging in public 18 communication of health science. Subsections discussed in text 19 20 21 22 23 24 25 26 27 28 29 30 31 32 . http://bmjopen.bmj.com/ 33 34 35 36 Methods 37 Non-exhaustive taxonomy of adversarial tactics 38 An non-exhaustive taxonomy was constructed from interviews on the experiences of 39 those communicating medical science to the public. While non-exhaustive, it forms a 40 useful basis for more systematic investigation. Adversarial experiences in communicating on September 23, 2021 by guest. Protected copyright. 41 medicine were broadly strati ed into ve distinct classes, as illustrated graphically in figure 42 43 1. These sub-types are detailed in detail in table 1. 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 21

Grimes and O’Connor Page 5 of 10 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 1 2 3 4 Table 1 A non-exhaustive taxonomy of common adversarial tactics 5 Misrepresentation of scientific evidence / expertise 6 Misrepresenting scientific arguments to make them “Mercury is toxic, yet scientists put it in vaccines!” 7 Straw-manning amenable to ridicule or attack, and to deflect or - This statement belies importance of dose and 8 obscure evidence that undermines a particular ignores the fact there is no evidence for harm from 9 narrative. thimerosal in vaccines. 10 Selective, manipulative filtering of scientific “THC kills cancer, but doctors don’t want you to 11 Cherry-picking / evidence or expert statements, taken out of context know about cannabis cures." - THC can kill cells Quote-mining and employed to undermine the evidence-base or in a petri dish, but killing plated cells is entirely 12 promote a narrative. different from treating human cancer. 13 14 Shifting the burden of Insisting the onus is on scientists to ‘disprove’ “GMOs are toxic, and scientists should prove us 15 proof claims rather than offering any evidence or rationale wrong" - This assertion is untrue, and onus lies on for assertions made. those making the claim to proffer evidence for it. 16 For peer reviewDiscreditation attemptsonly 17 Casting doubt on one’s ability to question claims “This patient advocate isn’t an expert, so they can’t 18 Questioning at hand. Typically, this ostensible scepticism is not say this vaccine is safe! " - One does not need to qualification extended to purported experts sharing claims be an expert immunologist in this case to accurately 19 supportive of the narrative. reflect the medical consensus on vaccination. 20 Claims that the speaker is compromised due to “Who’s paying you to say this?" - Unsubstantiated 21 Alleging vested interests some apparent conflict of interest or that experts allegation to deflect from absence of evidence for a 22 are otherwise lacking impartiality. narrative or claim. 23 24 Allegations that the scientist, physician, or patient “She’s part of a pharma cover-up to suppress Asserting Conspiracy advocate is part of some conspiracy to suppress the natural cancer cures!" - Appeals to conspiracy 25 theory truth or spread false information, either as a pawn theory function to distract from lack of evidence. 26 or an active player. 27 Dubious Amplification of pseudo-scientific narratives 28 Targeting traditional or online media outlets to Pitching dubious health claims to journalists as amplify dubious narratives, typically by-passing human interest stories - This approach was 29 Media Targeting (by- traditional gatekeepers (science / health journalists successfully used by anti-vaccine activists to push 30 passing the gatekeepers etc) who would otherwise be more likely to spot the dis credited link between autism and the MMR pseudoscience. vaccine between 1998-2000. 31

32 Astroturfing / Use of fake social media accounts / fictitious pressure Example: Accounts which spring up once an initial http://bmjopen.bmj.com/ sock-puppeting groups to provide an illusion of grass-roots support for 33 anti-fact site is blocked but which include a particular narrative misinformation consistent with the originator’s initial 34 social media accounts. 35 36 Celebrities and influencers can have There are numerous examples of this, especially in Celebrity endorsement disproportionately large impact on the perception of relation to anti-vaccine activism, including actors 37 public even in areas where they have no relevant and models being cited for their purported 38 expertise or knowledge. knowledge of complex health issues. 39 Malicious complaints / abuse of regulatory frameworks

The spreading of malicious claims regarding an “I’ve heard that doctor abuses patients" - on September 23, 2021 by guest. Protected copyright. 40 Poisoning the well / individual’s professional or personal conduct to Inflammatory slurs such as these are designed to 41 Smear campaigns undermine them or discourage others from discredit, and are not in any way substantiated, but 42 engaging with them. calculated to invoke disgust or contempt. 43 Exaggerated / misleading accounts of interactions 44 Vexatious complaints to Making calculated complaints to one’s employer or with public advocates and demands to censure employers threatening to do so in order to intimidate them them, typically aimed at an individual’s university 45 into silence. or employer. 46 47 Abusing procedures of professional bodies to target Registering complaints with a medical regulatory Vexatious complaints a researcher / medic who presents a challenge to a body against a doctor for their advocacy of 48 to regulatory bodies narrative. These may also include unwarranted evidence-based positions. Physicians especially 49 freedom of information requests or vexatious vulnerable, as typically all complaints must be 50 parliamentary questions. investigated, regardless of merit. 51 Interpersonal harassment and intimidation Harassment can take many forms, and personal Abusive language made publicly or in direct 52 Harassment / abuse abuse is perhaps most common. Threats (implied messages, and posting of private contact details, 53 and direct) are often employed. phone numbers, addresses etc. 54 Legal threats Legal notices and mechanisms can also be utilised Threatening to bring an advocate to court for alleged 55 to silence researchers questioning a narrative, from defamation, also used judiciously to limit 56 cease and desist notices to defamation claims. independent investigation on pseudoscientific 57 narratives 58 Implicit or explicit threats of physical violence is an Stalking of private abodes, explicit threats, or 59 Unfortunately potent method of intimidating actions like spitting etc. There are instances of 60 PhysicalFor intimidation peer review onlyscientists - http://bmjopen.bmj.com/site/about/guidelines.xhtml into silence. This includes threats of security being required for scientific meetings on physical violence or rape (the latter usually directed publicly contentious subjects, due to implications of at female discussants) or threats of violence. Page 7 of 21 BMJ Open

Grimes and O’Connor Page 6 of 10 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 1 2 3 4 Table 2 Survey participants 5 Variable Participants 6 Invited to partake N=142 7 Total completed N=101 8 Gender Female (N=55) Male (N = 44) 9 Non-Binary / Undisclosed (N=2) 10 Affiliation University / Medical Centre (N = 52) 11 Una liated (N = 26) Media organisation (N = 20) 12 Charity (N = 11) 13 Political organisation (N=4) 14 Profession Medical professional (N = 23) Scientist / Researcher (N = 20) 15 Science communications (N = 18) 16 For peerPatient Advocate review(N = 16) only 17 Health Policy (N=5) 18 Years involved Average 10.7 Years (Range 2 - 30) 19 20 21 Survey of negative experiences 22 142 people were invited to undertake a brief survey aimed at capturing what if any nega- 23 tive experiences they might have suffered whilst advocating evidence-based medical 24 information in public forums, with 101 (71.1%) responding. Selection criteria was based 25 on having more than 1000 twitter followers, predominantly focused on English-speaking 26 27 public communica-tion of health science and with media experience in both traditional and 28 online channels. This survey was undertaken for indicative purposes, with participants 29 invited from a variety of backgrounds and subject areas. Data was collected from a non- 30 randomised group with no expectation of transferable findings, and was thus exempt from 31 requiring HRA / HRB approval. Participant details are given in table 2. Topics covered / 32 channels of engagement are depicted in figure 2(a). http://bmjopen.bmj.com/ 33 34 Patient and Public Involvement 35 As patient advocates play a substantial role to play in combatting misinformation on medical 36 issues, several who met the inclusion criteria were invited to take part, with 15.8% of 37 respondents being patient advocates. 38 39

40 on September 23, 2021 by guest. Protected copyright. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 21

Grimes and O’Connor Page 7 of 10 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 1 2 3 4 5 6 Results 7 8 9 Topics covered, channels of engagement, and mediums for negative encounters are 10 shown in figure 2. Twitter is disproportionately represented, as prominence on that 11 platform was part of the selection criteria. Other fora cited included books, 12 documentaries, newspapers, podcasts - and in one instance criticism under 13 parliamentary privilege. A majority (N = 70, 69.3%) had experienced targeted ire from 14 at least one grouping. The most common groupings for negative reactions were anti- 15 vaccine and alternative medicine groups respectively (N = 43 each, 42.6%) followed 16 by dietary Foradvocates peer (N = 26, 25.7%), review ’wellness’ groups only (N = 17, 16.8%), religious and 17 chronic illness groups (both N = 15, 14.9%), anti- fluoride and autism focused groups 18 (both N =12, 11.9%). Others cited by 3 or less respondents included 19 electromagnetic hypersensitivity factions, conspiracy theorists and anti-GMO 20 organisations. 21 22 Figure 3 depicts types of experiences reported by participants, ranging from the 23 relatively minor to the severe. Of participants surveyed, a majority (N = 63, 62.4%) 24 reported negative fallout from public engagement had caused them at least some 25 negative mental health sequelae, including depression, anxiety, and stress. Most of 26 this was reported as minor, but considerable or significant mental health ramifications 27 28 were reported by 15 respondents. Of the participants, 20 (19.8%) were obligated to 29 seek police advice / legal counsel as a direct result of targeted actions associated with 30 their outreach work. Of those receiving vexatious complaints (N = 39, 38.6% of all 31 respondents), most (67.6%) felt supported or well supported by their institution,

32 employer, or professional body, with whilst 16.2% deemed support to be neutral, with http://bmjopen.bmj.com/ 33 an equal number (16.2%) feeling poorly or entirely unsupported. Predictably perhaps, 34 gender specific abuse was far more likely to be directed at women (40% of female 35 respondents) than men (6.8% of male respondents), with this difference being highly 36 significant (p << 0.001, calculated by Welch’s t-test) 37 38 39

40 on September 23, 2021 by guest. Protected copyright. 41 42 Figure 2 (a) Topics covered by participants (b) channels of engagement for subjects surveyed (c) 43 Fora for negative interactions 44 45 46 47 48 49 50 Figure 3 Proportion of negative experiences recorded including (a) abuse experienced (b) violence and intimidation (c) vexatious complaints. 51 52 53 54 55 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 21 BMJ Open

Grimes and O’Connor Page 8 of 10 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 1 2 Discussion 3 4 In a globalised information age, medical science can appear disconnected and aloof from 5 those it serves to help. Educational and professional bodies (including universities and medical 6 centres) have a unique societal role to inform their peers and public on evidence-based 7 medicine, and a responsibility to adjust to modern communications realities. We can collec- 8 tively no longer remain on the fence in supporting health information advocacy. While being 9 10 mindful not to over-infer from our survey, we can use it as a basis to make some suggestions. 11 It is vital to have proactive strategies in place to support those engaging in medical outreach, 12 and to maintain a high calibre for public discussion. It is also crucial that those engaging in 13 outreach are cognizant of the potential pitfalls, and afforded ample support. 14 15 Suggested guidelines for professional bodies and employers 16 For peer review only 17 While by no means comprehensive, the following guidelines might be bene cial in 18 establish-ing policy for dealing with issues that can arise. 19 20 1 Educational / professional organisations must recognise a commitment to support 21 evidence-based actions by their members. This may require oversight of such activities 22 and investment in the governance / training resources to protect members willing 23 24 to act as advocates. 25 2 Institutions and professional bodies should have robust measures in place to oversee 26 communication activities associated with their members, and to make assessments as to 27 whether individuals are communicating established facts in good faith or are 28 contributing to undermining of fact with potential legal / reputational damage. 29 30 3 When institutions receive complaints regarding members, the subject must be a orded 31 presumption of innocence rather than being served with reactionary and inflexible

32 procedure, lest the institution might become an unwilling tool against science. http://bmjopen.bmj.com/ 33 4 In case of disputes, competent and impartial individuals should be engaged to 34 independently assess complaints, cognisant of vital background and context. 35 5 Coordination between press offices and those engaged in outreach would improve com- 36 37 munication, preemptively identifying those likely to be targets for malicious tactics. 38 6 Support for those identified as being on engagement front-lines should be maintained, 39 with clear legal advice / institutional support for targeted members. 40 7 Organisations must be vocal in supporting public-facing members, willing to issue on September 23, 2021 by guest. Protected copyright. 41 strong rebuttals of vexatious complaints against individuals. 42 43 8 Professional bodies and employers should promote strive to promote both scientific 44 freedom of speech, and strive to champion evidence-based advocacy. 45 9 When possible, those expected to engage in outreach should be trained in methods 46 that reduce opportunities for personal attacks. 47 48 Conclusions 49 50 The question of how we best communicate health science in the modern era is an area where 51 more research is urgently required, especially on the role of social media, and optimum ways 52 physicians, researchers, and other public-facing figures can promote good medical science 53 and mitigate falsehoods. The suggestions herein ought to be taken as a starting point, with 54 discussion evolving as improved evidence materialises. There are wider problems implicit in all 55 this those communicating science cannot tackle in isolation; social media regulation 56 57 58 59 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 21

