Table 2.

Summary of articles planning or evaluating interventions or policies from social media during a disaster.

Disaster category and reference			Disaster type		Disaster year		Disaster location		Mental health issue		Social media platform		Number of posts (number of users)		Analysis
Natural disaster
	Baek et al [48]	Geophysical, accident		2011		Japan		Anxiety		Twitter		179,431 (NRa)		Time-series analysis
Human-made disaster
	Budenz et al [49]	Active shooter		2017		United States		Mental illness stigma		Twitter		38,634 (16,920)		Logistic regression
	Glasgow et al [50]	Active shooter		2011-2012		United States		Coping and social support		Twitter		NR		Classifier (unspecified), qualitative coding analysis
	Jones et al [6]	Active shooter		NR		United States		Psychological distress		Twitter		7824 (2515)		Time-series analysis
Epidemic or pandemic
	Abd-Alrazaq et al [51]	Epidemic or pandemic		2020		Global		Affective response		Twitter		167,073 (160,829)		Latent Dirichlet allocation
	He et al [52]	Epidemic or pandemic		2020		Americas and Europe		Depression, mood instability		YouTube		255 (NR)		Touchpoint needs analysis
	Massaad and Cherfan [53]	Epidemic or pandemic		2020		Undisclosed		Service access/needs		Twitter		41,329 (NR)		Generalized linear regression, k-means clustering
	Wang et al [54]	Epidemic or pandemic		2020		China		Subjective well-being		Sina Weibo		NR (5370)		Regression, analysis of variance
	Zhou et al [55]	Epidemic or pandemic		2020		China		Affective response		Sina Weibo		8,985,221 (NR)		Latent Dirichlet allocation

^aNR: not reported.