The COVID-19 Pandemic: Understanding COVID Vaccine Skeptics: News from Emergency Medicine

Some patients are suspicious of a loss and use disinformation as a means of coping

COVID-19 vaccine, disinformation, unproven therapies, personal loss:

Figure 1. Word frequencies in a COVID Facebook group

FU2-16
figure

Disinformation is a rapidly growing problem, and there are few better examples than the impact on medical care in the face of a global pandemic.

Even as excitement about the COVID-19 pandemic seems to be waning, I still find that speaking up about COVID with my patients — from asking about vaccination status to discussing tests — can often lead to tense interactions that tend to be medical disinformation. It may be difficult to empathize with patients whose ideas about medical reality are so different or even directly opposed to ours, but we must try to understand these patients, because ultimately our goal is to help them.

The methods

To better understand patients whose views on COVID-19 have been heavily influenced by disinformation, I looked for places to find an organic, ongoing source of COVID-specific disinformation. I didn’t have to look far and joined a Facebook group with more than 40,000 members whose purpose was to discuss the harms of the COVID vaccine.

I followed the group for more than a year and manually collected the most recent posts and comments for about two months to create a text-based dataset for analysis.

Results

I applied several text analytical methods to this dataset. First, I looked at raw word frequencies and created a word cloud to demonstrate them. (Figure 1.) Larger words represent a higher number of occurrences. I was struck by how often personal terms like “mom,” “family,” “friend,” and “husband” came up. The terms “God” and “prayers” also appeared frequently. A large number of words related to medical procedures, sometimes with slang (e.g., jabbed means “vaccinated”). These results seem to demonstrate a more personal, even spiritual, approach to medical beliefs than a scientific one.

Next I looked at the frequency inverse document frequency (tf-idf). This measures how many times a term occurs in a single piece of text, and the inverse of how many pieces of text it occurs. The goal is to find words that convey large amounts of information by finding an optimal balance between occurring all the time (e.g. the word “that”) and almost never (like a person’s name). The selected words with high tf-idf scores imply a high level of distrust towards the medical community, even suggesting a conspiracy. (Figure 2.)

I also tried to see what words occur together by creating a graphical representation of the data set, a visual representation of how words are linked together. (Figure 3.) This can be thought of in a similar way to how an autocomplete program might select the next word. Specifically, this graphic highlighted specific medical concerns (e.g., blood clots) as well as recommendations that were frequently discussed in the group: visiting the Front Line COVID-19 Critical Care Alliance website, using bentonite clay, and discussing heavy metals (and chelation therapy).

Finally, I conducted sentiment analysis using the NRC Word-Emotion Lexicon, comparing the entire text of this Facebook group to the text of several of mine News from emergency medicine Article. (Figure 4.) Negative mood, anxiety, sadness, and anger were much more common in the Facebook group than in the EMN Article.

Conclusions

From this analysis we can learn several things. First, medical disinformation is an emotional problem. People may turn to disinformation in response to negativity, fear, or sadness. It is possible that this is related to experiences with friends or family members.

Second, group members have a strong desire to help in a situation they perceive to be bad; They propose unproven therapies and try to share sources of disinformation with the aim of “opening your eyes”. These efforts may ultimately be misguided, but these negative results may actually reflect ambitious tendencies in the face of adversity.

FU3-16
Figure 2.:

Words that indicate great distrust in the medical community

FU4-16
Figure 3.:

Links to medical concerns and recommendations on the COVID Facebook group

FU5-16
Figure 4.:

Emotional sentiment in Facebook group texts compared to EMN articles

It can be helpful for us doctors to frame our patients’ suffering from disinformation in this context. It may be tempting to engage in a scientific debate with these patients, but it is likely that what is at stake is a primarily deep-seated emotional issue rooted in personal loss or tragedy, and the person is using the disinformation as an unproductive coping strategy.

It’s not as satisfying as winning an argument, but offering sympathy and understanding can be our best way to help these patients.

dr Belangeris the Chief Data Officer of TotalCare (totalcare.us), the President-elect of the American College of Emergency Physicians Manpower Division and Emergency Medical Officer in McKinney, TX.

Leave a Comment