Logistic regression model used to predict 30-day mortality
| Variable | Number | OR (95% CI) | P value |
| Previous emergency admissions | 5 249 891 | 1.03 (1.03 to 1.04) | <0.001 |
| Previous emergency admissions2 | 5 249 891 | 1.01 (1.01 to 1.01) | <0.001 |
| Previous ED attendances | 5 249 891 | 1.19 (1.18 to 1.20) | <0.001 |
| Previous ED attendances2 | 5 249 891 | 0.98 (0.98 to 0.98) | <0.001 |
| Elixhauser Comorbidity Index - vWs | 5 249 891 | 1.10 (1.10 to 1.10) | <0.001 |
| Elixhauser Comorbidity Index - vWs2 | 5 249 891 | 1.00 (1.00 to 1.00) | <0.001 |
| Age | 5 249 891 | 1.08 (1.08 to 1.08) | <0.001 |
| Age2 | 5 249 891 | 1.00 (1.00 to 1.00) | <0.001 |
| Four-hour breach rate in ED | 5 249 891 | 1.35 (1.33 to 1.37) | <0.001 |
| Male | 2 524 541 | Reference | |
| Female | 2 725 350 | 0.82 (0.81 to 0.82) | <0.001 |
| Deprivation decile | |||
| 0 (deprivation unknown)* | 52 384 | 1.06 (1.01 to 1.10) | <0.001 |
| 1 (least deprived) | 369 831 | Reference | |
| 2 | 448 792 | 1.04 (1.03 to 1.06) | <0.001 |
| 3 | 461 779 | 1.09 (1.07 to 1.11) | <0.001 |
| 4 | 478 846 | 1.12 (1.10 to 1.13) | <0.001 |
| 5 | 502 621 | 1.13 (1.11 to 1.14) | <0.001 |
| 6 | 523 041 | 1.15 (1.14 to 1.17) | <0.001 |
| 7 | 545 037 | 1.20 (1.18 to 1.22) | <0.001 |
| 8 | 576 244 | 1.23 (1.21 to 1.25) | <0.001 |
| 9 | 614 493 | 1.27 (1.25 to 1.29) | <0.001 |
| 10 (most deprived) | 676 823 | 1.32 (1.30 to 1.34) | <0.001 |
| Hour of attendance | |||
| 00:00-05:59 | 611 073 | Reference | |
| 06:00-11:59 | 1 268 524 | 0.99 (0.98 to 1.00) | 0.075 |
| 12:00-17:59 | 1 898 281 | 1.01 (1.00 to 1.02) | <0.001 |
| 18:00-23:59 | 1 472 013 | 0.98 (0.96 to 0.99) | <0.001 |
| Month of attendance | |||
| January to March | 1 485 052 | Reference | |
| April to June | 1 158 860 | 0.74 (0.73 to 0.75) | <0.001 |
| July to September | 1 226 749 | 0.63 (0.62 to 0.64) | <0.001 |
| October to December | 1 379 230 | 0.63 (0.63 to 0.64) | <0.001 |
| Year of attendance | |||
| 2016 | 1 654 846 | Reference | |
| 2017 | 2 700 055 | 0.60 (0.60 to 0.61) | <0.001 |
| 2018 | 894 990 | 0.27 (0.27 to 0.28) | <0.001 |
| ROC curve statistic=0.862 |
Total number of patients = 5 249 891.
2Denotes the use of quadratic terms (ie, the squares of the metrics) that were used in order to account for potential non-linearities in these continuous variables.
*Unknown deprivation includes missing data, overseas visitors, people with no permanent address and those with a new address that had not yet been classified.
ED, emergency department; ROC=receiver operating characteristic; vW, summary score developed by van Walraven.18