Table 1.
Overall length of stay estimates for England using the AFT and TC method, and for Manchester trusts using the MS method
| Method | Hospital trajectory | Mean | SD | N |
|---|---|---|---|---|
| TC | Hospital admission to outcome (no ICU) | 9.1 | 9.5 | 2794 |
| TC | Hospital admission to outcome (via ICU) | 17.3 | 13.1 | 2517 |
| TC | ICU entry to ICU exit | 13.4 | 13.8 | 1809 |
| TC | Hospital admission to ICU entry | 2.0 | 2.7 | 2983 |
| AFT | Hospital admission to outcome (no ICU) | 8.4 | 8.9 | 2805 |
| AFT | Hospital admission to outcome (via ICU) | 16.2 | 12.0 | 2555 |
| AFT | ICU entry to ICU exit | 12.4 | 12.8 | 1809 |
| AFT | Hospital admission to ICU entry | 2.0 | 2.7 | 2983 |
| Multistate | Hospital admission to outcome (no ICU) | 8.0 | 8.4 | 620 (786) |
| Multistate | Hospital admission to outcome (via ICU) | 29.7 | 22.9 | 73 (101) |
| Multistate | ICU entry to ICU exit | 18.9 | 18.0 | 92 (101) |
| Multistate | Hospital admission to ICU entry | 2.3 | 4.5 | 101 (786) |
Source: own elaboration using CHESS and MFT data. For the multi-state model, the sample size in brackets indicates the observed and censored data (including competing risks), with the first number indicating observed transitions. For TC, for sample size indicates the number of observed transitions, and for AFT the sample size is the number of observed and censored transitions