Grimes and O’Connor Page 9 of 10 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 1 2 3 4 5 particularly is a serious issue, both in regards to the spreading of misinformation / dis- 6 information, and with respect to procedures preventing the weaponisation of social media 7 platforms. Social media platforms must ultimately be made answerable to regulatory over- 8 sight, just as every other important aspect of life is; claims of innocence are unconvincing 9 when their business model is so clearly dependent on advertising engagement at the cost 10 of lives. The problem of poor reporting and false balance [24] in conventional media 11 outlets also must be considered, and there is significant scope for scientists and doctors 12 13 to contribute to policy in these areas. There is ample evidence that physicians and 14 scientists have an important role to play in combatting health disinformation, as has 15 recently been argued by the authors in relation to vaccination for the British Medical 16 Journal [25].For But equally, peer it is crucial review that those engaging only in this vital work have the 17 requisite support from their institutions, so that deleterious consequences of laudable 18 outreach work might be circumvented. It is increasingly clear that disinformation about 19 medicine and illness has become ubiquitous, with severe consequence for both our 20 21 collective health and public understanding of medical science. Scientists and physicians 22 must be at the vanguard of the pushback against these dangerous falsehoods – our 23 societal well-being depends upon it. 24 25 Electronic Supplementary Material 26 A full version of the questionnaire and further analysis can be found in the electronic sup- 27 plementary material. 28 29 30 Acknowledgements 31 The authors would like to thank the scientists, physicians, and patient advocates who gave their time to discuss the issues they face in communicating science, and for sharing their insights. In particular, the authors would like to profoundly thank Dr. Laura J http://bmjopen.bmj.com/ 32 Brennan for her invaluable input; Laura was present and helped conceive the concept from which this paper sprung. She was a 33 passionate advocate evidence-based science communication and especially for the HPV vaccine, championing it despite 34 enduring a life-limiting and difficult cervical cancer diagnosis. Following a con dence crisis in the vaccine in Ireland driven by anti-vaccine activism, her tireless work was a substantial factor in reversing the damage wrought by disinformation. Laura 35 passed away in March 2019, but while we have lost a friend and colleague, hers is a legacy that will resonate for generations. 36 Laura embodied precisely how vital health communication is to promoting public health; Ireland, and the world, owe her a huge 37 debt, and we dedicate this work to her memory. 38 39 Contributor statement All authors contributed equally to this work. 40 on September 23, 2021 by guest. Protected copyright. 41 Competing interests The authors declare that they have no competing interests. 42 43 Funding statement 44 The authors declare that they have no specific funding for this work. 45 Data Sharing 46 Full survey forms available in electronic supplementary material. 47 References 48 1. Shahab, L., McGowan, J.A., Waller, J., Smith, S.G.: Prevalence of beliefs about actual and mythical causes of cancer and 49 their association with socio-demographic and health-related characteristics: Findings from a cross-sectional survey 50 in england. European Journal of Cancer 103, 308{316 (2018) 2. Kata, A.: A postmodern pandora’s box: anti-vaccination misinformation on the internet. Vaccine 28(7), 1709{1716 (2010) 51 3. Poland, G.A., Jacobson, R.M.: The age-old struggle against the antivaccinationists. New England Journal of Medicine 52 364(2), 97{99 (2011) 53 4. Zimmerman, R.K., Wolfe, R.M., Fox, D.E., Fox, J.R., Nowalk, M.P., Troy, J.A., Sharp, L.K.: Vaccine criticism on the world wide web. Journal of medical internet research 7(2), 17 (2005) 54 5. Kata, A.: Anti-vaccine activists, web 2.0, and the postmodern paradigm{an overview of tactics and tropes used online by 55 the anti-vaccination movement. Vaccine 30(25), 3778{3789 (2012) 56 6. Grimes, D.R.: On the viability of conspiratorial beliefs. PloS one 11(1), 0147905 (2016) 7. Goertzel, T.: Conspiracy theories in science. EMBO reports 11(7), 493{499 (2010) 57 8. Jolley, D., Douglas, K.M.: The e ects of anti-vaccine conspiracy theories on vaccination intentions. PloS one 9(2), 89177 58 (2014) 9. Thornton, J.: Measles cases in Europe tripled from 2017 to 2018. British Medical Journal Publishing Group (2019) 59 10. Organization, W.H., et al.: Ten threats to global health in 2019. 2019. Retrieved January (2019) 60 For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 21 BMJ Open

Grimes and O’Connor Page 10 of 10 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 1 2 11. Freeze, R.A., Lehr, J.H.: The Fluoride Wars: How a Modest Public Health Measure Became America’s Longest-Running 3 Political Melodrama. John Wiley & Sons, ??? (2009) 12. Grimes, D.R.: Commentary on “are fluoride levels in drinking water associated with hypothyroidism prevalence in england? 4 a large observational study of gp practice data and fluoride levels in drinking water". J Epidemiol Community Health 69(7), 5 616{616 (2015) 13. Grimes, D.R., Bishop, D.V.: Distinguishing polemic from commentary in science: Some guidelines illustrated with the case 6 of sage and burgio (2017). Child development 89(1), 141{147 (2018) 7 14. Singh, S., Ernst, E.: Trick or Treatment: The Undeniable Facts About Alternative Medicine. WW Norton & Company, 8 (2008) 15. Grimes, D.R.: Proposed mechanisms for homeopathy are physically impossible. Focus on Alternative and 9 Complementary Therapies 17(3), 149{155 (2012) 10 16. The, L.O.: Oncology," fake" news, and legal liability. The Lancet. Oncology 19(9), 1135 (2018) 11 17. Johnson, S.B., Park, H.S., Gross, C.P., James, B.Y.: Complementary medicine, refusal of conventional cancer therapy, and survival among patients with curable cancers. JAMA oncology 4(10), 1375{1381 (2018) 12 18. Vosoughi, S., Roy, D., Aral, S.: The spread of true and false news online. Science 359(6380), 1146{1151 (2018) 13 14 19. Brady, W.J., Wills, J.A., Jost, J.T., Tucker, J.A., Van Bavel, J.J.: Emotion shapes the diffusion of moralized content in 15 social networks. Proceedings of the National Academy of Sciences 114(28), 7313{7318 (2017) 16 20. Smith, B.: Dr.For mercola: Visionary peer or quack? Chicago review (2012) only 17 21. Rosenthal, S., Weiss, T.W., Zimet, G.D., Ma, L., Good, M., Vichnin, M.: Predictors of hpv vaccine uptake among women 18 aged 19{26: importance of a physician’s recommendation. Vaccine 29(5), 890{895 (2011) 22. Corcoran, B., Clarke, A., Barrett, T.: Rapid response to hpv vaccination crisis in ireland. The Lancet 391(10135), 2103 19 (2018) 20 23. Kahan, D.M.: Misconceptions, misinformation, and the logic of identity-protective cognition (2017) 21 24. Grimes, D.R.: A dangerous balancing act. EMBO reports (2019) 22 25. Grimes, D.R.: “David Robert Grimes: Vaccines-How Can We Counter Misinformation Online?” The BMJ, 7 Nov. 2019, 23 blogs.bmj.com/bmj/2019/11/06/david-robert-grimes-vaccines-how-can-we-counter-misinformation-online/. 24 25 26 27 28 29 30 31

32 http://bmjopen.bmj.com/ 33 34 35 36 37 38 39

1 2 3 4 5 6 7 8 9 10 11 12 Questioning 13 14 qualifications Questioning 15 16 motivations 17 Shifting burden 18 19 of Proof 20 21 22 Alleging 23 24 25 Conspiracy 26 27 28 29 Cherrypicking / 30 31 Quote-mining 32 33 34 35 36 37 Discreditation Bypassing 38 39 Gatekeepers 40 41 42 43 For peer review only 44 Mispresentation 45 Straw- 46 47 manning 48 49 50 51 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 52 53 54 Dubious Sockpuppets / 55 56 Amplification Astroturfing 57 58 59 60

Violence / Intimidation Interpersonal Harassment

Malicious Celebrity Complaints Endorsement

Legal threats http://bmjopen.bmj.com/

Poisoning the well

Abuse / on September 23, 2021 by guest. Protected copyright. Threats Vexatious complaints Vexatious complaints to employers to professional bodies

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 21 BMJ Open

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 For peer review only 45 46 47 48 49 50 51 52 53

54 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 55 56 57 58 59 60 http://bmjopen.bmj.com/ on September 23, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 For peer review only 39 40 41 42 43 44 45

46 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 47 48 49 50 51 52 53 54 55 56 57 58 59 60 http://bmjopen.bmj.com/ on September 23, 2021 by guest. Protected copyright. Required investigation Dismissed without investigation

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 21 BMJ Open

* Required Information 1 Experiences in public communication of medical science 2 & health 3

4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 This survey is part of a research project that aims to better understand the experiences of medical 7 professionals, scientific researchers, patients advocates, health journalists, & policy makers in 8 communicating health issues to the public. It consists of 20 short questions, & should take 5-10 9 minutes to complete. All responses are anonymous, and their content will be used for research purposes only. If you might be willing to contribute more about your experiences we invite you to 10 inform us of this on the final page of the survey. For any questions, please e-mail 11 [email protected] 12 13

14 15 * 1. Please select that which best describes your primary role (Select one option) 16 17 Medical Professional Scientific Communications 18 Scientist / Researcher ForHealth Journalism peer review only 19 Patient Advocate Health Policy 20 21 Other (Please specify) ______22 23 24 * 2. Please specify your gender (Select one option) 25 Female Non-Binary 26 27 Male Prefer not to say 28

29 30 * 3. Please indicate which of the following organsiations you are professionally affiliated with that are directly relevant to 31 your outreach efforts.

32 University / research institute Media outlet 33 34 Medical institute Political grouping 35 NGO / Charity Non-affiliated (Patient Advocate / independent)

36 http://bmjopen.bmj.com/ Other (Please specify) ______37 38 39 * 4. Please select the topic or topics you frequently communicate about with the public. Please note that this includes topics 40 you might debunk or criticize. Select all that apply. 41 42 Vaccination Alternative Medicine Medicine in media 43 Dietary Science Pseudoscience / scams Medical science

44 on September 23, 2021 by guest. Protected copyright. 45 Cancer Chronic illness Bioethics 46 Health Policy Infectious diseases Interpretation of research

47 Reproductive / Sexual Health Pharmacology / drugs Gender issues 48 49 Patient Advocacy Psychological issues Genetics 50 Other (Please specify) ______51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 16 of 21

1 * 5. What media do you use for public engagement? Select all that apply. 2 3 Television (appearances) Newspapers / Magazines (advising / quoted) Facebook

4 Television (advising / quoted) Books (Author) Twitter BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 Radio (appearances) Books (Advising / quoted) Instagram 7 Radio (advising / quoted) Public talks Youtube 8 Newspapers / Magazines (Author) Podcasts Blogging 9 10 Other (Please specify) ______11

12 13 * 6. Approximately how long have you been communicating medical issues to the public? Please round to the nearest whole year 14 (Enter a value between 0 and 120) 15 16 17 18 For peer review only 19 * 7. Please select what best describes the extent of your public engagement (Select one option) 20 21 I am heavily involved in public outreach I am lightly involved in public outreach 22 I am considerably involved in public outreach My vocation is media or outreach centred 23 I am involved in public outreach 24 25 Other (Please specify) ______26 27 28 29 30 31 32 33 34 35

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 1 of 4 For any queries, please contact Dr. David Robert Grimes at [email protected] or [email protected] For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 17 of 21 BMJ Open

1 Experiences in public communication of medical science 2 & health 3

4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 Part II: Engagement experiences 7 8 9 10 8. In your opinion, how frequently do the following sentiments apply to your experiences of public engagement? 11 Always Frequently Sometimes Infrequently Never Unsure 12 *(a) I feel my efforts contribute to public 13 understanding 14 15 *(b) I find engagement prompts constructive and 16 respectful discussion 17 *(c) Engagement is mutually informative 18 For peer review only 19 *(d) My contributions are welcomed & appreciated 20 21 *(e) Engagement increases my personal 22 understanding

23 *(f) I feel I correct some misconceptions 24 25 *(g) I find my engagement changes minds & informs 26 27 *(h) I consider outreach personally rewarding 28 29 *(i) My efforts feel futile 30 *(j) Engagement takes a toll on my mental health & 31 well-being 32 33 *(k) I find my efforts are taken in good faith by 34 others 35

36 http://bmjopen.bmj.com/ * 9. Please indicate whether you have experienced any of the following negative experiences during your engagement 37 efforts. Select all that apply. Note that for the purposes of this survey, vexatious complaints are defined as those raised 38 chiefly to harass or intimidate.

39 Personal abuse Sexist or gender-specific abuse 40 41 Personal smears Racist of ethnic-specific abuse 42 Professional smears Intimidation attempts 43 Physical threats Vexatious legal threats

44 on September 23, 2021 by guest. Protected copyright. 45 Threats of sexual violence Vexatious complaints to employers 46 Harassment from individuals Vexatious complaints to professional bodies 47 Coordinated harassment from specific groups Persistent unwanted communication 48 49 Questioning of one's motivations No negative experiences 50 Other (Please specify) ______51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 18 of 21

1 * 10. How would you best describe your feelings about your outreach experiences to date? (Select one option) 2 3 Largely rewarding

4 Rewarding BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 Mixed 6 7 Not very rewarding 8 Not at all rewarding 9 Other (Please specify) ______10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 4 For any queries, please contact Dr. David Robert Grimes at [email protected] or [email protected]

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 21 BMJ Open

1 Experiences in public communication of medical science 2 & health 3

4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 * 11. If you have been the recipient of abuse or personal / professional smears stemming from your engagement, please 6 indicate which of the following you've experienced (check all that apply)

7 Assertions that you are dishonest or Aggressive or intimidating comments 8 deliberately misleading 9 Insults about your professional competence or Insults based on appearance, race, or gender 10 intelligence Malicious comments about one's motivations / allegations of 11 I have not experienced any of the above 12 'corruption' or 'shilling' 13 Malacious comments about one's personal or sexual conduct 14 Other (Please specify) ______15 16 17 * 12. Please indicate whether you have been subjected to any of the following (check all that apply) 18 For peer review only 19 Threats or implications of physical violence Persistent trolling / harassment by an individual 20 21 Physical violence or intimidation Persistent trolling / harassment by a group 22 Spreading of malicious rumours None of the above 23 Repeated unwanted communications 24 25 Other (Please specify) ______26

27 28 * 13. If you have been the victim of targeted abuse for your outreach work by a group or community, what best describes 29 that grouping? Choose all that apply if relevant 30 Anti-vaccine groups Religious groupings 31 32 Dietary advocates Alternative medicine advocates 33 Chronic illness groups "Wellness" groups

