Interrupted time series regression results assessing the impact of ERAS on LOS*
| Covariates | Parameter (SE) | P value |
| Intercept | 23.3 (16.9) | 0.178 |
| Average age | −0.03 (0.30) | 0.909 |
| Per cent open | 0.77 (0.11) | <0.001 |
| Per cent weekend | 0.45 (0.22) | 0.054 |
| Per cent ASA mild | −0.04 (0.07) | 0.556 |
| Per cent ASA missing | −0.12 (0.12) | 0.351 |
| Baseline trend | −0.43 (0.15) | 0.008 |
| Level change after intervention | 1.01 (2.00) | 0.621 |
| Trend change after intervention | 0.32 (0.17) | 0.067 |
| Adjusted R-square | 0.71 |
*We modelled monthly average LOS as a function of trend (procedure month), intervention period (0, 1), and covariates where there was significant difference between the preintervention and postintervention samples (age, ASA rating and weekend). The data had 40 periods (months).
ASA, American Society of Anesthesia; ERAS, enhanced recovery after surgery; LOS, length of stay.