Table 3.

Characteristics of negative and positive prepandemic (calendar year 2019) tweets and peripandemic (calendar year 2020) tweets.

Tweet sentiment	Total tweets (N=305,321), n (%)	Prepandemic tweets (n=139,561), n (%)	Peripandemic tweets (n=165,760), n (%)	Prepandemic vs peripandemic
				P value	z-statistic	Odds ratio (95% CI)
Negative tweets	156,406 (51.3)	65,164 (46.7)	91,242 (55)	<.001	45.94	1.39 (1.37-1.41)
Positive tweets	62,254 (20.4)	34,633 (24.8)	27,621 (16.7)	<.001	55.402	0.60 (0.59-0.61)
Neutral tweets	86,661 (28.3)	39,764 (28.5)	46,897 (28.3)	.22	1.22	0.99 (0.97-1.00)