34 Anti-fluoride groups Autism-focused groups 35 Electromagnetic hypersensitivity groups I have not encountered this 36 http://bmjopen.bmj.com/ 37 Other (Please elaborate as appropriate) ______38 39 40 * 14. Please indicate whether you have experienced any of the following. Note that for the purposes of this survey, vexatious 41 complaints are defined as those raised chiefly to harass or intimidate. 42 Vexatious complaint to employer or institution - complaint dismissed without investigation 43

44 Vexatious complaint to employer or institution - required investigation on September 23, 2021 by guest. Protected copyright. 45 Vexatious complaint to a professional body - complaint dismissed without investigation 46 47 Vexatious complaint to a professional body - required investigation 48 Vexatious legal complaint - complaint dismissed without investigation

49 Vexatious legal complaint - complaint required investigation 50 51 None of the above 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 20 of 21

1 * 15. If you have experienced a vexatious complaint, did you feel supported by your institution or professional body? (Select 2 one option) 3 Highly supported Not well supported

4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 Supported Entirely unsupported 6 Neutral Not applicable 7 8 9 * 16. If you have experienced abusive or diminishing comments, in which fora are these typically delivered or shared? 10 (check all that apply) 11 E-mail Youtube 12 13 Post Other social media 14 Phone calls Blogs 15 Texts Websites 16 17 Twitter Public television 18 Facebook Public radioFor peer review only 19 Instagram Not applicable 20 21 Other (Please specify) ______22 23 24 * 17. Have you ever had to take legal action or consult law-enforcement officials regarding malicious communications, 25 threats, or claims? (Select one option)

26 Yes Not applicable 27 28 No Prefer not to say 29 30 31 * 18. Have negative reactions to public engagement ever caused you mental health problems (depression, anxiety, stress, etc) or otherwise impeded your functioning? (Select one option) 32 33 Yes - severely Unsure

34 Yes - considerably No 35 Yes - to a minor degree Prefer not to say

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 21 BMJ Open

1 19. Please expand on noteworthy negative situations you have encountered while engaging in health outreach. This is 2 entirely optional, but helps us identify common experiences and pitfalls. 3

4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 8 9 10 11 12 13 14 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 * 20. Finally, which of the following changes (if any) do you feel would benefit those communicating health the most? (check 33 all that apply) 34 35 Increased support from professional bodies Practical legal advice

36 Increased support from academic institutions More proactive stances from institutions and bodies http://bmjopen.bmj.com/ 37 Increased support from advocacy / healthcare bodies Dedicated funding for outreach 38 Clearer guidelines and better training on public Better protective mechanisms from internet / social media 39 engagement companies 40 41 Improved emotional support services None of the above 42 Other (Please specify) ______43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 3 of 4 For any queries, please contact Dr. David Robert Grimes at [email protected] or [email protected]

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 21

1 Experiences in public communication of medical science 2 & health 3

4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 Thank you for taking part in this survey. Results will be treated anonymously, but if you are happy for us to contact you 7 and perhaps quote from your experiences if appropriate, please indicate this below. 8 9 21. If you would like to be kept up to date on results and follow-up, please enter your e-mail address below. 10 11 12

13 22. We'd be grateful if you'd also supply your twitter handle. This will not be shared without your consent, and is entirely 14 optional, but may prove useful in future work estimating network reach of health information. 15 16 17 18 For peer review only 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 4 For any queries, please contact Dr. David Robert Grimes at [email protected] or [email protected] For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from

Establishing a taxonomy of potential hazards associated with communicating medical science in the age of disinformation ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2019-035626.R1

Article Type: Original research

Date Submitted by the 06-Feb-2020 Author:

Complete List of Authors: Grimes, David; University of Oxford, & Queens University Belfast Brennan, Laura ; Not applicable , Not applicable (patient advocate, deceased) O'Connor, Robert; Irish Cancer Society

Primary Subject Public health Heading:

Secondary Subject Heading: Medical education and training

PUBLIC HEALTH, INFECTIOUS DISEASES, ONCOLOGY, MEDICAL Keywords:

JOURNALISM, MEDICAL EDUCATION & TRAINING http://bmjopen.bmj.com/

on September 23, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 24 BMJ Open

1 2 3

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 24

1 2 3

4 Establishing a taxonomy of potential hazards associated with BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 communicating medical science in the age of disinformation 7 David Robert Grimes1,2* Laura J Brennan†, and Robert O’Connor3 8 1. School of physical sciences, Dublin City University, Glasnevin, Dublin 9, Ireland 9 2. Department of Oncology, University of Oxford, Off Roosevelt Rd, Oxford, UK 10 3. Irish Cancer Society, Dublin 4, Dublin, Ireland 11 Correspondence: [email protected] / [email protected] 12 13 14 Abstract 15 Objectives: Disinformation on medical matters has become an increasing public health concern. Public 16 engagement by scientists, clinicians, and patient advocates can contribute towards public 17 18 understanding of medicine.For However,peer depth review of feeling on manyonly issues (notably vaccination and 19 cancer) can lead to adverse reactions for those communicating medical science, including vexatious 20 interactions and targeted campaigns. Our objective in this work is to establish a taxonomy of common 21 negative experiences encountered by those communicating medical science, and suggest guidelines 22 so that they may be circumvented. 23 24 Design: We establish a taxonomy of the common negative experiences reported by those 25 communicating medical science, informed by surveying medical science communicators with public 26 27 platforms. 28 29 Participants: 142 prominent medical science communicators (defined as having >1000 Twitter 30 followers and experience communicating medical science on social and traditional media platforms) 31 were invited to take part in a survey, with 101 responses. 32 33 Results: 101 responses were analysed. Most participants experienced abusive behaviour (91.9%), 34 including persistent harassment (69.3%) and physical violence and intimidation (5.9%). A substantial 35 number (38.6%) received vexatious complaints to their employers, professional bodies, or legal 36 http://bmjopen.bmj.com/ 37 intimidation. The majority (62.4%) reported negative mental health sequelae due to public outreach, 38 including depression, anxiety, and stress. A significant minority (19.8%) were obligated to seek police 39 advice or legal counsel due to actions associated with their outreach work. While the majority targeted 40 with vexatious complaints felt supported by their employer / professional body, 32.4% reported 41 neutral, poor or non-existent support. 42 43 Conclusions: Those engaging in public outreach of medical science are vulnerable to negative

44 on September 23, 2021 by guest. Protected copyright. repercussions, and we suggest guidelines for professional bodies and organisations to remedy some 45 46 of these impacts on front-line members. 47 48 Strengths and limitations 49  Individuals involved in the communication of medical science were surveyed to ascertain their 50 experiences in public outreach. 51  Results of this survey established a taxonomy of adverse reactions those communicating medical 52 science can encounter. 53  Based upon survey feedback, suggested guidelines for professional bodies, research institutes, and 54 medical centres to circumvent the worst impacts on members engaged in public outreach on medical 55 science are preferred. 56  Survey results should only be taken as indicative of the scope of the issue at this juncture. 57  Much further research is needed to ascertain how the medical community can best act to counter the 58 rise of medical disinformation whilst protecting practitioners 59 60 † Deceased 20th March 2019

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 24 BMJ Open

1 2 3

4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 Introduction 8 9 10 Despite being fundamental to societal well-being, many aspects of medical science remain poorly 11 understood and frequently distrusted. Disinformation undermining health science and evidence- 12 based medicine has increased markedly in the era of social media, and dangerous misconceptions 13 abound, from perceived cancer risks and ostensible cures [1] to dangerous falsehoods about 14 vaccination [2]. Improving public awareness and understanding of science and medicine is imperative 15 if we are to maintain continued progress in research endeavours, and scientists, physicians, and 16 science communicators have a crucial role to play in shaping public perceptions. Medical science is 17 largely publicly funded, and direct communication of research with the wider public can be 18 For peer review only 19 extraordinarily beneficial on a societal level. Accordingly, public engagement has become a 20 prerequisite for many funding bodies. Informed engagement by patient advocates and media figures 21 too can have marked impact on public understanding of medicine, empowering the public with facts 22 with which to make important health decisions. 23 24 Improving public understanding of medical science is vital, as there many scenarios where public 25 perception (or a vocal subset of that) is starkly at odds with scientific consensus. Frequently, medical 26 science contradicts a narrative strongly held by particular groups within the wider public. For our 27 28 purposes, we define a ‘narrative’ as a world-view or mindset shared by a given subgroup, which unifies 29 that grouping. Narratives are often articles of faith, empowerment or comfort, frequently 30 unsupported by available evidence or at odds with scientific consensus. For clarity, we concentrate 31 herein on situations where there is no reputable evidence for a narrative, or where overwhelming 32 scientific consensus is firmly against that viewpoint. 33 34 Misguided narratives can be supremely damaging, and the anti-vaccine movement is perhaps the 35 most obvious example of this. Despite their life-saving efficacy of vaccination, opposition has existed 36 http://bmjopen.bmj.com/ since the time of Jenner [3]. The rise of social media has seen significant propagation of anti-vaccine 37 38 narratives [4, 5, 6], driving uptake rates down and causing serious harm worldwide [7, 8]. In 2018, 39 Europe saw the highest number of cases of measles cases in 20 years, numbering over 82,525 cases 40 with at least 72 death - over fifteen-fold the figures from 2016 [9]. Such is the extent of the problem 41 that in 2019 the World Health Organisation described vaccine hesitancy as a “Top ten threat to global 42 health" [10]. Exposure to anti-vaccine conspiracy theory is a leading factor in parental intention to 43 vaccinate [8], and evidence to date suggests that the deluge of vaccine disinformation across social 44 media is extremely damaging to public understanding and health. on September 23, 2021 by guest. Protected copyright. 45 46 47 Other strongly-held narratives which clash with the weight of available scientific evidence include the 48 claims propounded by the anti-fluoride movement [11, 12], the beliefs of the electromagnetic 49 hypersensitivity movement [13], and the narratives of alternative medicine [14, 15]. Cancer patients 50 are especially vulnerable to misinformation, and frequently targeted by charlatans and the misguided 51 [16]. Consequences of this can be severe, with patients sometimes delaying or refusing conventional 52 treatment. The net result of this is diminished survival statistics for those who engage with cancer 53 pseudoscience, due to delayed treatment and sometimes rejection of conventional medicine. In some 54 instances, subscribing to un-proven or disproven modalities could approximately half survival time 55 56 [17]. 57 58 While health falsehoods have always existed, the social media age has created new avenues for 59 misinformation (misinformed advice) and disinformation (deliberate falsehoods) to propagate [16], 60 rapidly bringing discredited ideas and dangerous pseudoscience to vast new audiences.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 24

1 2 3 Scaremongering stories from dubious outlets propagate more readily than reliable fact-based 4 information from reputable sources [18, 19]. In 2016, over half of all cancer stories shared on BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 Facebook were medically unsound. Some have harnessed pseudoscience to sell questionable diets, 7 supplements, and books, to the detriment of public understanding. Internet health guru Joseph 8 Mercola, for example, made over $7 million in 2010 alone, proffering dubious treatments and advice 9 [20], including denigration of conventional therapies for cancer. Mercola is far from unique, and such 10 proclamations have huge potential for patient harm. 11 12 To overcome this challenge, public outreach by scientists, physicians, and evidence-based health 13 advocates must be a crucial element to counter damaging fictions and empower our community with 14 evidence-based information. A physician’s recommendation, for example, is central to parental 15 16 decisions to vaccinate [21]. Addressing patient concerns improves public health, and personal 17 engagement by researchers and physicians can have a positive impact on public perception. Patient 18 advocates and mediaFor figures have peer substantial review ability to shift public only perception; after Ireland saw HPV 19 vaccine uptake drop from 87% to 51%, an alliance of healthcare professionals, researchers, and 20 patient advocates were instrumental in countering the dominant falsehoods, and Ireland has seen a 21 dramatic recovery in vaccine uptake rates [22]. To make inroads against the deluge of dubious health 22 claims to which we’re subjected, it’s vital that scientists, clinicians, and patient groups must be on the 23 vanguard of efforts to counter-act misinformation. 24 25 26 Those engaging in public outreach, however, often encounter enmity for publicly advocating scientific 27 evidence. Scientific consensus often runs contrary to deeply-held beliefs, leading to certain groups 28 attempting to undermine legitimate public scientific discourse. Motivations for this are multi-faceted, 29 often depending on very specific circumstances. Conspiratorial thinking underpins many narratives, 30 and those attempting to communication science are often vilified as “shills", or agents of a nebulous 31 “Big Pharma". The phenomenon of identity protective cognition is also commonly encountered [23] 32 and narrative believers frequently attacking those who cast doubt on their beliefs. Even when handled 33 with sympathy and compassion, professional and patient advocates who challenge misconceptions 34 35 can become targets for certain individuals and groups. These negative responses can range from

36 verbal abuse to coordinated harassment campaigns, and even violence. http://bmjopen.bmj.com/ 37 38 Aside from being deeply unsettling, such responses can cause professional and personal problems for 39 those targeted. With the increasing emphasis on public engagement and ubiquity of the internet, this 40 subject warrant urgent consideration, as there are currently few clear guidelines for researchers, 41 clinicians, or patient advocates engaging in outreach work. Nor indeed is there a unified understanding 42 of how adverse effects can manifest, and institutions and professional bodies are typically ill-equipped 43 or muted in their support. This leaves those in the public eye or studying contentious topics vulnerable 44 on September 23, 2021 by guest. Protected copyright. 45 to vexatious complaints and even physical harm. Without awareness of this reality, a less than ideal 46 situation where professional bodies can potentially be weaponised against researchers can too easily 47 emerge. Accordingly, the aim of this work is to ascertain the typical experiences of those 48 communicating medical science and identify how these might be counteracted. 49 50 Figure 1 A non-exhaustive taxonomy of negative experiences encountered by individuals 51 52 engaging in public communication of health science. Subsections discussed in text. 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 24 BMJ Open

