Binary logistic regression models assessing the probability of readmission
| Variable | p Value | OR (95% CI) | AUC (95% CI); R2 |
| Model 1: LACE score only | |||
| LACE score | <0.0001 | 1.42 (1.41 to 1.43) | AUC=0.773 (0.768 to 0.779); R2=0.180 |
| Model 2: all variables from univariate analysis | |||
| Age | <0.0001 | 1.01 (1.01 to 1.02) | AUC=0.820 (0.815 to 0.825); R2=0.240 |
| Gender (female) | <0.0001 | 0.90 (0.85 to 0.95) | |
| Length of stay | <0.0001 | 0.98 (0.98 to 0.99) | |
| Admission type (emergency) | <0.0001 | 4.18 (3.77 to 4.64) | |
| Comorbidity score | <0.0001 | 0.96 (0.94 to 0.98) | |
| A&E visits in the previous 6 months | <0.0001 | 1.12 (1.11 to 1.13) | |
| LACE score | <0.0001 | 1.23 (1.21 to 1.25) | |
| Episodes per patient | <0.0001 | 1.07 (1.06 to 1.07) | |
| Model 3: four components of the LACE index | |||
| Length of stay | <0.0001 | 1.02 (1.01 to 1.03) | AUC=0.806 (0.801 to 0.812); R2=0.193 |
| Admission type (emergency) | <0.0001 | 4.79 (4.40 to 5.22) | |
| Comorbidity score | <0.0001 | 1.30 (1.26 to 1.33) | |
| A&E visits in the previous 6 months | <0.0001 | 1.28 (1.27 to 1.29) | |
| Model 4: reduced complexity model | |||
| A&E visits in the previous 6 months | <0.0001 | 1.36 (1.35 to 1.38) | AUC=0.815 (0.810 to 0.819); R2=0.130 |
| Episodes per patient in the previous year | <0.0001 | 1.03 (1.02 to 1.04) | |
A&E, accident and emergency; AUC, area under the receiver operating characteristic curve.