MétaCan
Menu
Back to cohort
Record W4291006670 · doi:10.2196/35937

Trends and Characteristics of #HIVPrevention Tweets Posted Between 2014 and 2019: Retrospective Infodemiology Study

2022· article· en· W4291006670 on OpenAlexfundvenueno aff
Raquel Burgess, Josemari Feliciano, Leonardo Lizbinski, Yusuf Ransome

Bibliographic record

VenueJMIR Public Health and Surveillance · 2022
Typearticle
Languageen
FieldSocial Sciences
TopicSocial Media in Health Education
Canadian institutionsnot available
FundersYale School of Public Health, Yale UniversityCanadian Institutes of Health ResearchNational Center for Advancing Translational SciencesNational Institute of Mental HealthYale University
KeywordsCondomCategorizationHuman immunodeficiency virus (HIV)MedicineComputer scienceFamily medicineArtificial intelligence

Abstract

fetched live from OpenAlex

BACKGROUND: Twitter is becoming an increasingly important avenue for people to seek information about HIV prevention. Tweets about HIV prevention may reflect or influence current norms about the acceptability of different HIV prevention methods. Therefore, it may be useful to empirically investigate trends in the level of attention paid to different HIV prevention topics on Twitter over time. OBJECTIVE: The primary objective of this study was to investigate temporal trends in the frequency of tweets about different HIV prevention topics on Twitter between 2014 and 2019. METHODS: We used the Twitter application programming interface to obtain English-language tweets employing #HIVPrevention between January 1, 2014, and December 31, 2019 (n=69,197, globally). Using iterative qualitative content analysis on samples of tweets, we developed a keyword list to categorize the tweets into 10 prevention topics (eg, condom use, preexposure prophylaxis [PrEP]) and compared the frequency of tweets mentioning each topic over time. We assessed the overall change in the proportions of #HIVPrevention tweets mentioning each prevention topic in 2019 as compared with 2014 using chi-square and Fisher exact tests. We also conducted descriptive analyses to identify the accounts posting the most original tweets, the accounts retweeted most frequently, the most frequently used word pairings, and the spatial distribution of tweets in the United States compared with the number of state-level HIV cases. RESULTS: PrEP (13,895 tweets; 20.08% of all included tweets) and HIV testing (7688, 11.11%) were the most frequently mentioned topics, whereas condom use (2941, 4.25%) and postexposure prophylaxis (PEP; 823, 1.19%) were mentioned relatively less frequently. The proportions of tweets mentioning PrEP (327/2251, 14.53%, in 2014, 5067/12,971, 39.1%, in 2019; P≤.001), HIV testing (208/2251, 9.24%, in 2014, 2193/12,971, 16.91% in 2019; P≤.001), and PEP (25/2251, 1.11%, in 2014, 342/12,971, 2.64%, in 2019; P≤.001) were higher in 2019 compared with 2014, whereas the proportions of tweets mentioning abstinence, condom use, circumcision, harm reduction, and gender inequity were lower in 2019 compared with 2014. The top retweeted accounts were mostly UN-affiliated entities; celebrities and HIV advocates were also represented. Geotagged #HIVPrevention tweets in the United States between 2014 and 2019 (n=514) were positively correlated with the number of state-level HIV cases in 2019 (r=0.81, P≤.01). CONCLUSIONS: Twitter may be a useful source for identifying HIV prevention trends. During our evaluation period (2014-2019), the most frequently mentioned prevention topics were PrEP and HIV testing in tweets using #HIVPrevention. Strategic responses to these tweets that provide information about where to get tested or how to obtain PrEP may be potential approaches to reduce HIV incidence.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.001
metaresearch head score (Gemma)0.008
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Observational · Consensus signal: Observational
GenreCandidate signal: Empirical · Consensus signal: Empirical
Teacher disagreement score0.016
Threshold uncertainty score0.031

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0010.008
Meta-epidemiology (narrow)0.0000.000
Meta-epidemiology (broad)0.0000.001
Bibliometrics0.0040.004
Science and technology studies0.0010.000
Scholarly communication0.0010.003
Open science0.0010.002
Research integrity0.0010.001
Insufficient payload (model declined to judge)0.0030.001

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.067
GPT teacher head0.404
Teacher spread0.336 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designObservational
Domainnot available
GenreEmpirical

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations7
Published2022
Admission routes2
Has abstractyes

Explore more

Same venueJMIR Public Health and SurveillanceSame topicSocial Media in Health EducationFrench-language works237,207