1 2 3

4 Methods BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 Sample recruitment and selection criteria 8 The main recruitment fora for this study was online discussion groups for physicians, scientists, and 9 10 patient advocates communicating aspects of medical science to the general public across social and 11 traditional media. At the height of the Irish HPV vaccine confidence crisis [22], several physicians and 12 scientists (based in Ireland and the UK) in these groups sought advice for negative experiences, 13 including threats and malicious complaints to their employers and professional bodies, all of which 14 were eventually dismissed. Group members across Europe echoed similar experiences in 15 communicating vaccination science and in other health issues, and almost identical adverse reports 16 came from colleagues across the America and Africa. Informal interviews were conducted on foot of 17 this in these fora to identify common issues, as to the authors’ knowledge there is no existent 18 For peer review only literature on the topic. 19 20 21 On basis of these interviews and related fora discussions, a survey was created, including free-form 22 sections where subjects were free to expand on their own experiences. The wording of this survey is 23 included in the supplementary material. The selection criteria was specifically for those 24 communicating medical science both on social media (defined as having over 1000 followers on 25 twitter) and in mainstream channels (defined as invited appearances on public television, radio, and / 26 or in the form of newspaper articles & invited comment). From this, 142 individuals world-wide (from 27 across Europe, America, Africa, and Asia) working predominantly in the English language were 28 29 identified as fitting the criteria and invited to partake, of whom 101 (71.1%) responded. This survey 30 was undertaken for indicative purposes and was collected from a non-randomised group with no 31 expectation of transferable findings. Accordingly, it was thus exempt from requiring HRA / HRB 32 approval, though informed consent was sought and obtained prior to subjects partaking, with all data 33 appropriately anonymised. Subject details are given in table 1. 34 35 36 Patient and Public Involvement http://bmjopen.bmj.com/ 37 38 39 As patient advocates play a substantial role to play in combatting misinformation on medical issues, 40 several who met the inclusion criteria were invited to take part, with 15.8% of respondents being 41 42 patient advocates. One co-author of this work (LBJ) was a prominent patient advocate. 43

44 Table 1: Participant details on September 23, 2021 by guest. Protected copyright. 45 Inclusion and completion Invited to take part (N = 142) Total Completed (N = 101) 46 Gender Female (N=55) 47 Male (N=44) 48 Non-Binary / Undisclosed (N=2) 49 Affiliation University / Medical Centre (N=52) 50 Unaffiliated (N=26) Media organisation (N=20) 51 Charity (N=11) 52 Political organisation (N=4) 53 Profession Medical Professional (N=23) 54 Scientist / Researcher (N=20) 55 Science communications (N=16) Patient Advocate (N=16) 56 Health Policy (N=5) 57 Years active Average: 10.7 years (Range 2-30 years) 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 24

1 2 3

4 Table 2 A non-exhaustive taxonomy of common adversarial tactics BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 Misrepresentation of scientific evidence / expertise 6 Straw- Misrepresenting scientific arguments to make them amenable to ridicule or “Mercury is toxic, yet scientists put it in vaccines!” - This 7 manning attack, and to deflect or obscure evidence that undermines a particular narrative. statement belies importance of dose and ignores the fact there is 8 no evidence for harm from thimerosal in vaccines. 9 Cherry- Selective, manipulative filtering of scientific evidence or expert statements, taken “THC kills cancer, but doctors don’t want you to know about 10 picking / out of context to undermine evidence-base or promote a narrative. cannabis cures." - THC can kill cells in a petri dish, but killing 11 Quote-mining plated cells is entirely different from treating human cancer. 12 13 Shifting the Insisting the onus is on scientists to ‘disprove’ claims rather than offering any “GMOs are toxic, and scientists should prove us wrong" - This burden of evidence or rationale for assertions made. assertion is untrue, and onus lies on those making the claim to 14 proof proffer evidence for it. 15 16 Discreditation attempts 17 Questioning Casting doubt on one’s ability to question claims at hand. Typically, ostensible “This patient advocate isn’t an expert, so they can’t say this 18 qualification scepticism is not extendedFor to claims supportivepeer of the narrative. review onlyvaccine is safe!" - One does not need to be an expert immunologist in this case to accurately reflect medical consensus. 19 20 Alleging Claims that the speaker is compromised due to some apparent conflict of interest “Who’s paying you to say this?" - Unsubstantiated allegation to 21 vested or that experts are otherwise lacking impartiality. deflect from absence of evidence for a narrative or claim. 22 interests 23 Asserting Allegations that the scientist, physician, or patient advocate is part of some “She’s part of a pharma cover-up to suppress natural cancer 24 Conspiracy conspiracy to suppress the truth or spread false information, either as a pawn or cures!" - Appeals to conspiracy theory function to distract from 25 theory an active player. lack of evidence. 26 27 28 Dubious amplification of pseudoscientific narratives 29 Media Targeting traditional or online media outlets to amplify dubious narratives, Pitching dubious health claims to journalists as human interest Targeting typically by-passing gatekeepers (science / health journalists etc) who would stories - This approach was successfully used by anti-vaccine 30 otherwise be more likely to spot pseudoscience. activists to push the dis credited link between autism and the 31 MMR vaccine between 1998-2000. 32 Astroturfing 33 / sock- Use of fake social media accounts / fictitious pressure groups to provide an Example: Accounts which spring up once an initial anti-fact site is puppeting illusion of a wider grass-roots support for a particular narrative blocked but which include misinformation consistent with the 34 originator’s initial social media accounts. 35

36 Celebrity Celebrities and influencers can have disproportionately large impact on the There are numerous examples of this, especially in relation to http://bmjopen.bmj.com/ 37 endorsement perception of public even in areas where they have no relevant expertise or anti-vaccine activism, including actors and models being cited for knowledge. their purported knowledge of complex health issues. 38 39 Malicious complaints / abuse of regulatory frameworks 40 Poisoning the The spreading of malicious claims regarding an individual’s professional or “I’ve heard that doctor abuses patients" - Inflammatory slurs such 41 well / Smear personal conduct to undermine them or discourage others from engaging with as these are designed to discredit, and are not in any way 42 campaigns them. substantiated, but calculated to invoke disgust or contempt. 43

44 Vexatious Making calculated complaints to one’s employer or threatening to do so in order Exaggerated / misleading accounts of interactions with public on September 23, 2021 by guest. Protected copyright. 45 complaints to to intimidate them into silence. advocates and demands to censure them, typically aimed at an 46 employers individual’s university or employer. 47 Vexatious Abusing procedures of professional bodies to target a researcher / medic who Registering complaints with a medical regulatory body against a 48 complaints to presents a challenge to a narrative. These may also include unwarranted freedom doctor for their advocacy of evidence-based positions. Physicians 49 regulatory of information requests or vexatious parliamentary questions. especially vulnerable, as typically all complaints must be 50 bodies investigated, regardless of merit. 51 Intimidation Harassment / Harassment can take many forms, and personal abuse is perhaps most common. Abusive language made publicly or in direct messages, and 52 abuse Threats (implied and direct) are often employed. posting of private contact details, phone numbers, addresses etc. 53 54 Threatening to bring an advocate to court for alleged defamation, 55 Legal threats Legal notices and mechanisms can also be utilised to silence researchers also used judiciously to limit independent investigation on 56 questioning a narrative, from cease and desist notices to defamation claims. pseudoscientific narratives 57 Physical Implicit or explicit threats of physical violence is an uUnfortunately, potent Stalking of private abodes, explicit threats, or actions like spitting 58 intimidation method of intimidating scientists into silence. This includes threats of physical etc. There are instances of security being required for scientific 59 violence or rape (the latter usually directed at female discussants) meetings on publicly contentious subjects, due to implications of 60 or threats of violence.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 24 BMJ Open

1 2 3

4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 Results 7 8 9 142 individuals were approached to undertake the survey, with 101 responding (response rate: 10 71.1%). In addition to survey questions on known problems, participants were also invited to expand 11 on noteworthy negative situations they may have encountered while engaging in health outreach. 12 This section was entirely optional, and 53 participants (52.5% of subjects) opted to share their 13 experiences. Example responses are shown below; please note that some responses have been edited 14 or partially redacted to exclude potentially identifying information and preserve anonymity. 15 16 “Accusations - including by one Senator - that [we are] uncaring, dismissive, neglectful, arrogant, or 17 18 paid by pharma companiesFor when peer advocating review for vaccines.” (Misrepresentation) only 19 20 “I find my expertise is questioned - this often seems to be when men find it difficult to accept women 21 with intelligence and qualifications. Sexist insults are a typical go-to response.” (Discreditation) 22 23 “The worst one that hurt me professionally and personally was that activists gathered my emails using 24 [Freedom of Information Requests] and handed chosen packets of them with a story to different 25 reporters.” (Misrepresentation / Discreditation / Dubious Amplification) 26 27 28 “Persistent negative comments on twitter; usually it doesn't last long but it can feel very intense while 29 it's happening!” (Intimidation) 30 31 “I have been served with a SLAPP lawsuit in order to silence my outreach work. Frequently receive 32 harassing emails, malicious comments made on blog.” (Malicious Complaints) 33 34 “Social media co-ordinated intimidation, implied threats of legal action (for defamation). Mocking, 35 undermining, condescension and attacks for being an industry shill, although.. I am just a patient 36 http://bmjopen.bmj.com/ 37 advocate. Being called a liar, that I never had cancer, that I deserved cancer due to my attitude, that I 38 have been mutilated by conventional medical treatment, and that I am no longer a woman (having 39 had mastectomy for cancer). That my cancer will return and I deserve that.” (Dubious Amplification / 40 Misrepresentation / Discreditation) 41 42 “I have had anti-vaccine organizations and individuals attempt to prevent my public appearances and 43 have been the subject of numerous online smear campaigns accusing me of being ‘a shill for Big 44 Pharma’ etc.” (Discreditation / Dubious Amplification) on September 23, 2021 by guest. Protected copyright. 45 46 47 “Those who attack me very frequently try to do it by targeting me at my job, sending bogus complaints 48 to my bosses and the university. From my observation, that is the go-to attack, the first thing these 49 groups do.” (Malicious Complaints) 50 51 “I had to contact the police, who visited the person who was harassing me. I also involved social 52 services. We bought a CCTV to monitor our front door after a strange envelope was hand delivered. 53 The person involved has targeted several people before and continues to target individuals who 54 advocate vaccination.” (Intimidation) 55 56 57 “Abuse and accusations of corruption are the most common adverse reaction I get. Sometimes a 58 particular group petition one's employer and try to create trouble for them. I have been lucky in the 59 past when this happened to have had supportive universities who appreciate my outreach work. I have 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 24

1 2 3 in the past had slightly unhinged individuals writing rambling, implicitly threatening letters to my office 4 which ultimately required police intervention.” (Discreditation / Malicious Complaints / Intimidation) BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 “The worst are gendered insults (being called a cunt, etc.) and rape/death threats. I have had one 8 empty legal threat that was widely publicized.” (Intimidation / Malicious Complaints) 9 10 “Regular threats to sue for defamation.” (Malicious Complaints) 11 12 “Attempts to get me fired, public records act requests for emails, verbal attacks on my children.” 13 (Malicious Complaints, Intimidation) 14 15 16 “One of the most unpleasant things is that certain people or groupings will use very underhanded 17 tactics to respond to perceived criticism. If they can't refute the science, it isn't uncommon for them to 18 go after you personally,For alleging peer all manner ofreview things to anyone onlywho'll listen; that you're incompetent, 19 or unethical, or perverted. It seems they throw things wildly to see what sticks, but it can be 20 extraordinarily unpleasant to endure.” (Dubious Amplification, Discreditation) 21 22 “My main concern has been obsessed individuals who declare their enmity and seem to be 23 unconstrained by civil norms.” (Intimidation) 24 25 26 “Homeopathy advocates looked up my Linkedin profile and called my employer to complain about my 27 comments on the radio. My employer did not support me and I ended up having to stop the activity I 28 had been planning.” (Misrepresentation / Malicious Complaints) 29 30 “Being threatened with physical violence” (Intimidation) 31 32 “A delusional supporter of [an individual] I wrote about accused me and my lawyer of stalking him and 33 killing his in-laws. He sent accusing emails to the faculty of my school and all the police departments 34 35 in my state. [They] also accused me of being a terrorist and complained about me to the FBIs Terrorism

36 Joint Task Force. That gave me many nervous, sleepless nights.” (Discreditation / Malicious http://bmjopen.bmj.com/ 37 Complaints) 38 39 “Death threats received, employer unhelpful, sorted myself” (Intimidation) 40 41 “I haven't experienced many negative encounters because I would say I am only lightly involved in 42 public engagement. However the reason I don't become more heavily involved in this area is fear of 43 this kind of abuse and vexatious complaints to my employer or regulatory body.” (Malicious 44 on September 23, 2021 by guest. Protected copyright. 45 Complaints) 46 47 Based upon this and survey responses, a non-exhaustive taxonomy was constructed detailing common 48 experiences of those communicating medical science to the public. While non-exhaustive, it forms a 49 useful basis for more systematic investigation. Adversarial experiences in communicating medicine 50 were broadly stratified into five distinct classes, illustrated graphically in figure 1, with these sub-types 51 detailed in detail in table 2. 52 53 54 Participant details are given in table 2. Topics covered / channels of engagement and fora for abusive 55 interactions are depicted in figure 2, informed by survey questions 4,5, and 16. Twitter is 56 disproportionately represented, as prominence on that platform was part of the selection criteria. 57 Other fora cited included books, documentaries, newspapers, podcasts - and in one instance criticism 58 under parliamentary privilege. The vast majority of those surveyed (N = 94, 93.1%) reported being the 59 recipient of personal abuse of professional smears in the course of their outreach efforts. A majority 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 24 BMJ Open

