Table 4. Effect size estimates of main effects from the exploratory analyses.
| Outcome | Weekend service: ‘current’ versus ‘newly developed’ | No weekend service: Trial 1 versus Trial 2 | Sensitivity: Trial 1 with 1-month washout period | ||||
|---|---|---|---|---|---|---|---|
| Effect size | ICC* | Effect size | ICC* | Effect size | Inferiority | ICC* | |
| Primary | |||||||
| Length of stay (days) | 0.12 (−0.20 to 0.43) p = 0.46 |
S: 0.0002 W: 0.06 E: 0.89 |
1.03 (0.73 to 1.32) p < 0.001£ |
S: <0.0001 W: 0.01 E: 0.91 |
0.30 (0.01 to 0.60) p = 0.04£ |
Non-inferior | S: <0.0001 W: 0.05 E: 0.91 |
| Length of stay (log transformed) | 0.05 (0.01 to 0.09) p = 0.02€ |
S: 0.0001 W: 0.12 E: 0.90 |
0.11 (0.07 to 0.15) p < 0.001£ |
S: 0.0001 W: 0.06 E: 0.90 |
0.05 (0.02 to 0.09) p = 0.005£ |
N/A | S: <0.0001 W: 0.01 E: 0.91 |
| Proportion of patients staying longer than expected | 0.01 (−0.01 to 0.04) p = 0.31 |
S: 0.83 W: 0.86 |
0.01 (−0.003 to 0.03) p = 0.004£ |
S: 0.87 W: 0.89 |
0.03 (−0.004 to 0.06) p = 0.09 |
Uncertain inferiority | S: 0.83 W: 0.87 |
| Unplanned readmission within 28 days | 0.01 (−0.01 to 0.02) p = 0.42 |
S: <0.0001 W: 0.45 |
−0.001 (−0.02 to 0.01) p = 0.95 |
S: <0.0001 W: 0.50 |
0.01 (−0.01 to 0.03) p = 0.31 |
Uncertain inferiority | S: <0.0001 W: 0.39 |
| Proportion of patients with any adverse event | −0.002 (−0.01 to 0.01) p = 0.74 |
S: 0.02 W: 0.68 |
0.02 (0.008 to 0.04) p = 0.02£ |
S: <0.0001 W: 0.59 |
0.01 (−0.01 to 0.03) p = 0.33 |
Uncertain inferiority | S: <0.0001 W: 0.71 |
| Secondary | |||||||
| Proportion of patients discharged to aged care | −0.004 (−0.01 to 0.003) p = 0.30 |
S: 0.14 W: 0.58 |
−0.01 (−0.01 to 0.003) p = 0.09 |
S: 0.23 W: 0.63 |
−0.001 (−0.01 to 0.01) p = 0.89 |
N/A | S: 0.31 W: 0.70 |
| Cost to the healthcare system per admission (Australian dollars) | −558 (−1,086 to −30) p = 0.04£ |
S: 0.0001 W: 0.02 E: 0.86 |
1,224 (745 to 1,702) p < 0.001£ |
S: 0.0001 W: 0.01 E: 0.76 |
952 (494 to 1,410) p < 0.001£ |
N/A | S: 0.0001 W: 0.03 E: 0.79 |
| Proportion of patients discharged on a Saturday or Sunday | −0.01 (−0.04 to 0.02) p = 0.34 |
S: <0.0001 W: 0.43 |
0.05 (−0.01 to 0.10) p = 0.10 |
S: <0.0001 W: 0.47 |
0.02 (−0.004 to 0.04) p = 0.12 |
N/A | S: <0.0001 W: 0.51 |
Main effects are interpreted as the impact of being exposed to the ‘current’ service, exposure to Trial 1, and exposure to the ‘current’ service for the three analyses, respectively. Data in parentheses are 95% CIs.
*Intraclass correlation coefficients (ICCs) derived from mixed-effects generalised linear models partitioned at the site (S), ward (W), and patient episode (E) levels.
£Statistically significant (superiority hypothesis, 2-tailed p < 0.05).
N/A, not applicable.