Fluid management of surgical patients has changed dramatically in recent years. Recent studies have shown that so called 'restrictive' fluid strategies can improve outcomes such as length of stay and time to fitness for discharge. Furthermore, new monitoring technologies have been developed that allow the clinician to better monitor the effects of intravenous fluid therapy. The fluid management of free flap reconstruction surgery is controversial. As vasoactive agent use during the anesthetic is not the preferred intervention for the treatment of hypotension, fluid therapy is often administered to correct hemodynamic instability. A liberal fluid strategy may predispose the flap to edema, venous congestion and ultimately flap failure. A restrictive fluid strategy however has (at this institution) anecdotally been reported to lead to a higher incidence of deep venous thrombosis (DVT) and pulmonary embolus (PE). Thus, there is clinical equipoise as to the optimal fluid management for these patients. Non-invasive cardiac output (CO) monitoring based on the arterial pulse contour is a nascent technological advance that has been widely validated in the operating theatre. The calculation of CO is based on the assumption that stroke volume is proportional to the integral of the area under the arterial pressure versus time curve. With these monitors, the variation in stroke volume during mechanical ventilation is also calculated and this can be utilized as an estimate of the patient's volume status. A stroke volume variation (SVV) of less than 15% has been shown to be indicative that a patient is no longer volume responsive. In our study, we plan to recruit patients undergoing microvascular free flap reconstruction. Patients will be randomized to either standard care, or goal directed fluid therapy. All patients will receive a pre-operative 5 ml kg-1 bolus of normal saline. Subsequently, all patients will then receive lactated ringers at a rate of 5 ml kg-1 hr-1. Patients in the intervention group will have their CO and SVV monitored by a Flo-Trac pulse contour monitor (Edwards Life Sciences, Irvine, CA). Patients in the intervention group will receive 6% tetrastarch boluses (Voluven, Fresenius Kabi, Hamburg, Germany) if their SVV exceeds 15%. Patients in the control arm will receive fluid at a rate and type at the discretion of the attending anesthesiologist to maintain hemodynamic stability and a urine output of 0.5 ml kg-1 hr-1. Intra-operative variables studied will include total fluid administered (colloid and crystalloid), urine output, SVV and CO (both at the start and end of surgery). Postoperative variables examined will be urine output, daily fluid balance (days 0-3), incidence of DVT, PE, pulmonary edema, rate of re-operation, daily serum creatinine, chloride and bicarbonate concentration, and incidence of flap failure.

Measurements of flap oxygenation and water content will be made with near infrared spectroscopy at baseline, and during the immediate post-operative period (days 1-7). We hypothesize that the goal directed group would have more intra-operative colloid given with a reduced incidence of adverse outcomes.