1 2 3 (N = 70, 69.3%, survey question 13) had experienced targeted abuse from at least one particular 4 grouping. The most common groupings for negative reactions were anti-vaccine and alternative BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 medicine groups respectively (N = 43 each, 42.6%) followed by dietary advocates (N = 26, 25.7%), 7 ’wellness’ groups (N = 17, 16.8%), religious and chronic illness groups (both N = 15, 14.9%), anti- 8 fluoride and autism focused groups (both N =12, 11.9%). Others cited by 3 or less respondents 9 included electromagnetic hypersensitivity factions, conspiracy theorists and anti-GMO organisations. 10 11 Figure 3 depicts types of experiences reported by participants, ranging from the relatively minor to 12 the severe, informed by survey responses to questions 11, 12, and 14. Of participants surveyed, a 13 majority (N = 63, 62.4%) reported fallout from public engagement had caused them at least some 14 negative mental health sequelae, including depression, anxiety, and stress. Most of this was reported 15 16 as minor, but considerable or significant mental health ramifications were reported by 15 respondents 17 (14.9%). Of the participants, 20 (19.8%) were obligated to seek police advice / legal counsel as a direct 18 result of targeted actionsFor associated peer with their review outreach work. Ofonly those receiving vexatious complaints 19 (N = 39, 38.6% of all respondents), most (67.6%) felt supported or well supported by their institution, 20 employer, or professional body, with whilst 16.2% deemed support to be neutral, with an equal 21 number (16.2%) feeling poorly or entirely unsupported. Predictably perhaps, gender specific abuse 22 was far more likely to be directed at women (40% of female respondents) than men (6.8% of male 23 respondents), with this difference being highly significant (p << 0.001, calculated by Welch’s t-test) 24 25 26 Table 3 depicts frequency of different experiences (positive and negative) reported by respondents, 27 taken from data in survey question 8. In response to survey question 9, 29.7% (N=30) responded that 28 they found outreach largely rewarding, 38.6% rewarding (N=39), 29.7% mixed (N=30), and ~2% not 29 very rewarding (N=2). Changes respondents felt would most improve outreach work is depicted in 30 figure 4 (from survey question 20). Free-form responses to under category of “other” included 31 Improving media science literacy (false balance and issues with platforming of anti-science views were 32 repeatedly mentioned), legal defence funds, better coordination of professional bodies, robust 33 infrastructure on social media to report disinformation, and better support from police organisations. 34 35 Table 3: Frequency of experiences with outreach 36 http://bmjopen.bmj.com/ 37 Statement Always Frequently Sometimes Infrequently Never Unsure Engagement is mutually informative 6 31 52 10 2 0 38 Engagement changes minds & informs 1 22 64 7 1 6 39 My contributions are welcome & appreciated 1 56 39 3 0 0 40 My efforts contribute to public understanding 4 47 49 1 0 0 41 My efforts are taken in good faith 3 51 38 6 0 3 42 My efforts feel futile 1 9 50 34 7 0 43 Engagement takes a toll on mental health 2 12 44 26 16 0

44 on September 23, 2021 by guest. Protected copyright. 45 46 Figure 2 (a) Topics covered by participants (b) channels of engagement for subjects 47 surveyed (c) Fora for negative interactions 48 49 50 Figure 3 Proportion of negative experiences recorded including (a) abuse experienced (b) 51 violence and intimidation (c) vexatious complaints. 52 53 Figure 4 Changes respondents deemed most likely to benefit medical science 54 communication the most. 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 24

1 2 3

4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 8 Discussion 9 10 In a globalised information age, medical science can appear disconnected and aloof from those it 11 serves to help. Educational and professional bodies (including universities and medical centres) have 12 a unique societal role to inform their peers and public on evidence-based medicine, and a 13 14 responsibility to adjust to modern communications realities. We can collectively no longer remain on 15 the fence in supporting health information advocacy. While being mindful not to over-infer from our 16 survey, we can use it as a basis to make some suggestions. It is vital to have proactive strategies in 17 place to support those engaging in medical outreach, and to maintain a high calibre for public 18 discussion. It is also crucialFor that peer those engaging review in outreach are onlycognizant of the potential pitfalls, and 19 afforded ample support. Given the gendered nature of much of the abuse reported, it seems likely 20 that the hostile environment encountered online could dissuade many talented female 21 communicators from engaging fully, to focus on but one example. It is also important to note that 22 23 despite the sometimes fraught nature of medical science outreach, a majority of respondents (68.2%, 24 N = 69) found the undertakings rewarding or very rewarding. More can be done to avoid common 25 pitfalls of this endeavour, however, much of which is outlined in this manuscript. 26 27 One potential weakness of the survey is the potential for ambiguous definition. Alternative medicine, 28 for example, is typically defined as ostensible medical interventions for which there is insufficient or 29 disconfirmatory evidence; for example, the National Science Board define it as referring to “all 30 treatments that have not been proven effective using scientific methods”. As no specific definition 31 32 was given in the survey text, it is possible respondents substituted their own meaning to some extent. 33 As those surveyed were drawn from science communicators with significant media profiles however, 34 it might be expected that their definitions were more unified that a typical respondent might be. There 35 is also some unavoidable ambiguity with terms such as “abuse” and “smears”. There is a level of

36 subjectivity to these terms, which respondents were left to define themselves. This renders the http://bmjopen.bmj.com/ 37 responses potentially subjective, although the free-form responses do indicate behaviours that could 38 be seen as objectively abusive. 39 40 There is also a serious point that must be at least considered – that advocates for medical science may 41 42 on occasion engage in ill-advised tactics or unhelpful rhetoric. Nor does one’s expertise render them 43 infallible, and it is certainly possible that advocates for science might sometimes engage in a

44 counterproductive fashion. To ascertain this requires some context and nuance, especially for on September 23, 2021 by guest. Protected copyright. 45 academic and medical institutions whose staff might be the subject of complaints. But rather than be 46 reactionary, it is imperative that bodies and institutions have robust policy for dealing with issues that 47 might arise, both so they can correct errant behaviour by members and so that they cannot be 48 weaponised by malicious complaints. Based on the feedback to this survey and wider discussion on 49 the topic, we offer the following suggestions to employers and professional bodies whose members 50 51 might engage in the communication of medical science. While by no means comprehensive, the 52 following guidelines might be beneficial towards establishing policy for dealing with issues that can 53 arise. 54 55 Suggested guidelines for professional bodies and employers 56 57 1. Educational / professional organisations must recognise a commitment to support evidence-based 58 actions by their members. This may require oversight of such activities and investment in the 59 governance / training resources to protect members willing to act as advocates. 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 24 BMJ Open

1 2 3 2. Institutions and professional bodies should have robust measures in place to oversee communication

4 activities associated with their members, and to make assessments as to whether individuals are BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 communicating established facts in good faith or are contributing to undermining of fact with 6 potential legal / reputational damage. 7 8 3. When institutions receive complaints regarding members, the subject must be afforded presumption 9 of innocence rather than being served with reactionary and inflexible procedure, lest the institution 10 might become an unwilling tool against science. 11 12 13 4. In case of disputes and complaints, competent and impartial individuals should be engaged to 14 independently assess complaints, cognisant of vital background and context. 15 16 5. Coordination between press offices and those engaged in outreach would improve communication, 17 pre-emptively identifying those likely to be targets for malicious tactics. 18 For peer review only 19 6. Support for those identified as being on engagement frontlines should be maintained, with clear legal 20 advice / institutional support for targeted members. 21 22 7. Organisations must be vocal in supporting public-facing members, willing to issue strong rebuttals of 23 24 vexatious complaints against individuals. 25 26 8. Professional bodies and employers should promote strive to promote both scientific freedom of 27 speech and strive to champion evidence-based advocacy. 28 29 9. When possible, those expected to engage in outreach should be trained in methods that reduce 30 opportunities for personal and professional attacks. 31 32 33 34 Conclusions 35

36 The question of how we best communicate health science in the modern era is an area where more http://bmjopen.bmj.com/ 37 research is urgently required, especially on the role of social media, and optimum ways physicians, 38 researchers, and other public-facing figures can promote good medical science and mitigate 39 40 falsehoods. The suggestions herein ought to be taken as a starting point, with discussion evolving as 41 improved evidence materialises. There are wider problems implicit in all this those communicating 42 science cannot tackle in isolation; social media regulation particularly is a serious issue, both in regards 43 to the spreading of misinformation / dis-information, and with respect to procedures preventing the

44 potential weaponisation of social media platforms. Social media platforms must ultimately be made on September 23, 2021 by guest. Protected copyright. 45 answerable to regulatory over-sight, just as every other important aspect of life is; claims of innocence 46 are unconvincing when their business model is so clearly dependent on advertising engagement at 47 the cost of lives. The problem of poor reporting and false balance [24] in conventional media outlets 48 also must be considered, and there is significant scope for scientists and doctors to contribute to policy 49 50 in these areas. There is ample evidence that physicians and scientists have an important role to play 51 in combatting health disinformation, as has recently been argued by one of the authors in relation to 52 vaccination for British Medical Journal opinion [25]. But equally, it is crucial that those engaging in this 53 vital work have the requisite support from their institutions, so that deleterious consequences of 54 laudable outreach work might be circumvented. It is increasingly clear that disinformation about 55 medicine and illness has become ubiquitous, with severe consequence for both our collective health 56 and public understanding of medical science. Scientists and physicians must be at the vanguard of the 57 pushback against these dangerous falsehoods – our societal well-being depends upon it. 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 12 of 24

1 2 3

4 Electronic Supplementary Material BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 A full version of the questionnaire and further analysis can be found in the electronic supplementary 8 material. 9 10 11 12 Acknowledgements 13 14 15 The authors would like to thank the scientists, physicians, and patient advocates who gave their time 16 to discuss the issues they face in communicating science, and for sharing their insights, and the 17 reviewers for their useful recommendations. In particular, the authors would like to profoundly thank 18 Dr Laura J Brennan forFor her invaluable peer input; Laurareview was present andonly helped conceive the concept from 19 which this paper sprung. She was a passionate advocate of evidence-based science communication 20 and especially for the HPV vaccine, championing it despite enduring a life-limiting and difficult cervical 21 cancer diagnosis. Following a confidence crisis in the vaccine in Ireland driven by anti-vaccine activism, 22 her tireless work was a substantial factor in reversing the damage wrought by disinformation. Laura 23 passed away 20th March 2019, but while we have lost a friend and colleague, hers is a legacy that will 24 25 resonate for generations. Laura embodied precisely how vital health communication is to promoting 26 public health; Ireland, and the world, owe her a huge debt, and we dedicate this work to her memory. 27 28 29 30 Contributor statement 31 DRG, LJB & ROC conceived the concept for this paper, and worked on survey design. DRG & ROC 32 performed the analysis and wrote the manuscript. 33 34 35

36 http://bmjopen.bmj.com/ 37 Competing interests 38 The authors declare that they have no competing interests. 39 40 41 42 Funding statement 43 The authors declare that they have no specific funding for this work. 44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 Data Sharing 49 Full survey forms available in electronic supplementary material. 50 51 52 53 References 54 55 1. Shahab, L., McGowan, J.A., Waller, J., Smith, S.G.: Prevalence of beliefs about actual and mythical causes of cancer and their 56 association with socio-demographic and health-related characteristics: Findings from a cross-sectional survey in england. European Journal 57 of Cancer 103, 308{316 (2018) 58 59 2. Kata, A.: A postmodern pandora’s box: anti-vaccination misinformation on the internet. Vaccine 28(7), 1709-1716 (2010) 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 13 of 24 BMJ Open

1 2 3 3. Poland, G.A., Jacobson, R.M.: The age-old struggle against the antivaccinationists. New England Journal of Medicine 364(2), 97{99 (2011) 4 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 4. Zimmerman, R.K., Wolfe, R.M., Fox, D.E., Fox, J.R., Nowalk, M.P., Troy, J.A., Sharp, L.K.: Vaccine criticism on the world wide web. 6 Journal of medical internet research 7(2), 17 (2005) 7 8 5. Kata, A.: Anti-vaccine activists, web 2.0, and the postmodern paradigm{an overview of tactics and tropes used online by the anti- 9 vaccination movement. Vaccine 30(25), 3778{3789 (2012) 10 6. Grimes, D.R.: On the viability of conspiratorial beliefs. PloS one 11(1), 0147905 (2016) 11 12 7. Goertzel, T.: Conspiracy theories in science. EMBO reports 11(7), 493{499 (2010) 13 14 8. Jolley, D., Douglas, K.M.: The e ects of anti-vaccine conspiracy theories on vaccination intentions. PloS one 9(2), 89177 (2014) 15 9. Thornton, J.: Measles cases in Europe tripled from 2017 to 2018. British Medical Journal Publishing Group (2019) 16 17 10. Organization, W.H., et al.: Ten threats to global health in 2019. 2019. Retrieved January (2019) 18 For peer review only 11. Freeze, R.A., Lehr, J.H.: The Fluoride Wars: How a Modest Public Health Measure Became America’s Longest-Running Political 19 Melodrama. John Wiley & Sons, ??? (2009) 20 21 12. Grimes, D.R.: Commentary on “are fluoride levels in drinking water associated with hypothyroidism prevalence in england? a 22 large observational study of gp practice data and fluoride levels in drinking water". J Epidemiol Community Health 69(7), 616{616 (2015) 23 13. Grimes, D.R., Bishop, D.V.: Distinguishing polemic from commentary in science: Some guidelines illustrated with the case of sage 24 and burgio (2017). Child development 89(1), 141{147 (2018) 25 26 14. Singh, S., Ernst, E.: Trick or Treatment: The Undeniable Facts About Alternative Medicine. WW Norton & Company, (2008) 27 15. Grimes, D.R.: Proposed mechanisms for homeopathy are physically impossible. Focus on Alternative and Complementary 28 Therapies 17(3), 149{155 (2012) 29 30 16. The, L.O.: Oncology," fake" news, and legal liability. The Lancet. Oncology 19(9), 1135 (2018) 31 32 17. Johnson, S.B., Park, H.S., Gross, C.P., James, B.Y.: Complementary medicine, refusal of conventional cancer therapy, and survival among patients with curable cancers. JAMA oncology 4(10), 1375{1381 (2018) 33 34 18. Vosoughi, S., Roy, D., Aral, S.: The spread of true and false news online. Science 359(6380), 1146{1151 (2018) 35 19. Brady, W.J., Wills, J.A., Jost, J.T., Tucker, J.A., Van Bavel, J.J.: Emotion shapes the diffusion of moralized content in social networks. 36 http://bmjopen.bmj.com/ Proceedings of the National Academy of Sciences 114(28), 7313{7318 (2017) 37 38 20. Smith, B.: Dr. mercola: Visionary or quack? Chicago (2012) 39 40 21. Rosenthal, S., Weiss, T.W., Zimet, G.D., Ma, L., Good, M., Vichnin, M.: Predictors of hpv vaccine uptake among women aged 19- 26: importance of a physician’s recommendation. Vaccine 29(5), 890{895 (2011) 41 42 22. Corcoran, B., Clarke, A., Barrett, T.: Rapid response to hpv vaccination crisis in ireland. The Lancet 391(10135), 2103 (2018) 43

