Cancer: A model topic for misinformation researchers

Abstract

Although cancer might seem like a niche subject, we argue that it is an excellent model topic for misinformation researchers, and an ideal area of application given its importance for society. We first discuss the prevalence of cancer misinformation online and how it has the potential to cause harm. We next examine the financial incentives for those who profit from disinformation dissemination, how people with cancer are a uniquely vulnerable population, and why trust in science and medical professionals is particularly relevant to this topic. We finally discuss how belief in cancer misinformation has clear objective consequences and can be measured with treatment adherence and health outcomes such as mortality. In sum, cancer misinformation could assist the characterization of misinformation belief and be used to develop tools to combat misinformation in general.

In 2018, a jury ordered Robert O. Young to pay $105 million to a cancer patient. Mr. Young is a naturopath who had persuaded the patient to forgo evidence-based treatment in favor of his alkaline diet, resulting in the patient’s progression to stage 4 cancer [1]. Today, he still pedals his message that cancer is not a cell, but an acidic poisonous liquid, via his personal website, YouTube channel (11,000 subscribers), Instagram (17,000 followers), Twitter (40,000 followers), and over 60 books available for purchase on Amazon. Although he has twice been convicted for posing as a medical doctor without a license and currently faces new felony charges [2], he has an active online presence. Online misinformation is a concern to many people with cancer [3], and in the current information ecosystem discerning fact from fiction can be quite a challenge. The case of Robert O. Young propagating dangerous pseudoscientific beliefs and maintaining a substantial online presence despite legal actions, exemplifies how the dissemination of misinformation in the realm of cancer remains a pressing and critical concern.

Although cancer might seem like a niche subject, we argue that it is an excellent model topic for misinformation researchers, and an ideal area of application given its importance for society. It is estimated that 40% of people will be diagnosed with cancer in their lifetimes [4] and is the second leading cause of death [5], making it relevant to many individuals and their families. Furthermore, cancer is extremely complex. The term is used to describe hundreds of diseases, with thousands of treatment options [6]. Even if a person with curable breast cancer understands a great deal about their own cancer, they will know very little about accurate information and treatment options for a friend’s incurable pancreatic cancer. Unlike other topics, cancer is not a subject taught at schools, and even if it were, it is a fast-moving science where new and better treatments are constantly being developed. In the current paper we discuss why cancer is the ideal topic to study misinformation and intervention: (i) cancer misinformation is extremely prevalent online, (ii) it has potential to cause great harm, (iii) there are significant financial incentives and thus an abundance of disinformation sources, (iv) people with cancer are a uniquely vulnerable population, (v) trust in science and medical professionals is extremely important, (vi) the consequences of belief in misinformation and acting on it are objective.

1.1. Prevalence online

We define cancer misinformation as any information counter to current scientific evidence (or consensus), concerning cancer prevention, cause, or treatment. One of the most comprehensive studies of prevalence to-date was Johnson et al. [7], who examined the amount of misinformation in 200 of the most popular social media articles about cancer on Facebook, Reddit, Pinterest, and Twitter. They found that a third could be considered misinformation. However, the exact percentage of cancer misinformation online differs greatly between studies. See Table 1 where the range of misinformation prevalence is between 11% (for breast cancer content on Pinterest [8]), to 100% (for prostate cancer screening content on TikTok [9]). Future research could examine whether the variation in misinformation prevalence is due to meaningful differences between the platforms, types of cancer, or rather differences in the operationalization of what counts as misinformation between researchers.

Table 1.

Studies reporting prevalence of cancer misinformation online since 2020.

Publication	Cancer type	Platform	Misinformation definition and operationalization	Total cases rated	Percentage
Abramson et al. [9]	Prostate cancer (screening)	TikTok, YouTube	Three reviewers rated content on a 5-point scale (strongly disagree - strongly agree) for "this video does not contain misinformation" compared to most recent American Urological Association and National Comprehensive Cancer Network guidelines; ≥4 considered to be moderate to high misinformation.	The top 50 videos populated for the search term “prostate cancer screening” were rated.	100% (TikTok) and 88% (YouTube)
Johnson et al. [7]	Breast, prostate, colorectal, and lung	Facebook, Reddit, Pinterest, Twitter	Two domain specific oncologist reviewers rated content on 5-point scale (true information-false information) with ≥ 3 considered to be misinformation.	The top 50 articles from each cancer type were rated, 200 articles in total.	33%
Kureyama et al. [10]	Cancer (general)	Twitter	Two physician reviewers rated content on a 5-point scale (true information - false information); ≥ 3 considered to be misinformation.	The top 100 tweets with the highest number of “likes” were rated.	44%
Loeb et al. [11]	Bladder cancer	YouTube	Content coded by urologists on a 5-point scale; >2 considered to be moderate to high amount of misinformation.	The first 150 YouTube videos about “bladder cancer” were rated.	21%
Scott et al. [12]	Prostate cancer	Podcasts	Content compared to national guidelines using a 5-point scale, ≥ 3 considered moderate to high misinformation.	The first 50 unique podcast episodes from Spotify and Apple Podcast for the search term “prostate cancer” were rated, 100 total.	12%
Wilner and Holton [8]	Breast cancer	Pinterest	Content compared to reputable sources such as the National Cancer Institute and coded as containing misinformation if there was one or more false claims.	Using the search terms “breast cancer” and “breast + cancer” 797 Pinterest posts were rated.	11% (51% of pins that made factual claims)
Xu, Myrie et al. [13]	Prostate cancer	Instagram	Content reviewed by physicians with clinical expertise using a 5-point scale; ≥ 3 was considered moderate to high levels of misinformation.	62 still-images and 64 videos were examined.	40%
Xu, Taylor et al. [14]	Prostate cancer	Tik Tok	Content was scored for the presence of misinformation on a 5-point scale.	55 videos with the hashtag “prostate cancer”	15%
Xue et al. [15]	Bladder, prostate, and kidney	TikTok	2 authors coded whether or not the video had obvious misinformation.	61 most watched and relevant videos.	36%

