Content Analysis of Idiopathic Pulmonary Fibrosis-Related Information on Twitter
Bibliographic record
Abstract
Abstract Background Information regarding idiopathic pulmonary fibrosis (IPF) on the internet is often outdated, inaccurate, and potentially harmful. Twitter is a social media platform that allows users to post content in the form of “tweets”. Objective We sought to assess the prevalence of inaccurate information regarding IPF on Twitter. We hypothesized that foundations and medical organizations would be the least likely to post inaccurate information and that inaccurate tweets would have higher user engagement. Methods All tweets posted between 2011 and 2019 were gathered using “snscrape” on Python 3.8 while searching for the phrase “idiopathic pulmonary fibrosis”. Quantitative analysis was performed to describe trends in IPF-related tweet frequency over time. A subset of tweets made between 2018 and 2019 was screened for verifiable medical statements, which were then analyzed for accuracy compared with contemporary clinical practice guidelines, with descriptive statistics reported. Logistic regression was used to compare tweet accuracy and recommendation of nonindicated therapies across sources, with adjustment for tweet age and character count. Wilcoxon rank-sum tests were used to determine if user engagement (favorites, retweets, and replies) differed between accurate and inaccurate tweets. Results A total of 16,787 tweets were identified between 2011 and 2019. Between 2018 and 2019, 4,861 tweets were included, of which 1,612 (33%) contained verifiable medical statements. Tweets from sources other than foundations or medical organizations were more likely to contain inaccurate information and to recommend nonindicated therapies in both unadjusted and adjusted analyses. News and media sources had the highest odds of communicating potentially harmful information in both adjusted (odds ratio [OR], 12.00; 95% confidence interval [CI], 5.87–27.16) and unadjusted (OR, 11.62; 95% CI, 5.70–26.21) analyses when compared with foundations and medical organizations. Tweets containing inaccurate information had significantly lower numbers of favorites and retweets (P < 0.001 for both). Conclusion Misinformation regarding IPF is present on Twitter and is more often presented by news and media sources. Medically inaccurate tweets displayed less user engagement than accurate tweets. This differs from findings on IPF-related information on YouTube and Facebook, which may reflect differences in both author and consumer qualities across social media platforms.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.002 | 0.002 |
| Meta-epidemiology (narrow) | 0.000 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.002 |
| Science and technology studies | 0.001 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
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.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one teacher head, not a consensus.
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".