Stillbirth Discourse on Instagram and X (Formerly Twitter): Content Analysis
Notice bibliographique
Résumé
Background: Stillbirth, the loss of a fetus after the 20th week of pregnancy, affects about 1 in 160 deliveries in the United States and nearly 1 in 70 globally. It profoundly affects parents, often resulting in grief, depression, anxiety, and posttraumatic stress disorder, exacerbated by societal stigma and a lack of public awareness. However, no comprehensive analysis has explored social media discussions of stillbirth. Objective: This study aimed to analyze stillbirth-related content on Instagram and X (formerly Twitter) by (1) identifying dominant themes using topic modeling, evaluated using latent Dirichlet allocation, non-negative matrix factorization (NMF), and BERTopic; (2) detecting influential hashtags via co-occurrence network analysis; (3) examining sentiments and emotions using transformer-based models; (4) categorizing visual representations of stillbirth on Instagram (Meta) through manual image analysis with a predefined codebook; and (5) screening for misinformation relating to stillbirth on X. Methods: Stillbirth-related posts were collected via RapidAPI (N=27,395), with Instagram posts (#stillbirth: n=7415; #stillbirthawareness: n=8312; 2023-2024) and X posts (#stillbirth: n=11,668; 2020-2024) analyzed using Python 3.12.7 (Python Software Foundation), with NetworkX for hashtag co-occurrence networks and the PageRank algorithm; comparative analyses were restricted to 2023-2024 due to Instagram application programming interface constraints. Topic modeling was evaluated using latent Dirichlet allocation, NMF, and BERTopic, with coherence scores guiding our model selection. Sentiment and emotion were analyzed using transformer-based RoBERTa and DistilRoBERTa. Misinformation screening was applied to X posts. On Instagram, 2 representative image samples (n=366) were manually categorized using a predefined codebook, with the interrater reliability being assessed using Cohen Kappa. Results: Health-related hashtags (eg, #COVID19) appeared more frequently on X. Topic modeling showed that NMF achieved the highest coherence scores (#stillbirthawareness=0.624 and #stillbirth=0.846 on Instagram, #stillbirth=0.816 on X). Medical misinformation appeared in 27.8% (149/536) of tweets linking COVID-19 vaccines to stillbirth. In the image analysis, "Image of text" was most common, followed by remembrance visuals (eg, gravesites and stillborn infants). The interrater reliability was strong, κ=0.837 (95% CI 0.773-0.891) and κ=0.821 (95% CI 0.755-0.879), with high Pearson correlation (r=0.999; P<.001) and no significant difference (χ²7=12.4; P=.09). The sentiment analysis found that positive sentiments exceeded negative sentiments. The emotion analysis showed that fear and sadness were dominant, with fear being more prevalent on X. Conclusions: Instagram emphasizes emotional expression while X focuses on public health and informational content. Evidence-based communication is necessary to counter misinformation, especially on X, whose real-time affordances amplify fear-based narratives during crises, such as COVID-19. In addition, Instagram's visual and commemorative content offers an opportunity to legitimize parental grief and to validate and humanize loss by directly involving bereaved parents in awareness campaigns. Platform-specific strategies and stronger moderation could enhance health discourse credibility. Future research should examine targeted approaches to counter misinformation and assist affected populations.
Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.
Comment cette classification a été obtenuedéplier
Prédiction machine sur la base complète
Imitation des enseignantsNi prévalence calibrée, ni vérité terrain. Validation humaine à venir. Le volet Gemma est une étiquette directe du modèle pour chaque travail de la base, lue sur la notice réduite au titre. Le volet Codex est un classifieur appris des 10 348 étiquettes directes de Codex et calibré sur les taux pondérés de l'échantillon; les champs sans appui suffisant ne portent aucun appel Codex. Le mode candidate est l'union des deux volets; le consensus est leur intersection. Ces sorties portent le statut machine_predicted_unvalidated et ne sont pas des étiquettes humaines.
Scores du classifieur distillé par catégorie (deux têtes)
| Catégorie | Codex | Gemma |
|---|---|---|
| Métarecherche | 0,001 | 0,004 |
| Méta-épidémiologie (sens strict) | 0,000 | 0,000 |
| Méta-épidémiologie (sens large) | 0,000 | 0,000 |
| Bibliométrie | 0,003 | 0,003 |
| Études des sciences et des technologies | 0,001 | 0,001 |
| Communication savante | 0,001 | 0,002 |
| Science ouverte | 0,000 | 0,001 |
| Intégrité de la recherche | 0,000 | 0,000 |
| Charge utile insuffisante (le modèle a refusé de juger) | 0,003 | 0,001 |
Scores machine (provisoires)
Les deux têtes enseignantes du modèle étudiant, lues sur ce travail. Un score ordonne la base pour la relecture; il n'affirme jamais une catégorie, et le statut de validation accompagne chaque rangée tel quel.
Scores de référence d'un modèle non mature (critères de maturité non atteints, 7 itérations). Un score ordonne; il n'affirme jamais une catégorie.
score_only:v0-immature-baseline · tel quel depuis la passe de notation : score_only signifie que le nombre peut ordonner les travaux, et qu'aucune étiquette de catégorie n'en découleClassification
machine, non validéePrédiction automatique; un appel candidat d’une seule source (Gemma direct ou Codex distillé), pas un consensus.
Le détail, modèle par modèle et score par score, se trouve en fin de page sous « Comment cette classification a été obtenue ».