Preoperative counselling and preparation is a key component of ERP protocols. Of equal importance is the preoperative identification of patients at risk of developing perioperative fluid and electrolyte imbalance. Use of the H.E.A.D pneumonic may aid this process:

There is level I evidence in support of shortened preoperative fasting protocols, and numerous national anaesthetic societies now permit solid food intake up to 6 h before and clear noncarbonated fluids up to 2 h before induction of anaesthesia. However, patients may not readily follow these guidelines and in clinical practice (even within the context of ERP protocols) alterations in theatre schedules mean it not uncommon to find patients fasted from food for longer periods (even up to 18 h). It is, therefore, of importance that patients are encouraged and provided the opportunity to maintain oral fluid intake (ideally carbohydrate containing drinks) up to 2 h preoperatively to avoid fluid depletion. Similarly, mechanical bowel preparation leads to losses of salt and water, and does not seem to decrease risk of infection when used without oral antibiotics, at least in open colon resection [11]. The routine use of mechanical bowel preparation is discouraged in ERP protocols [6]. If used, patients should receive supplemental intravenous fluid therapy to replace GI losses and ensure zero fluid balance. Induction of anaesthesia in patients with a fluid deficit further reduces the effective circulatory volume by decreasing sympathetic tone. Finally, use of premedication, hypnotics and long-acting sedatives reduce patients’ ability to drink and mobilise postoperatively thereby hampering early recovery, which is a key aim of ERP protocols.

Intraoperative assessment of fluid status is difficult, as a formal physical examination cannot always be conducted. Traditionally heart rate, blood pressure, and urine output have been used to guide intraoperative fluid therapy; however, volume deficits may not become apparent until they exceed 10 % of body weight. Other common intraoperative confounders such as activation of nociceptive pathways by surgical stimulation and changes in body temperature may distort interpretation of the patient’s real-time volume status. Finally, the use of static measurements such as end-diastolic and central venous pressure to estimate volume responsiveness can be influenced by numerous factors including comorbid cardiovascular pathologies and CO₂ pneumoperitoneum during laparoscopic surgery. The latter affects cardiovascular parameters through effects on reduced preload, and hypercarbia-induced vasodilatation and myocardial depression. Goal-directed therapy (GDT) principally aims to guide intravenous fluid and vasopressor/inotropic therapy using measurements of cardiac output or other similar parameters to improve stroke volume, cardiac index and splanchnic perfusion. A number of devices such as the transesophageal Doppler (TOD), arterial pulse contour analysis, lithium dilution and transpulmonary thermodilution techniques can be used to monitor and direct GDT. Algorithms usually involve intraoperative measurement of stroke volume corrected flow time (FTc) in the descending aorta and administering a 200–250 mL fluid bolus over 5–10 min if FTc is <0.35 s. A stroke volume increase of more than 10 % or an FTc < 0.35 s indicates intravascular hypovolemia. Conversely, if the stroke volume does not increase after the initial bolus or if the FTc is >0.4 s, a further bolus is not necessary and background continuous crystalloid infusion is continued. Such an individualised intraoperative algorithm improves splanchnic perfusion without causing excessive interstitial oedema. Previous meta-analyses comparing GDT with conventional therapy reported reduced incidence of postoperative and gastrointestinal complications, need for ICU stay and LOHS in patients undergoing major surgery [12, 13]. However, recent meta-analysis of 31 studies of 5292 patients did not demonstrate difference in mortality [5]. Furthermore, appraisal of these ‘historical’ GDT studies highlighted no data comparing GDT with patients receiving ‘restrictive’ fluid therapy (near-zero fluid balance). Two recent studies have demonstrated no differences between GDT and no GDT in patients managed with ERP protocols, with avoidance of postoperative fluid overload in occurrence of postoperative complications and LOHS [14, 15]. Thus, the value of GDT within the setting of ERP patients receiving zero-balance fluid therapy remains unclear. Moreover, the use of hydroxyethyl starch for GDT has been limited severely by the European Medicines Agency’s cautions [16] on the use of starch in light of the recent randomised controlled trials [17–20] showing harm caused by starch when used for resuscitation. However, a recent study has suggested that either crystalloid or HES may be used with equal efficacy for flow-directed fluid therapy, but both groups of patients received in excess of 5 L of fluid on the day of surgery, which is far in excess of what most patients being managed with ERP protocols receive [21]. A number of hospitals have now moved to gelatins for GDT, but at present there is no evidence to suggest that gelatins are equivalent to starches for this indication. Surprisingly, there is little published data on the utility of goal-directed technologies (GDT) within the setting of laparoscopic (CO₂ pneumoperitoneum) major abdominal surgery. Furthermore, the specific effects of differing pressures of CO₂ pneumoperitoneum (low-pressure [e.g. 8–10 mmHg] vs. normal/high pressure [e.g. 12–15 mmHg]) and steep Trendelenburg/reverse Trendelenburg positioning on GDT parameters have not been studied. That said, however, preliminary data suggest haemodynamic consequences in fluid optimised patients with pneumoperitoneum include decreased stroke volume, cardiac output, and oxygen delivery, coupled with increased systemic vascular resistance [22]. A study in patients undergoing laparoscopic colorectal surgery within an enhanced recovery setting failed to demonstrate differences in intraoperative indexed oxygen delivery (DO₂I) between patients who received epidural versus spinal analgesia. Furthermore, this study identified patients with DO₂I <400 mL/min/m² had higher incidence of anastomotic leak (22 %) than patients with DO₂I >400 mL/min/m² (1.8 %) [22].