44 23. Kahan, D.M.: Misconceptions, misinformation, and the logic of identity-protective cognition (2017) on September 23, 2021 by guest. Protected copyright. 45 24. Grimes, D.R.: A dangerous balancing act. EMBO reports (2019) 46 47 25. Grimes, D.R.: “David Robert Grimes: Vaccines-How Can We Counter Misinformation Online?” The BMJ, 7 Nov. 2019, 48 blogs.bmj.com/bmj/2019/11/06/david-robert-grimes-vaccines-how-can-we-counter-misinformation-online/. 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 14 of 24

Violence / Intimidation Interpersonal Harassment

Malicious Celebrity Complaints Endorsement

Legal threats http://bmjopen.bmj.com/

Poisoning the well

Abuse / on September 23, 2021 by guest. Protected copyright. Threats Vexatious complaints Vexatious complaints to employers to professional bodies

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 15 of 24 (a) Topics coveredBMJ Open by subjects

70 1 2 3 4 5 60 6 7 8 9 10 50 11(%) 12 13 14 15 40 16 17 18 19 30 20 21

22Percentage respondants 23 24 20 25 26 27 28 29 10 30 31 32 33 0 34 / 35 36 37 Cancer Patient Genetics Bioethics & Media disease Medicine 38 Drugs Advocacy Infectious medicine scams Alternative Vaccination Research / Psychology 39 / Health Policy

Interpretation 40 Gender Issues Sexual Health Chronic Illness Reproductive Pharmacology Dietary Science 41 Pseudoscience 42 43 (b) Mediums employed by subjects 44 100 45 46 47 48 49 For peer review only 50 51 80 52 53 54 55 56(%) 57 58 60 59 60

40 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from Percentage respondants

Twitter

Books

Radio (Author) Books YouTube Blogging Radio Podcasts Facebook Instagram (Author) (Advising) (Advising) Television (Advising) Television (Advising) Public talks Periodicals Periodicals (Appearing) (Appearing) (c) Fora for negative interactions

80 http://bmjopen.bmj.com/

(%) 60 on September 23, 2021 by guest. Protected copyright. 40 Percentage respondants

Post

Blogs Radio Phone Twitter E-mail Other Text For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml forums

YouTube Websites Other message Facebook Instagram Television

Social Media (a) AbuseBMJ experienced Open Page 16 of 24

1 2 3 None 4 5 6 7 8 Malicious insinituations 9 10 over personal conduct 11 12 13 14 15 Insults based on 16 appearance / gender / race 17 18 19 20 21 22 Aggressive / intimidating comments 23 24 25 26 27 28 Assertions of dishonesty 29 30 31 32 33 34 35 Allegations of corruption / 'shilling' 36 37 38 39 40 41 Insults over competance / intelligence 42 43 44 45 46 047 20 40 60 80 48 49 Percentage respondants (%) 50 51 (b) Violence and intimidation experienced 52 53 54 55 Violence / 56 57intimidation 58 59 60 For peer review only

None

Threats / implications of violence

Persistent harassment by a group

Repeated unwanted communications BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from

Persistent harassment by an individual

0 10 20 30 40 50 60 Percentage respondants (%) (c) Vexatious complaints experienced

Required investigation 30 Dismissed without investigation

(%) 25 http://bmjopen.bmj.com/

20 on September 23, 2021 by guest. Protected copyright. 15 Percentage respondants

0 Employer / For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtmlLegal Professional Institution Complaint Body Page 17 of 24 BMJ Open

1 2 3None 4 5 6 7 BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 8 9 Other 10 11 12 13 14 Improved emotional 15 16 support services 17 18 19 Clearer guidelines / training 20 21 22 on public engagement 23 24 25 Increased support from 26 For peer review only 27 advocacy / healthcare bodies 28 29 30 31 32 Practical legal advice 33 34 35 36 37 38 Dedicated outreach funding

39 http://bmjopen.bmj.com/ 40 41 42 More proactive stance 43 44 from institutions / bodies 45 46

47 on September 23, 2021 by guest. Protected copyright. 48 Increased support from 49 50 academic institutions 51 52 53 Increased support from 54 55 56 professional bodies 57 58 59 Better protective mechanisms 60 from internet / social media companies

0 10 20 30For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml40 50 60 70 Percentage respondants (%) BMJ Open Page 18 of 24

* Required Information 1 Experiences in public communication of medical science 2 & health 3

29 30 * 3. Please indicate which of the following organsiations you are professionally affiliated with that are directly relevant to 31 your outreach efforts.

32 University / research institute Media outlet 33 34 Medical institute Political grouping 35 NGO / Charity Non-affiliated (Patient Advocate / independent)

44 on September 23, 2021 by guest. Protected copyright. 45 Cancer Chronic illness Bioethics 46 Health Policy Infectious diseases Interpretation of research

47 Reproductive / Sexual Health Pharmacology / drugs Gender issues 48 49 Patient Advocacy Psychological issues Genetics 50 Other (Please specify) ______51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 19 of 24 BMJ Open

1 * 5. What media do you use for public engagement? Select all that apply. 2 3 Television (appearances) Newspapers / Magazines (advising / quoted) Facebook

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 Experiences in public communication of medical science 2 & health 3

39 Personal abuse Sexist or gender-specific abuse 40 41 Personal smears Racist of ethnic-specific abuse 42 Professional smears Intimidation attempts 43 Physical threats Vexatious legal threats

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 21 of 24 BMJ Open

1 * 10. How would you best describe your feelings about your outreach experiences to date? (Select one option) 2 3 Largely rewarding

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 4 For any queries, please contact Dr. David Robert Grimes at [email protected] or [email protected]

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 22 of 24

1 Experiences in public communication of medical science 2 & health 3

49 Vexatious legal complaint - complaint required investigation 50 51 None of the above 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 23 of 24 BMJ Open

1 * 15. If you have experienced a vexatious complaint, did you feel supported by your institution or professional body? (Select 2 one option) 3 Highly supported Not well supported

34 Yes - considerably No 35 Yes - to a minor degree Prefer not to say

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 24 of 24

1 19. Please expand on noteworthy negative situations you have encountered while engaging in health outreach. This is 2 entirely optional, but helps us identify common experiences and pitfalls. 3

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 3 of 4 For any queries, please contact Dr. David Robert Grimes at [email protected] or [email protected]

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 25 of 24 BMJ Open

1 Experiences in public communication of medical science 2 & health 3

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Establishing a taxonomy of potential hazards associated with communicating medical science in the age of disinformation ForJournal: peerBMJ Open review only Manuscript ID bmjopen-2019-035626.R2

Article Type: Original research

Date Submitted by the 13-Apr-2020 Author:

Primary Subject Public health Heading:

Secondary Subject Heading: Medical education and training

PUBLIC HEALTH, INFECTIOUS DISEASES, ONCOLOGY, MEDICAL Keywords:

JOURNALISM, MEDICAL EDUCATION & TRAINING http://bmjopen.bmj.com/

on September 23, 2021 by guest. Protected copyright.

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 1 of 24 BMJ Open

1 2 3

36 http://bmjopen.bmj.com/ 37 38 39 40 41 42 43

44 on September 23, 2021 by guest. Protected copyright. 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 2 of 24

1 2 3

4 Establishing a taxonomy of potential hazards associated with BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 communicating medical science in the age of disinformation 7 David Robert Grimes1,2* Laura J Brennan†, and Robert O’Connor3 8 1. School of physical sciences, Dublin City University, Glasnevin, Dublin 9, Ireland 9 2. Department of Oncology, University of Oxford, Off Roosevelt Rd, Oxford, UK 10 3. Irish Cancer Society, Dublin 4, Dublin, Ireland 11 Correspondence: [email protected] / [email protected] 12 13 14 Abstract 15 Objectives: Disinformation on medical matters has become an increasing public health concern. Public 16 engagement by scientists, clinicians, and patient advocates can contribute towards public 17 18 understanding of medicine.For However,peer depth review of feeling on manyonly issues (notably vaccination and 19 cancer) can lead to adverse reactions for those communicating medical science, including vexatious 20 interactions and targeted campaigns. Our objective in this work is to establish a taxonomy of common 21 negative experiences encountered by those communicating medical science, and suggest guidelines 22 so that they may be circumvented. 23 24 Design: We establish a taxonomy of the common negative experiences reported by those 25 communicating medical science, informed by surveying medical science communicators with public 26 27 platforms. 28 29 Participants: 142 prominent medical science communicators (defined as having >1000 Twitter 30 followers and experience communicating medical science on social and traditional media platforms) 31 were invited to take part in a survey, with 101 responses. 32 33 Results: 101 responses were analysed. Most participants experienced abusive behaviour (91.9%), 34 including persistent harassment (69.3%) and physical violence and intimidation (5.9%). A substantial 35 number (38.6%) received vexatious complaints to their employers, professional bodies, or legal 36 http://bmjopen.bmj.com/ 37 intimidation. The majority (62.4%) reported negative mental health sequelae due to public outreach, 38 including depression, anxiety, and stress. A significant minority (19.8%) were obligated to seek police 39 advice or legal counsel due to actions associated with their outreach work. While the majority targeted 40 with vexatious complaints felt supported by their employer/professional body, 32.4% reported 41 neutral, poor or non-existent support. 42 43 Conclusions: Those engaging in public outreach of medical science are vulnerable to negative

44 on September 23, 2021 by guest. Protected copyright. repercussions, and we suggest guidelines for professional bodies and organisations to remedy some 45 46 of these impacts on front-line members. 47 48 Strengths and limitations 49  Individuals prominently involved in the communication of medical science across different media were 50 surveyed to ascertain their experiences in public outreach. 51  Participants were from around the world, but predominantly communicated in the English language. 52  Self-selection bias in this survey is unavoidable, and findings cannot be taken as generalisable. 53  Accordingly, survey results should only be taken as indicative of the scope of the issue at this juncture. 54  Much further research is needed to ascertain how the medical community can best act to counter the 55 rise of medical disinformation whilst protecting practitioners. 56 57 58 59 60 † Deceased 20th March 2019

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 3 of 24 BMJ Open

1 2 3

4 Introduction BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 Despite being fundamental to societal well-being, many aspects of medical science remain poorly 8 understood and frequently distrusted. Disinformation undermining health science and evidence- 9 based medicine has increased markedly in the era of social media, and dangerous misconceptions 10 abound, from perceived cancer risks and ostensible cures [1] to dangerous falsehoods about 11 vaccination [2]. Improving public awareness and understanding of science and medicine is imperative 12 if we are to maintain continued progress in research endeavours, and scientists, physicians, and 13 science communicators have a crucial role to play in shaping public perceptions. Medical science is 14 largely publicly funded, and direct communication of research with the wider public can be 15 16 extraordinarily beneficial on a societal level. Accordingly, public engagement has become a 17 prerequisite for many funding bodies. Informed engagement by patient advocates and media figures 18 too can have markedFor impact on peer public understanding review of medicine, only empowering the public with facts 19 with which to make important health decisions. 20 21 Improving public understanding of medical science is vital, as there many scenarios where public 22 perception (or a vocal subset of that) is starkly at odds with scientific consensus. Frequently, medical 23 science contradicts a narrative strongly held by particular groups within the wider public. For our 24 25 purposes, we define a ‘narrative’ as a world-view or mindset shared by a given subgroup, which unifies 26 that grouping. Narratives are often articles of faith, empowerment or comfort, frequently 27 unsupported by available evidence or at odds with scientific consensus. For clarity, we concentrate 28 herein on situations where there is no reputable evidence for a narrative, or where overwhelming 29 scientific consensus is firmly against that viewpoint. 30 31 Misguided narratives can be supremely damaging, and the anti-vaccine movement is perhaps the 32 most obvious example of this. Despite the life-saving efficacy of vaccination, opposition has existed 33 since the time of Jenner [3]. The rise of social media has seen significant propagation of anti-vaccine 34 35 narratives [4, 5, 6], driving uptake rates down and causing serious harm worldwide [7, 8]. In 2018,

36 Europe saw the highest number of cases of measles cases in 20 years, numbering over 82,525 cases http://bmjopen.bmj.com/ 37 with at least 72 deaths - over fifteen-fold the figures from 2016 [9]. Such is the extent of the problem 38 that in 2019 the World Health Organisation described vaccine hesitancy as a “Top ten threat to global 39 health" [10]. Exposure to anti-vaccine conspiracy theory is a leading factor in parental intention to 40 vaccinate [8], and evidence to date suggests that the deluge of vaccine disinformation across social 41 media is extremely damaging to public understanding and health. 42 43 Other strongly-held narratives which clash with the weight of available scientific evidence include the 44 on September 23, 2021 by guest. Protected copyright. 45 claims propounded by the anti-fluoride movement [11, 12], the beliefs of the electromagnetic 46 hypersensitivity movement [13], and the narratives of alternative medicine [14, 15]. Cancer patients 47 are especially vulnerable to misinformation, and frequently targeted by charlatans and the misguided 48 [16]. Consequences of this can be severe, with patients sometimes delaying or refusing conventional 49 treatment. The net result of this is diminished survival statistics for those who engage with cancer 50 pseudoscience, due to delayed treatment and sometimes rejection of conventional medicine. In some 51 instances, subscribing to un-proven or disproven modalities could approximately half survival time 52 [17]. 53 54 55 While health falsehoods have always existed, the social media age has created new avenues for 56 misinformation (misinformed advice) and disinformation (deliberate falsehoods) to propagate [16], 57 rapidly bringing discredited ideas and dangerous pseudoscience to vast new audiences. 58 Scaremongering stories from dubious outlets propagate more readily than reliable fact-based 59 information from reputable sources [18, 19]. In 2016, over half of all cancer stories shared on 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 4 of 24

