Table 3.
Ordinal logistic (GENLIN) models were predicting determinants of ERA and NUERA.
| Hypothesis test | 95% confidence interval | |||||||
|---|---|---|---|---|---|---|---|---|
| Predictors of ERA | β | Std. Er. | Wald χ2 | df | p-value | OR | Lower | Upper |
| Gender (male) | 0.434 | 0.1198 | 13.126 | 1 | <0.001 | 1.544 | 1.221 | 1.952 |
| Frailty | 0.256 | 0.0663 | 14.844 | 1 | <0.001 | 1.291 | 1.134 | 1.47 |
| Comorbidity | 0.123 | 0.0313 | 15.356 | 1 | <0.001 | 1.131 | 1.063 | 1.202 |
| Predictors of NUERA | ||||||||
| Social network | −0.607 | 0.2498 | 5.901 | 1 | 0.015 | 0.545 | 0.334 | 0.889 |
| Pulmo-Cardio-Vascular Function (Medium score) |
–.410 | 0.1484 | 7.620 | 1 | 0.006 | 1.506 | 1.126 | 2.015 |
| Pulmo-Cardio-Vascular Function (Bad score) |
−0.053 | 0.3479 | 0.023 | 1 | 0.880 | 0.949 | 0.480 | 1.876 |
ERA, emergency room access. Dependent variables: absolute number of ERA and NUERA. Covariate's variable = Comorbidity and frailty. Test omnibus χ2(3) =51.73, P < 0.001; ERA AIC = 662.89 and BIC = 698.77. NUERA, not-urgent emergency room access. Dichotomic variable = Social network. NUERA test omnibus χ2(3) = 14.049, P = 0.003; NUERA Akaike information criteria (AIC) = 93.75 and Bayesian information criteria (BIC) = 124.52.