Effect of goal-directed haemodynamic therapy in free flap reconstruction for head and neck cancer.

Kim, H J; Kim, E J; Lee, H J; Min, J Y; Kim, T W; Choi, E C; Kim, W S; Koo, B-N

BACKGROUND In free flap reconstruction for head and neck cancer, achieving a haemodynamic target using excessive fluid infusion is associated with decreased flap survival rates and extended hospital stays. We hypothesized that goal-directed haemodynamic therapy would improve flap survival rates and shorten hospitalization periods. METHODS Patients scheduled for free flap reconstruction were randomly assigned to a goal-directed haemodynamic therapy group (n = 31) or a conventional haemodynamic therapy control group (n = 31). The control group received extra bolus fluid and ephedrine or norepinephrine to maintain a mean arterial pressure ≥ 65 mmHg. The goal-directed haemodynamic therapy group received a colloid solution as the extra bolus fluid to maintain a stroke volume variation < 12%; dobutamine, ephedrine, or norepinephrine was administered to maintain a cardiac index ≥ 2.5 l/min/m2 and mean arterial pressure ≥ 65 mmHg. Enhanced recovery after surgery protocols were not used except for fluid therapy. An otolaryngologist blinded to group assignments assessed flap outcomes and classified them as 'survival,' 'at risk' or 'failure.' RESULTS The hospitalization period was not significantly different between the groups. The goal-directed haemodynamic therapy group had significantly shorter intensive care unit stays and a higher flap survival rate. The crystalloid volume was significantly lower in goal-directed haemodynamic therapy group. Reoperation rates, post-operative complications, and laboratory data including inflammatory markers were similar between the groups. CONCLUSION Compared to conventional haemodynamic therapy, goal-directed haemodynamic therapy does not reduce hospitalization periods; it may, however, reduce the length of intensive care unit stays and increase flap survival rates. Further studies including multi-centre trials with larger sample sizes are warranted.