Table 2.
Health Library and Physician’s Database searches for smell disorders, taste disorders, and coronavirus fitted with a trend of COVID-19 cases, including P values of cases in model, LRTa, ANOVAb, and AICc, and model improvement information.
| Database | P value of cases in model | LRT ANOVA, P value |
AIC | Model improvement | |
| Health Library |
|
|
|
|
|
|
|
Searches for smell disorders | <.001 | <.001 | From 752.71 to 725.58 (Reduced) | Improved |
|
|
Searches for taste disorders | <.001 | <.001 | From 10,464.04 to 5524.93 (Reduced) | Improved |
|
|
Searches for new coronavirus | <.001 | >.99 | From 1141.26 to 5,642,226.89 (Increased) | Not improved |
| Physician’s Database |
|
|
|
|
|
|
|
Searches for smell disorders | .76 | .77 | From 380.26 to 382.17 (Increased) | Not improved |
|
|
Searches for taste disorders | >.99 | .63 | From 358.68 to 360.46 (Increased) | Not improved |
|
|
Searches for coronavirus | <.001 | .001 | From 754.74 to 745.94 (Reduced) | Improved |
aLRT: likelihood ratio test.
bANOVA: analysis of variance.
cAIC: Akaike information criterion.