Table 4.
Estimated coefficients obtained using autoregressive integrated moving average with exogenous variables models.
| Variables and parameters | Coefficient (β) | 95% CI | P value | ||||
| Number of hospitalized patients; best model: (1,0,1) | |||||||
|
|
Cta (23)b | –24.12 | –41.08 to –7.16 | .005 | |||
|
|
ARc (1) | .99 | .95 to 1.02 | <.001 | |||
|
|
MAd (1) | –.87 | –.96 to –.78 | <.001 | |||
| Number of COVID-19 deaths; best model: (1,0,2) | |||||||
|
|
Ct (30) | –1.52 | –2.86 to –.18 | .03 | |||
|
|
AR (1) | .96 | .89 to 1.03 | <.001 | |||
|
|
MA (1) | –1.07 | –1.22 to –.92 | <.001 | |||
|
|
MA (2) | .21 | .09 to .34 | .001 | |||
| Number of positive tests; best model: (1,0,1) | |||||||
|
|
Ct (30) | –16.87 | –28.93 to –4.82 | <.001 | |||
|
|
AR (1) | .96 | .84 to 1.07 | <.001 | |||
|
|
MA (1) | –.89 | –1.06 to –.71 | <.001 | |||
| Number of hospitalized patients under 5 years old; best model: (1,0,1) | |||||||
|
|
Ct (9) | –.60 | –1.68 to .47 | .27 | |||
|
|
AR (1) | .96 | .84 to 1.07 | <.001 | |||
|
|
MA (1) | –.89 | –1.06 to –.71 | <.001 | |||
| Number of hospitalized patients aged 5-17 years (1,0,1) | |||||||
|
|
Ct (13) | –.40 | –1.30 to .50 | .39 | |||
|
|
AR (1) | .97 | .92 to 1.03 | <.001 | |||
|
|
MA (1) | –.89 | –.99 to –.79 | <.001 | |||
| Number of hospitalized patients aged 18-59 years; best model: (1,0,1) | |||||||
|
|
Ct (23) | –11.87 | –21.81 to –1.94 | .02 | |||
|
|
AR (1) | .99 | .95 to 1.02 | <.001 | |||
|
|
MA (1) | –.85 | –.94 to –.76 | <.001 | |||
| Number of hospitalized patients over 60 years old; best model: (1,0,1) | |||||||
|
|
Ct (23) | –11.44 | –17.82 to –5.07 | <.001 | |||
|
|
AR (1) | .99 | .96 to 1.02 | <.001 | |||
|
|
MA (1) | –.90 | –.98 to –.81 | <.001 | |||
aCt: cycle threshold.
bThe numbers in parentheses indicate the lag in days.
cAR: autoregressive.
dMA: moving average.