1 2 3 Facebook were medically unsound. Some have harnessed pseudoscience to sell questionable diets, 4 supplements, and books, to the detriment of public understanding. Internet health guru Joseph BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 Mercola, for example, made over $7 million in 2010 alone, proffering dubious treatments and advice 7 [20], including denigration of conventional therapies for cancer. Mercola is far from unique, and such 8 proclamations have huge potential for patient harm. 9 10 To overcome this challenge, public outreach by scientists, physicians, and evidence-based health 11 advocates must be a crucial element to counter damaging fictions and empower our community with 12 evidence-based information. A physician’s recommendation, for example, is central to parental 13 decisions to vaccinate [21]. Addressing patient concerns improves public health, and personal 14 engagement by researchers and physicians can have a positive impact on public perception. Patient 15 16 advocates and media figures have substantial ability to shift public perception; after Ireland saw HPV 17 vaccine uptake drop from 87% to 51%, an alliance of healthcare professionals, researchers, and 18 patient advocates wereFor instrumental peer in countering review the dominant only falsehoods, and Ireland has seen a 19 dramatic recovery in vaccine uptake rates [22]. To make inroads against the deluge of dubious health 20 claims to which we’re subjected, it’s vital that scientists, clinicians, and patient groups must be on the 21 vanguard of efforts to counter-act misinformation. 22 23 Those engaging in public outreach, however, often encounter enmity for publicly advocating scientific 24 25 evidence. Scientific consensus often runs contrary to deeply-held beliefs, leading to certain groups 26 attempting to undermine legitimate public scientific discourse. Motivations for this are multi-faceted, 27 often depending on very specific circumstances. Conspiratorial thinking underpins many narratives, 28 and those attempting to communication science are often vilified as “shills", or agents of a nebulous 29 “Big Pharma". The phenomenon of identity protective cognition is also commonly encountered [23] 30 and narrative believers frequently attacking those who cast doubt on their beliefs. Even when handled 31 with sympathy and compassion, professional and patient advocates who challenge misconceptions 32 can become targets for certain individuals and groups. These negative responses can range from 33 verbal abuse to coordinated harassment campaigns, and even violence. 34 35

36 Aside from being deeply unsettling, such responses can cause professional and personal problems for http://bmjopen.bmj.com/ 37 those targeted. With the increasing emphasis on public engagement and ubiquity of the internet, this 38 subject warrants urgent consideration, as there are currently few clear guidelines for researchers, 39 clinicians, or patient advocates engaging in outreach work. Nor indeed is there a unified understanding 40 of how adverse effects can manifest, and institutions and professional bodies are typically ill-equipped 41 or muted in their support. This leaves those in the public eye or studying contentious topics vulnerable 42 to vexatious complaints and even physical harm. Without awareness of this reality, a less than ideal 43 situation where professional bodies can potentially be weaponised against researchers can too easily 44 on September 23, 2021 by guest. Protected copyright. 45 emerge. Accordingly, the aim of this work is to ascertain the typical experiences of those 46 communicating medical science and identify how negative impacts might be counteracted. 47 48 Figure 1 A non-exhaustive taxonomy of negative experiences encountered by individuals 49 engaging in public communication of health science. Subsections discussed in text. 50 51 52 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 5 of 24 BMJ Open

1 2 3

4 Methods BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 Sample recruitment and selection criteria 8 The main recruitment fora for this study were online discussion groups for physicians, scientists, and 9 10 patient advocates communicating aspects of medical science to the general public across social and 11 traditional media. At the height of the Irish HPV vaccine confidence crisis [22], several physicians and 12 scientists (based in Ireland and the UK) in these groups sought advice for negative experiences, 13 including threats and malicious complaints to their employers and professional bodies, all of which 14 were eventually dismissed. Group members across Europe echoed similar experiences in 15 communicating vaccination science and in other health issues, and almost identical adverse reports 16 came from colleagues across the Americas and Africa. Informal interviews were conducted on foot of 17 this in these fora to identify common issues, as to the authors’ knowledge there is no existent 18 For peer review only literature on the topic. 19 20 21 Based on these interviews and related fora discussions, a survey was created, including free-form 22 sections where subjects were free to expand on their own experiences. The wording of this survey is 23 included in the supplementary material. The participant selection criteria were specifically for those 24 communicating medical science both on social media (defined as having over 1000 followers on 25 twitter) and in mainstream channels (defined as invited appearances on public television, radio, and / 26 or in the form of newspaper articles & invited comment). With this participant selection criteria, 142 27 individuals world-wide (from across Europe, America, Africa, and Asia) working predominantly in the 28 29 English language were identified and invited to partake, of whom 101 (71.1%) responded. This survey 30 was undertaken for indicative purposes and was collected from a non-randomised group with no 31 expectation of transferable findings. Accordingly, the Health Research Authority (HRA) decision tool 32 (online at http://www.hra-decisiontools.org.uk/research/ ) indicated specific ethical approval was not 33 required, with the research governance body of Queen’s University Belfast (the lead author’s primary 34 affiliation at the time) confirming ethical approval for the survey was not required. In all cases, 35 informed consent was sought and obtained prior to subjects partaking, with all data appropriately 36 anonymised. Subject details are given in table 1. http://bmjopen.bmj.com/ 37 38 39 40 Patient and Public Involvement 41 42 43 As patient advocates play a substantial role in combatting misinformation on medical issues, several

44 who met the inclusion criteria were invited to take part, with 15.8% of respondents being patient on September 23, 2021 by guest. Protected copyright. 45 advocates. One co-author of this work (LBJ) was a prominent patient advocate. 46 47 Table 1: Participant details 48 Inclusion and completion Invited to take part (N = 142) 49 Total Completed (N = 101) 50 Gender Female (N=55) 51 Male (N=44) Non-Binary/Undisclosed (N=2) 52 Affiliation University/Medical Centre (N=52) 53 Unaffiliated (N=26) 54 Media organisation (N=20) 55 Charity (N=11) Political organisation (N=4) 56 Profession Medical Professional (N=23) 57 Scientist/Researcher (N=20) 58 Science communications (N=16) 59 Patient Advocate (N=16) 60 Health Policy (N=5)

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 6 of 24

1 2 3 Years active Average: 10.7 years (Range 2-30 years)

4 Table 2 A non-exhaustive taxonomy of common adversarial tactics BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 Misrepresentation of scientific evidence / expertise 6 Straw- Misrepresenting scientific arguments to make them amenable to ridicule or “Mercury is toxic, yet scientists put it in vaccines!” - This 7 manning attack, and to deflect or obscure evidence that undermines a particular narrative. statement belies importance of dose and ignores the fact there is no evidence for harm from thimerosal in vaccines. 8 9 Cherry- Selective, manipulative filtering of scientific evidence or expert statements, taken “THC kills cancer, but doctors don’t want you to know about 10 picking / out of context to undermine evidence-base or promote a narrative. cannabis cures." - THC can kill cells in a petri dish, but killing 11 Quote-mining plated cells is entirely different from treating human cancer. 12 Shifting the Insisting the onus is on scientists to ‘disprove’ claims rather than offering any “GMOs are toxic, and scientists should prove us wrong" - This 13 burden of evidence or rationale for assertions made. assertion is untrue, and onus lies on those making the claim to 14 proof proffer evidence for it. 15 16 Discreditation attempts Questioning Casting doubt on one’s ability to question claims at hand. Typically, ostensible “This patient advocate isn’t an expert, so they can’t say this 17 qualification scepticism is not extended to claims supportive of the narrative. vaccine is safe!" - One does not need to be an expert 18 For peer review onlyimmunologist in this case to accurately reflect medical consensus. 19 20 Alleging Claims that the speaker is compromised due to some apparent conflict of interest “Who’s paying you to say this?" - Unsubstantiated allegation to 21 vested or that experts are otherwise lacking impartiality. deflect from absence of evidence for a narrative or claim. interests 22 23 Asserting Allegations that the scientist, physician, or patient advocate is part of some “She’s part of a pharma cover-up to suppress natural cancer 24 Conspiracy conspiracy to suppress the truth or spread false information, either as a pawn or cures!" - Appeals to conspiracy theory function to distract from 25 theory an active player. lack of evidence. 26 27 Dubious amplification of pseudoscientific narratives 28 Media Targeting traditional or online media outlets to amplify dubious narratives, Pitching dubious health claims to journalists as human interest 29 Targeting typically by-passing gatekeepers (science / health journalists etc) who would stories - This approach was successfully used by anti-vaccine 30 otherwise be more likely to spot pseudoscience. activists to push the dis credited link between autism and the 31 MMR vaccine between 1998-2000. 32 Astroturfing / sock- Use of fake social media accounts / fictitious pressure groups to provide an Example: Accounts which spring up once an initial anti-fact site is 33 puppeting illusion of a wider grass-roots support for a particular narrative blocked but which include misinformation consistent with the 34 originator’s initial social media accounts. 35 Celebrity Celebrities and influencers can have disproportionately large impact on the There are numerous examples of this, especially in relation to 36 http://bmjopen.bmj.com/ endorsement perception of public even in areas where they have no relevant expertise or anti-vaccine activism, including actors and models being cited for 37 knowledge. their purported knowledge of complex health issues. 38 39 Malicious complaints / abuse of regulatory frameworks 40 Poisoning the The spreading of malicious claims regarding an individual’s professional or “I’ve heard that doctor abuses patients" - Inflammatory slurs such 41 well / Smear personal conduct to undermine them or discourage others from engaging with as these are designed to discredit, and are not in any way campaigns them. substantiated, but calculated to invoke disgust or contempt. 42 43

44 Vexatious Making calculated complaints to one’s employer or threatening to do so in order Exaggerated/misleading accounts of interactions with public on September 23, 2021 by guest. Protected copyright. 45 complaints to to intimidate them into silence. advocates and demands to censure them, typically aimed at an employers individual’s university or employer. 46 47 Vexatious Abusing procedures of professional bodies to target a researcher / medic who Registering complaints with a medical regulatory body against a 48 complaints to presents a challenge to a narrative. These may also include unwarranted freedom doctor for their advocacy of evidence-based positions. Physicians 49 regulatory of information requests or vexatious parliamentary questions. especially vulnerable, as typically all complaints must be 50 bodies investigated, regardless of merit. Intimidation 51 Harassment / Harassment can take many forms, and personal abuse is perhaps most common. Abusive language made publicly or in direct messages, and 52 abuse Threats (implied and direct) are often employed. posting of private contact details, phone numbers, addresses etc. 53 54 Threatening to bring an advocate to court for alleged defamation, 55 Legal threats Legal notices and mechanisms can also be utilised to silence researchers also used judiciously to limit independent investigation on questioning a narrative, from cease and desist notices to defamation claims. pseudoscientific narratives 56 57 Physical Implicit or explicit threats of physical violence is an uUnfortunately, potent Stalking of private abodes, explicit threats, or actions like spitting 58 intimidation method of intimidating scientists into silence. This includes threats of physical etc. There are instances of security being required for scientific 59 violence or rape (the latter usually directed at female discussants) meetings on publicly contentious subjects, due to implications of or threats of violence. 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 7 of 24 BMJ Open

1 2 3

4 Results BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 142 individuals were approached to undertake the survey, with 101 responding (response rate: 8 71.1%). In addition to survey questions on known problems, participants were also invited to expand 9 on noteworthy negative situations they may have encountered while engaging in health outreach. 10 This section was entirely optional, and 53 participants (52.5% of subjects) opted to share their 11 experiences. Example responses are shown below; please note that some responses have been edited 12 or partially redacted to exclude potentially identifying information and preserve anonymity. 13 14 15 “Accusations - including by one Senator - that [we are] uncaring, dismissive, neglectful, arrogant, or 16 paid by pharma companies when advocating for vaccines.” (Misrepresentation) 17 18 “I find my expertise isFor questioned peer - this often review seems to be when only men find it difficult to accept women 19 with intelligence and qualifications. Sexist insults are a typical go-to response.” (Discreditation) 20 21 “The worst one that hurt me professionally and personally was that activists gathered my emails using 22 [Freedom of Information Requests] and handed chosen packets of them with a story to different 23 reporters.” (Misrepresentation/Discreditation/Dubious Amplification) 24 25 26 “Persistent negative comments on twitter; usually it doesn't last long but it can feel very intense while 27 it's happening!” (Intimidation) 28 29 “I have been served with a SLAPP lawsuit in order to silence my outreach work. Frequently receive 30 harassing emails, malicious comments made on blog.” (Malicious Complaints) 31 32 “Social media co-ordinated intimidation, implied threats of legal action (for defamation). Mocking, 33 34 undermining, condescension and attacks for being an industry shill, although.. I am just a patient 35 advocate. Being called a liar, that I never had cancer, that I deserved cancer due to my attitude, that I

36 have been mutilated by conventional medical treatment, and that I am no longer a woman (having http://bmjopen.bmj.com/ 37 had mastectomy for cancer). That my cancer will return and I deserve that.” (Dubious Amplification / 38 Misrepresentation/Discreditation) 39 40 “I have had anti-vaccine organizations and individuals attempt to prevent my public appearances and 41 have been the subject of numerous online smear campaigns accusing me of being ‘a shill for Big 42 43 Pharma’ etc.” (Discreditation/Dubious Amplification)

