Table 3.

Pearson correlation coefficients and P values for the search term “fatigue.” Time lags with respect to COVID-19 incidence data are also presented^a.

Lag, daysb	Fatiguec
	r	P value
22	0.3926	<.001
23	0.3782	<.001
24	0.3632	.001
25	0.4947	<.001
26	0.5296	<.001
27	0.5171	<.001
28	0.5480	<.001
29	0.5253	<.001
30	0.4720	<.001
31	0.5342	<.001
32	0.5016	<.001
33	0.5427	<.001
34	0.5521	<.001
35	0.5664	<.001
36	0.6350d	<.001
37	0.4981	<.001
38	0.4711	<.001
39	0.4388	<.001
40	0.4631	<.001
41	0.4915	<.001
42	0.5325	<.001

^aPearson correlation coefficients and P values for fatigue are based on the comparison between Google Trends searches for the term and daily polymerase chain reaction-positive cases.

^bThis column refers to the days of lag between the two variables being compared.

^cThe searches for “fatigue” correlated less strongly than the searches for the symptoms in Table 1, but they do show stronger correlations after 36 days. Therefore, we presented these results in a separate table with a different scale of days.

^dThis is the highest correlation with regard to searches for “fatigue.”