SHBC1154

K.S.CHAN¹, S.P.JUNNARKAR², B.WANG², Y.P.TAN², J.K.LOW², C.W.T.HUEY², V.G.SHELAT²

MOH Holdings Pte Ltd (MOHH)¹, Tan Tock Seng Hospital²

Pancreaticoduodenectomy is technically challenging with high risk of major morbidity. Prehabilitation aims to improve physiologic function with intent to improve clinical outcomes. This study aims to compare the impact of an outpatient prehabilitation program (PP) versus standard care (SC) in patients undergoing elective pancreaticoduodenectomy.

This is a prospective cohort study comparing patients who underwent PP versus SC prior to elective pancreaticoduodenectomy from Jan 2016-Dec 2020. The PP includes dietician, physiotherapist, case manager and anaesthesia services. Outcomes evaluated include morbidity, length of stay, 30-day re-admission and mortality.

Fifty-one (71.8%) and 20 (28.2%) patients received PP and SC respectively. Median age was 65; 59 (83.1%) patients underwent open surgery. Ductal adenocarcinoma was the commonest histology (49.3%). Overall median operating time and estimated blood loss was 495min(interquartile range (IQR) 425-570) and 300mL(IQR 200-500) respectively. Patient demographics were comparable. Univariate analysis showed comparable length of stay (days) (PP: 12.5(IQR 8.8-18), SC 10(IQR 7.5-15.5), p=0.138), overall morbidity (PP: n=28/50(56%), SC: n=11/21(52.4%), p=0.780), post-operative pancreatic fistula (PP: n=12/50(24%), SC: n=3/21(14.3%), p=0.360) and delayed gastric emptying (PP: n=3/50(6%), SC: 3/21(14.3%), p=0.252). Multivariate analysis showed higher incidence of any morbidity (Odds ratio (OR) 2.01, p=0.546), major morbidity (Clavien-Dindo ≥Grade 3A) (OR 2.98, p=0.471) and 30-day readmission (OR 4.70, p=0.328) in PP but was not statistically significant. Overall 30-day readmission was 23.9% and 90-day mortality was 1.4%.

Prehabilitation for patients undergoing pancreaticoduodenectomy did not improve post-operative outcomes which may be due to selection bias. Data on exercise compliance with larger sample size is required for further validation.