Table 1 Thematic Analysis (10) and Structural Topic Modelling Topics (30) Described and Organised.

From: What do people really think about the RSV vaccine? Study of unsolicited text replies from adults over 60

Thematic Analysis

Structural Topic Modelling

Topic

Brief Description

Prev.

Topic

Brief Description

Prev.

Unknown sender

Unknown who the sender is, curiosity.

0.88%

NA

  

Wrong recipient

Who this is, passed away, people who’ve moved.

0.81%

Wrong recipient

Key words: PASSED AWAY, with a mix of Medicare, Medicaid, cost, insurance. Mixed valence.

2.92%

Help

Replies that solicit more information.

7.01%

Sum prevalence of all sub-topics

19.04%

RSV vs. flu FAQ

Most inform about receiving both RSV and flu vaccines. Some confusion about the difference between the two, inquiries. Generally positive.

7.10%

An RSV what?

Questions regarding “What is an RSV vaccine?” Many have the sentiment of not knowing what the acronym stands for. Mixed valence.

4.64%

Vaccine vagaries

Questions about the vaccine, a fair amount of person names, and action language i.e. “I will get.” Mixed valence.

2.95%

Hints of hesitation

Most suggest inquiry with words like “could,” “do we,” “really then why” or uncertainty about what to do. Generally negative.

2.74%

Help me out

Key word: NEED. Do I need a…, I need a refill, need a ride. Generally positive.

1.61%

Image related

Images not displaying, comments on images.

0.14%

NA

  

Stop

Indented to opt out of the communications. Misspellings.

17.82%

Sum prevalence of all sub-topics

14.35%

Complex quits

Unsubscribe keywords along with additional text. Mixed valence.

6.32%

Kind declines

Opt out messages with more varied structure. Pleasant in tone, courteous. Positive.

4.39%

Frustrated farewells

Key words: STOP, UGH. Emotions, frustration, disappointment, “come on.” Negative.

3.63%

Already vaccinated

The sender, and often other family members, have already gotten one or both shots.

8.16%

Sum prevalence of all sub-topics

6.12%

Already vaccinated (location)

Heavily focused on WHERE the RSV vaccine was obtained. Positive.

3.96%

Already vaccinated (time)

Heavily focused on WHEN the RSV vaccine was obtained. Positive.

2.16%

Will not get vaccinated

No to the vaccine and unhappy replies such as negative sentiment, swear words, etc.

20.49%

Sum prevalence of all sub-topics

22.80%

Pleading no with a side of adverse event

Refusals. Often repeated use of “no,” references to medical reasons of contraindication. Frustrated or desperate typing. Negative.

8.05%

I don’t want [something]

Something not wanted, whether the vaccine, the message, or an unrelated item/event. Negative.

4.92%

Moral disgust

Key words: extremely aggressive swear words + ETHICS. Corrupt pharma, lies, poison. Negative.

2.95%

Of God and Country

Religious messages, lack of trust in government, highly moralised anti-vaccine language. Negative.

2.08%

COVID-19 related

Key words: COVID-19. Flavour of distrust. Negative.

2.23%

Hard no

“I do not take vaccines,” “I will never…” but with a flavour of being contraindicated. Negative.

1.34%

Distrust of pharma

Accusing pharma of high prices, insufficiently researched products, and malfeasance. Negative.

1.23%

Pharmacy-related

Mentions of [pharmacy] that do not belong to the other topics.

3.69%

NA

See topics under Help.

 

Benign

Readable and pleasant but not always tied to the intervention.

15.14%

Sum prevalence of all sub-topics

7.83%

Future intention

Key word: WILL. “I will come in,” “I will pick up,” “I will get back to you.” Mixed valence.

3.62%

Pollyanna

Key word: LOVE. “morning, babe, I love you,” “having a good day,” heartwitharrow. Positive.

2.51%

Food

Key words related to food. Grocery lists, recipes, etc. No valence.

1.70%

Actions, auto-replies, etc.

Auto-replies, actions, illegible, blank, emojis, etc.

25.85%

Sum prevalence of all sub-topics

26.94%

Corporate replies

Feature trademarks/wordmarks, product words, or repeated same emoji. No valence.

6.50%

Emojis as reactions

Key words: [emoji] to “[pharmacy]: [name]…”). Tearsofjoy, cryingface, etc. Mixed valence.

3.35%

Mostly happy emojis

Key words: LOL + emojis. Positive.

3.17%

Just nonsense

Just nonsense. No valence.

3.08%

Non-English responses

Spanish language and nonsense words. No valence.

3.00%

Emoji potpourri

Multiple different emojis. No clear theme or meaning. Mixed valence.

2.63%

Distrustful nonsense

Nonsense with a flavour of distrust. Negative.

2.37%

Semantic disintegrity

Repeated characters. No semantic integrity. No valence.

1.82%

Foreign language

French words, non-English proper names, or legible typos. No valence.

1.02%