Table 3 Frequency with which each candidate variable was selected using backwards model selection in 1000 bootstrap samples
From: Impact of data source and time reference of functional status on hospital mortality prediction
Variable | Model Including Pre-admission ADL Impairment | Model Including On-admission ADL Impairment | ||
|---|---|---|---|---|
Interview | Medical Record | Interview | Medical Record | |
ADL impairment | 1000 | 1000 | 1000 | 614 |
Pneumonia | 996 | 994 | 969 | 999 |
AMI | 945 | 961 | 971 | 963 |
Renal Failure | 828 | 841 | 838 | 832 |
Sepsis | 809 | 808 | 823 | 862 |
CCI ≥ 3 | 750 | 758 | 730 | 902 |
Chinese | 570 | 637 | 611 | 642 |
Age | 269 | 272 | 213 | 633 |
CCI = 0 | ||||
CCI = 1 | 188 | 201 | 232 | 247 |
CCI = 2 | 267 | 255 | 270 | 166 |
Sheltered home resident | 251 | 208 | 147 | 110 |
Diabetes Mellitus with End Organ Damage | 10 | 5 | 2 | 3 |