Table 2.
Summary of policy impacts derived from PSA.
| Group | Pre | Expected post | Observed post | Change (95% CI) | Difference-in-difference change (95% CI) | |
|---|---|---|---|---|---|---|
| Non-ICU days | PSP | 4411 | 6481 | 5467 | − 1013 [− 2051, − 24] | − 596 [− 942, − 276] |
| Non-PSP | 682 | 847 | 793 | − 55 [− 232, 120] | ||
| ICU days | PSP | 589 | 862 | 751 | − 111 [− 309, 74] | − 172 [− 222, − 148] |
| Non-PSP | 142 | 160 | 172 | 11 [− 30, 55] | ||
| Transfers | PSP-mixed | 140 | 209 | 139 | − 70 [− 108, − 34] | − 83 [− 97, − 83] |
| Non-PSP | 85 | 106 | 115 | 9 [− 15, 33] |
Change column refers to the difference between observed and expected outcomes. Difference-in-difference analysis subtracts non-PSP change from PSP change.
IQR interquartile range reflecting 25th, 75th percentile of simulations, PSP Paediatric Sepsis Pathway, PSP-CH Children’s Hospital included in the PSP, PSP-mixed all participating PSP sites excluding PSP-CH.