Introduction Our hospital has an established ERAS protocol for colorectal patients. We aimed to assess whether Oesophageal Doppler (OD) and thoracic epidural affected length of stay (LOS) or complications rates.

Method We analysed a prospectively-collected database of patients undergoing colorectal surgery from 25/11/11 to 14/08/13. There were 231 patients; 121 M: 110 F with a median age of 62 yrs (range 18–92). LOS was typically left-skewed and was transformed using log. Ordinal Least Squares regression was conducted with log[LOS] as the dependant variable. Logistic regression was used to identify factors associated with complications. Variables were analysed separately; those with an association at P < 0.10 were combined in a multivariate analysis in a forward stepwise approach.

Results Our ERAS protocol routinely utilised 11 constituents of the ERAS compliance group guidelines. Epidural and IV fluid use were under investigation. The median no of ERAS elements per patient was 9 (IQR 7–10). Patients were dichotomised into ERAS compliant (≥9) 143/231 (62%) or ERAS non-compliant (<9) 88/231 (38%). Resections were classified as colonic 105/231 (45%), rectal/combined 83/231 (36%) or none 43/231 (19%). Approach was laparoscopic in 95/231 (41%) cases; conversion rate 21% (20/95). OD was utilised in 117/231 (51%) cases and was associated with a mean rise in IV fluid given of 702ml (P = 0.004, t-test). Thoracic epidural was placed in 37/231 (16%) patients. Overall complication rate was 94/231 (41%); surgical complications were seen in 32%, renal 9%, respiratory 8% and cardiac 4%.

Multivariate analysis independently linked increased LOS with surgical complications (P < 0.0001), renal complications (P = 0.0001), longer operating times (P = 0.0047), creation of stoma (P = 0.0259) and non-compliance with ERAS protocol (P = 0.056). ERAS compliance did not significantly affect LOS in the laparoscopic cohort, whereas in the open group this retained significance (P = 0.0107).

Development of any complication was associated with ASA grade of >2 (P = 0.001), use of IV opiates >48 hrs (P = 0.0120) and rectal surgery (P = 0.0879). In the laparoscopic cohort, only higher intra-op fluid volume (P = 0.0444) and IV opiates (P = 0.0169) correlated with complication rate. In this cohort, fluid volume predicted complications in the non-OD group (n = 41; P = 0.0253), but not the OD group (n = 52, P = 0.8863).

Conclusion Compliance with the ERAS protocol reduced LOS in patients having open surgery. Boosting outpatient support for patients with new stomas may reduce LOS. OD and epidural were not correlated with LOS. In the laparoscopic cohort, higher complication rates were seen with use of IV opiates >48 hrs and greater volumes of fluid intra-op. IN this group of patients, OD may safely identify patients who require fluid resuscitation whilst avoiding overload.

Disclosure of interest None Declared.