Regardless of the exact prevalence, the non-negligible amount reported in each study makes it ideal to examine misinformation engagement online (exposure, sharing, liking, re-tweeting etc.). Furthermore, it is extremely likely that cancer patients themselves are engaging with the content given that 94% of cancer patients look for cancer information on the internet [16]. In fact, Johnson et al. [7] found that people engaged with false information at a higher rate than neutral or accurate information. This aligns with Loeb et al. [11,17], who found that there was a significant negative correlation between scientific quality and viewer engagement for both prostate and bladder cancers on YouTube. In other words, the lower the quality of the information, the more people who viewed it and gave it a “thumbs up”.

1.2. The potential to cause harm

As a field, we ultimately care about the reduction of online misinformation because of the potential for it to cause real-world harm. Cancer misinformation is the epitome of harmful, given that it can ultimately lead to higher risks of mortality [18]. Johnson et al. [7] asked domain specific oncologists to rate online cancer misinformation for harm. They concluded that out of the misinformation found on Facebook, Reddit, Pinterest, and Twitter, 77% was potentially harmful. This included economic harm (out-of-pocket financial costs associated with treatment or travel), harmful action (potentially toxic effects of the suggested test or treatment), harmful interactions (known or unknown medical interactions with curative therapies), or harmful inaction (information that could lead to delay or not seeking medical attention for treatable or curable conditions). Kureyama et al. [10] replicated Johnson et al.’s [7] methods on Twitter and found that 69% of tweets containing misinformation were potentially harmful. These studies show that the cancer misinformation online is far from benign.

1.3. Financial incentives

One of the reasons for the high prevalence of cancer misinformation is likely because there are substantial financial incentives in this area. With the sheer number of cancer diagnoses, there is a great deal of demand for miracle cancer cures or preventative treatments. This creates an information ecosystem with no shortage of disinformation sources who intentionally spread inaccurate information for their own monetary gain, and sell herbal remedies, educational books, establish treatment institutes, or serve as cancer coaches [19,20]. In 2022, it was estimated that the alternative medicine industry was worth $30 billion in the USA alone [21]. Warner et al. [22] searched Pinterest for “cancer recipes” and found that 68% of creators were attempting to generate a profit, with only 18% of creators disclosing health- or nutrition-related credentials.

Financial incentives have also contributed to the deterioration of the information ecosystem by creating numerous predatory journals: fake journals that publish purely for profit and do not conform to the peer review process [23]. Many disinformation sources publish in predatory journals as an attempt to appear legitimate, allowing them to manufacture scientific productivity metrics like the “h-index” [24]. However, not all people who publish in predatory journals are disinformation sources, some are legitimate scientists who are unaware or looking for easy peer review [25]. Scholars who study cancer are targeted particularly aggressively by predatory journals due to the amount of funding in the biomedical sciences [26]. In fact, Manca et al. [27] found that the National Cancer Institute was the most frequent funder of predatory journal articles from the US National Institutes of Health, showing the desperate need for stricter regulations. In sum, cancer is a good topic for researchers who wish to study the impact of cancer disinformation sources online, the supply and demand of disinformation, as well predatory journals’ financial exploitation of the open access model of publishing.

1.4. Vulnerable population

People with cancer are particularly vulnerable to misinformation for several reasons. First, psychological distress frequently accompanies a cancer diagnosis. People are often desperate and willing to try anything, so are primed to be taken advantage of by disinformation sources promising miracles [28]. Anxiety, stress, and uncertainty can prompt information seeking or information avoidance [29], with 40% of individuals diagnosed within the last 6 months not wanting information about their cancer [30]. Indeed, information avoidance has been shown to be an even bigger predictor of misinformation exposure than information seeking [31]. A cancer diagnosis might also mean that people have less cognitive resources to devote to fact-checking. Cancer-related fatigue is frequently reported by patients during and after treatment for cancer and includes debilitating physical and cognitive symptoms [32]. In fact, cognitive impairments are a frequent treatment-related side effect of cancer survivors that may persist for years after the end of primary treatment [33]. Information overload can also occur, which can lead to psychological distress [34]. This shows how important it is that clear, accurate information is readily available for cancer patients and their families.