44 on September 23, 2021 by guest. Protected copyright. 45 “Those who attack me very frequently try to do it by targeting me at my job, sending bogus complaints 46 to my bosses and the university. From my observation, that is the go-to attack, the first thing these 47 groups do.” (Malicious Complaints) 48 49 “I had to contact the police, who visited the person who was harassing me. I also involved social 50 services. We bought a CCTV to monitor our front door after a strange envelope was hand delivered. 51 The person involved has targeted several people before and continues to target individuals who 52 53 advocate vaccination.” (Intimidation) 54 55 “Abuse and accusations of corruption are the most common adverse reaction I get. Sometimes a 56 particular group petition one's employer and try to create trouble for them. I have been lucky in the 57 past when this happened to have had supportive universities who appreciate my outreach work. I have 58 in the past had slightly unhinged individuals writing rambling, implicitly threatening letters to my office 59 which ultimately required police intervention.” (Discreditation/ Malicious Complaints/Intimidation) 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 8 of 24

1 2 3

4 “The worst are gendered insults (being called a cunt, etc.) and rape/death threats. I have had one BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 empty legal threat that was widely publicized.” (Intimidation/Malicious Complaints) 7 8 “Regular threats to sue for defamation.” (Malicious Complaints) 9 10 “Attempts to get me fired, public records act requests for emails, verbal attacks on my children.” 11 (Malicious Complaints, Intimidation) 12 13 “One of the most unpleasant things is that certain people or groupings will use very underhanded 14 tactics to respond to perceived criticism. If they can't refute the science, it isn't uncommon for them to 15 16 go after you personally, alleging all manner of things to anyone who'll listen; that you're incompetent, 17 or unethical, or perverted. It seems they throw things wildly to see what sticks, but it can be 18 extraordinarily unpleasantFor to endure.”peer (Dubious review Amplification/Discreditation) only 19 20 “My main concern has been obsessed individuals who declare their enmity and seem to be 21 unconstrained by civil norms.” (Intimidation) 22 23 “Homeopathy advocates looked up my Linkedin profile and called my employer to complain about my 24 25 comments on the radio. My employer did not support me and I ended up having to stop the activity I 26 had been planning.” (Misrepresentation/Malicious Complaints) 27 28 “Being threatened with physical violence” (Intimidation) 29 30 “A delusional supporter of [an individual] I wrote about accused me and my lawyer of stalking him and 31 killing his in-laws. He sent accusing emails to the faculty of my school and all the police departments 32 in my state. [They] also accused me of being a terrorist and complained about me to the FBIs Terrorism 33 Joint Task Force. That gave me many nervous, sleepless nights.” (Discreditation/Malicious Complaints) 34 35

36 “Death threats received, employer unhelpful, sorted myself” (Intimidation) http://bmjopen.bmj.com/ 37 38 “I haven't experienced many negative encounters because I would say I am only lightly involved in 39 public engagement. However the reason I don't become more heavily involved in this area is fear of 40 this kind of abuse and vexatious complaints to my employer or regulatory body.” (Malicious 41 Complaints) 42 43 Based upon this and survey responses, a non-exhaustive taxonomy was constructed detailing common 44 on September 23, 2021 by guest. Protected copyright. 45 experiences of those communicating medical science to the public. While non-exhaustive, it forms a 46 useful basis for more systematic investigation. Adversarial experiences in communicating medicine 47 were broadly stratified into five distinct classes, illustrated graphically in figure 1, with these sub-types 48 detailed in table 2. 49 50 Participant details are given in table 1. Topics covered, channels of engagement, and fora for abusive 51 interactions are depicted in figure 2, informed by survey questions 4, 5, and 16. Twitter is 52 53 disproportionately represented, as prominence on that platform was part of the selection criteria. 54 Other fora cited included books, documentaries, newspapers, podcasts - and in one instance criticism 55 under parliamentary privilege. The vast majority of those surveyed (N = 94, 93.1%) reported being the 56 recipient of personal abuse of professional smears in the course of their outreach efforts. A majority 57 (N = 70, 69.3%, survey question 13) had experienced targeted abuse from at least one particular 58 grouping. The most common groupings for negative reactions were anti-vaccine and alternative 59 medicine groups respectively (N = 43 each, 42.6%) followed by dietary advocates (N = 26, 25.7%), 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 9 of 24 BMJ Open

1 2 3 ’wellness’ groups (N = 17, 16.8%), religious and chronic illness groups (both N = 15, 14.9%), anti- 4 fluoride and autism focused groups (both N =12, 11.9%). Others cited by 3 or less respondents BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 included electromagnetic hypersensitivity factions, conspiracy theorists and anti-GMO organisations. 7 8 Figure 3 depicts types of experiences reported by participants, ranging from the relatively minor to 9 the severe, informed by survey responses to questions 11, 12, and 14. Of participants surveyed, a 10 majority (N = 63, 62.4%) reported fallout from public engagement had caused them at least some 11 negative mental health sequelae, including depression, anxiety, and stress. Most of this was reported 12 as minor, but considerable or significant mental health ramifications were reported by 15 respondents 13 (14.9%). Of the participants, 20 (19.8%) were obligated to seek police advice/legal counsel as a direct 14 result of targeted actions associated with their outreach work. Of those receiving vexatious complaints 15 16 (N = 39, 38.6% of all respondents), most (67.6%) felt supported or well supported by their institution, 17 employer, or professional body, with whilst 16.2% deemed support to be neutral, with an equal 18 number (16.2%) feelingFor poorly peer or entirely unsupported.review Predictably only perhaps, gender specific abuse 19 was far more likely to be directed at women (40% of female respondents) than men (6.8% of male 20 respondents), with this difference being highly significant (p << 0.001, calculated by Welch’s t-test) 21 22 Table 3 depicts frequency of different experiences (positive and negative) reported by respondents, 23 taken from data in survey question 8. In response to survey question 9, 29.7% (N=30) responded that 24 25 they found outreach largely rewarding, 38.6% rewarding (N=39), 29.7% mixed (N=30), and ~2% not 26 very rewarding (N=2). Changes respondents felt would most improve outreach work are depicted in 27 figure 4 (from survey question 20). Free-form responses to this question included; improving the 28 media’s scientific literacy (false balance and the platforming of anti-science views were repeatedly 29 mentioned); the establishment of legal defence funds; better coordination of professional bodies; 30 robust infrastructure on social media to report disinformation, and better support from police 31 organisations. 32 33 Table 3: Frequency of experiences with outreach 34 Statement Always Frequently Sometimes Infrequently Never Unsure 35 Engagement is mutually informative 6 31 52 10 2 0

36 Engagement changes minds & informs 1 22 64 7 1 6 http://bmjopen.bmj.com/ 37 My contributions are welcome & appreciated 1 56 39 3 0 0 38 My efforts contribute to public understanding 4 47 49 1 0 0 39 My efforts are taken in good faith 3 51 38 6 0 3 My efforts feel futile 1 9 50 34 7 0 40 Engagement takes a toll on mental health 2 12 44 26 16 0 41 42 43

44 Figure 2 (a) Topics covered by participants (b) channels of engagement for subjects on September 23, 2021 by guest. Protected copyright. 45 surveyed (c) Fora for negative interactions 46 47 Figure 3 Proportion of negative experiences recorded including (a) abuse experienced (b) 48 49 violence and intimidation (c) vexatious complaints. 50 51 Figure 4 Changes respondents deemed most likely to benefit medical science 52 communication the most. 53 54 55 56 57 58 59 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml BMJ Open Page 10 of 24

1 2 3

4 Discussion BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 6 7 In a globalised information age, medical science can appear disconnected and aloof from those it 8 serves to help. Educational and professional bodies (including universities and medical centres) have 9 a unique societal role to inform their peers and public on evidence-based medicine, and a 10 responsibility to adjust to modern communications realities. We can collectively no longer remain on 11 the fence in supporting health information advocacy. While being mindful not to over-infer from our 12 survey, we can use it as a basis to make some suggestions. It is vital to have proactive strategies in 13 14 place to support those engaging in medical outreach, and to maintain a high calibre for public 15 discussion. It is also crucial that those engaging in outreach are cognizant of the potential pitfalls, and 16 afforded ample support. Given the gendered nature of much of the abuse reported, it seems likely 17 that the hostile environment encountered online could dissuade many talented female 18 communicators fromFor engaging peer fully, to focus review on but one example. only It is also important to note that 19 despite the sometimes fraught nature of medical science outreach, a majority of respondents (68.2%, 20 N = 69) found the undertakings rewarding or very rewarding. This is encouraging, but it is crucial we 21 are aware too of the adverse effects that can arise from communicating medical science, many of 22 which are outlined in this manuscript. 23 24 25 One potential weakness of the survey is the potential for ambiguous definition. Alternative medicine, 26 for example, is typically defined as ostensible medical interventions for which there is insufficient or 27 disconfirmatory evidence; for example, the National Science Board define it as referring to “all 28 treatments that have not been proven effective using scientific methods”. As no specific definition 29 was given in the survey text, it is possible respondents substituted their own meaning to some extent. 30 As those surveyed were drawn from science communicators with significant media profiles however, 31 it might be expected that their definitions were more unified than a typical respondent might be. 32 33 There is also some unavoidable ambiguity with terms such as “abuse” and “smears”. There is a level 34 of subjectivity to these terms, which respondents were left to define themselves. This renders the 35 responses potentially subjective, although the free-form responses do indicate behaviours that could

36 be seen as objectively abusive. http://bmjopen.bmj.com/ 37 38 There is also a serious point that must be at least considered – that advocates for medical science may 39 on occasion engage in ill-advised tactics or unhelpful rhetoric. Nor does one’s expertise render them 40 infallible, and it is certainly possible that advocates for science might sometimes engage in a 41 counterproductive fashion. To ascertain this requires some context and nuance, especially for 42 43 academic and medical institutions whose staff might be the subject of complaints. But rather than be

44 reactionary, it is imperative that bodies and institutions have robust and considered policies for on September 23, 2021 by guest. Protected copyright. 45 dealing with issues that might arise. The benefits of this are two-fold; firstly, so that errant behaviour 46 by members can be corrected. But equally importantly, cognisance of the reality of vexatious 47 complaints also means that bodies and institutions can implement measures to ensure that their 48 procedures cannot be weaponised by malicious actors. Based on the feedback to this survey and wider 49 discussion on the topic, we offer the following suggestions to employers and professional bodies 50 whose members might engage in the communication of medical science. While by no means 51 52 comprehensive, the following guidelines might be beneficial towards establishing policy for dealing 53 with issues that can arise. 54 55 Suggested guidelines for professional bodies and employers 56 57 1. Educational/professional organisations must recognise a commitment to support evidence-based 58 actions by their members. This may require oversight of such activities and investment in the 59 governance/training resources to protect members willing to act as advocates. 60

For peer review only - http://bmjopen.bmj.com/site/about/guidelines.xhtml Page 11 of 24 BMJ Open

1 2 3 2. Institutions and professional bodies should have robust measures in place to oversee communication

4 activities associated with their members, and to make assessments as to whether individuals are BMJ Open: first published as 10.1136/bmjopen-2019-035626 on 5 July 2020. Downloaded from 5 communicating established facts in good faith or are contributing to undermining of facts with 6 potential legal/reputational damage. 7 8 3. When institutions receive complaints regarding members, the subject must be afforded presumption 9 of innocence rather than being served with reactionary and inflexible procedures, lest the institution 10 might become an unwilling tool against science. 11 12 13 4. In case of disputes and complaints, competent and impartial individuals should be engaged to 14 independently assess complaints, cognisant of vital background and context. 15 16 5. Coordination between press offices and those engaged in outreach would improve communication, 17 pre-emptively identifying those likely to be targets for malicious tactics. 18 For peer review only 19 6. Support for those identified as being on engagement frontlines should be maintained, with clear legal 20 advice/institutional support for targeted members. 21 22 7. Organisations must be vocal in supporting public-facing members, willing to issue strong rebuttals of 23 24 vexatious complaints against individuals. 25 26 8. Professional bodies and employers should strive to promote both scientific freedom of speech and to 27 champion evidence-based advocacy. 28 29 9. When possible, those expected to engage in outreach should be trained in methods that reduce 30 opportunities for personal and professional attacks. 31 32 33 34 Conclusions 35

36 The question of how we best communicate health science in the modern era is an area where more http://bmjopen.bmj.com/ 37 research is urgently required, especially on the role of social media, and optimum ways physicians, 38 researchers, and other public-facing figures can promote good medical science and mitigate 39 40 falsehoods. The suggestions herein ought to be taken as a starting point, with discussion evolving as 41 improved evidence materialises. There are wider problems implicit in all this that those 42 communicating science cannot tackle in isolation; social media regulation particularly is a serious 43 issue, both in regards to the spreading of misinformation/disinformation, and with respect to

44 procedures preventing the potential weaponisation of social media platforms. Social media platforms on September 23, 2021 by guest. Protected copyright. 45 must ultimately be made answerable to regulatory over-sight, just as every other important aspect of 46 life is; claims of innocence are unconvincing when their business model is so clearly dependent on 47 advertising engagement at the cost of lives. The problem of poor reporting and false balance [24] in 48 conventional media outlets also must be considered, and there is significant scope for scientists and 49 50 doctors to contribute to policy in these areas. There is ample evidence that physicians and scientists 51 have an important role to play in combatting health disinformation, as has recently been argued by 52 one of the authors in relation to vaccination for British Medical Journal opinion [25]. But equally, it is 53 crucial that those engaging in this vital work have the requisite support from their institutions, so that 54 deleterious consequences of laudable outreach work might be circumvented. It is increasingly clear 55 that disinformation about medicine and illness has become ubiquitous, with severe consequences for 56 both our collective health and public understanding of medical science. Scientists and physicians must 57 be at the vanguard of the pushback against these dangerous falsehoods – our societal well-being 58 59 depends upon it. 60