1.5. Trust

Trusted sources are essential for the topic of cancer because the accurate information is often highly complex, such as individualized treatment plans. It therefore cannot always be Googled or learned for oneself. Cancer management has continued to increase in complexity due to advances in personalized medicine, novel therapeutics, and technology. Between 1996 and 2019, the National Comprehensive Cancer Network Clinical Practice Guidelines increased 762% in page number (26 to 198 pages) and 370% in mean decision paths available (30 to 111 decision paths; [35]). Thankfully, trust in scientists, doctors, and nurses, is consistently high. For example, nurses have been at the top of the Gallup poll of honesty and ethics for the past 21 years, with doctors a close second [36]. However, a great deal of misinformation research on trust and source credibility has been conducted on political misinformation [37]), where trust is notoriously low. To put this into context, in 2022, 79% of the population said that nurses were high in honesty and ethical standards, but members of congress were second last at 9%, narrowly beating telemarketers [36]. Studying trust in the context of cancer and trusted medical sources might shed new light on misinformation. For example, the finding that expertise is less important than trustworthiness when correcting misinformation [38,39], might be different for health topics (see Vraga and Bode [40]). However, the field is yet to take advantage of cancer misinformation and trust empirically, with most of the research currently qualitative (see Teplinsky et al. [41] and Bylund et al. [42]).

1.6. Consequences of belief as an objective measure

Cancer is a topic where a person’s beliefs about evidence-based medicine is directly related to treatment adherence and their subsequent cancer control. We argue that the immediacy in health outcomes is quite rare in misinformation research. For example, if a person decides not to receive the COVID-19 vaccine, this does not guarantee that they themselves will catch the virus, and if they do, they will not necessarily be hospitalized. However, for cancer, there are very real consequences for health decisions that impact that very individual, sometimes within months. Johnson et al. [18] found that cancer patients who use alternative treatments over conventional treatments had more than a twofold increase in likelihood of death (with colorectal and breast cancer having a 4.5 and 5.6 times greater risk of death, respectively) over a five year period. The alternative medicine group were more likely to be young, female, have a higher income, have a higher level of education, and a more advanced cancer stage. Future research should examine the casual link between belief, behavior, and detrimental health outcomes, which is extremely rare in misinformation research.

Researchers interested in pre-existing worldviews will also be able to examine individuals who are more likely to believe cancer misinformation upon diagnosis. Scherer et al. [43] found that people are more susceptible to health information (including cancer) when they have less education, less health literacy, less trust in health care, and more positive attitudes towards alternative medicine. Belief in alternative medicines is common, with 35% of Americans believing that cancer can be cured through alternative therapies alone without standard cancer treatment [44]. Furthermore, approximately one fifth of people have taken alternative medicine in lieu of evidence-based medicine [45]. See Peterson et al. [46] for a review of complementary and alternative medicine in the context of cancer. Nan et al. [47] found that additional factors contributing to health misinformation susceptibility included reduced knowledge about the subject, lower numeracy and literacy (digital literacy, health, science, media), lower trust in science, conservative ideology, and conspiratorial thinking.

1.7. Future research and challenges

Cancer is a model topic for misinformation researchers and an ideal area of application given its importance for society. However, it is currently poorly characterized and extremely underutilized by researchers. Studies examining engagement (such as likes, shares, and views), and source credibility are rare, and intervention studies are in their infancy. A small body of literature has been steadily growing, including studies on inoculation [48], and combined inoculation and corrections (see Vraga et al. [49]). However, interventions such as warnings, accuracy prompts, lateral reading, social norms, health literacy [50], and algorithmic intervention are lacking and could be extremely informative.

Cancer misinformation is also topical due to recent decisions made by the US government and social media platforms. First, in June 2023, a $154 million research program to advance health communication science from the NIH was “paused” indefinitely the evening prior to release. The announcement stated that the NIH had to reconsider its scope in the context of the current regulatory and legal landscape around communication platforms [51]. Many presume that these decisions were due to mounting political pressure due to the politicization of misinformation research [52,53]. This concerning precedent makes it even more important to pursue research agendas such as cancer misinformation and its amelioration. Second, in August 2023, YouTube announced that it is updating its misinformation policies by removing cancer misinformation about treatment and publishing a playlist of informative videos in collaboration with the Mayo Clinic [54]. There is hope that in the future that channels such as those by Robert O. Young will be removed and that recommender systems won’t expose people to additional misinformation (see Yoon et al. [55] for an example about deworming medication for dogs being presented as an alternative cancer medicine). However, the efficacy of this change should be quantified, should data be made available to researchers [56].

In sum, cancer is an excellent topic for researchers due to the prevalence of misinformation online and its potential to cause harm. The consequences of misinformation belief can also be measured objectively with treatment adherence and health outcomes such as mortality. It is a good subject for researchers who are interested in a host of research directions, including disinformation sources and the importance of trust. Cancer misinformation is not only an interesting and important topic in its own right, but we could learn a great deal about the characterization of misinformation, psychological mechanisms underlying belief updating, and use it to develop new tools for how to combat misinformation more generally.