Standard perioperative management in gastrointestinal surgery

Preoperative laboratory testing should be performed for all patients prior to gastrointestinal surgery. At minimum, the testing should consist of: Certain procedures or clinical conditions may require additional assessments. Occasionally, a laboratory test may be required on the day of surgery (e.g. serum potassium levels after extensive mechanical bowel preparation (MBP) or glucose levels for patients with severe diabetes mellitus). Recent data have indicated that an elevated preoperative level of brain natriuretic peptide is associated with increased cardiac morbidity after major surgery, but it remains to be seen whether this level will be routinely determined for patients with cardiac risk factors [4‐6].

Preoperative 12-channel electrocardiography (ECG) allows for screening of as-yet undetected cardiac disorders. It also serves as a control should perioperative cardiac complications occur. ECG should be performed for patients who:

Although it is not routine practice, many clinicians recommend that for patients with coronary artery disease who underwent high-risk surgery, an additional ECG should be obtained immediately after surgery as well as on days 1 and 2 postoperatively.

The sensitivity of conventional chest radiography to detect pathophysiologic conditions in asymptomatic patients is relatively low. However, X-ray images may always serve as a basis for comparison should perioperative complications occur. While there is currently no recommendation for preoperative chest radiography for patients with an American Society of Anesthesiologists score 1–2, regardless of the patient’s age, it is indicated for patients who:

Preoperative echocardiography should be performed on patients who:

Preoperative carotid Doppler ultrasonography should be performed on patients who:

The definition of being “high risk” for poor outcome after surgery is nebulous, as it is influenced by many variables that vary from patient to patient and from one surgical procedure to another [7]. The surgeon and anaesthesiologist need to jointly evaluate the potential perioperative risk for each patient and the intended procedure. Relevant clinical conditions that characterise high-risk surgical patients are listed in Table 1, and the top 10 clinical variables that influence 30-day and long-term mortality are highlighted in Table 2 (modified from Ackland and Edwards [7] and Khuri et al. [8]).

In an attempt to facilitate perioperative risk assessment, a variety of scoring systems have been developed that incorporate the patient’s age and comorbidities and the complexity of the surgical procedure. Well-known systems include the Physiological and Operative Severity Score for the Enumeration of Mortality and Morbidity [9] and the Estimation of Physiologic Ability and Surgical Stress score [10]. However, both systems indicate a general risk for complications and do not specify or pinpoint any specific complication. Accordingly, their implementation into routine clinical practice has proven to be difficult. Data from our University Medical Center suggest that the subjective opinion (“gut feeling”) of the surgeon is a good predictor of postoperative outcome, especially in nonemergency surgery [11].

More recently, a risk calculator for colorectal surgery has been developed by the National Surgical Quality Improvement Program registry of the American College of Surgeons. After a patient's variables are entered, the risk probabilities for adverse outcome are calculated [12]. However, only registry members can access the calculator, and it is not clear whether this US hospital-based tool is applicable to European institutions.

Overall, gastrointestinal surgery is associated with a medium cardiac risk [13‐16]. However, due to an ageing population, an increasing incidence of coronary artery disease and the increasing complexity of surgical procedures, postsurgical cardiac complications are now a leading cause of morbidity and mortality. Particularly cardiac insufficiency is emerging more and more as a risk factor for perioperative adverse outcome, even compared with cardiac ischemia [17]. Comorbidities that increase the cardiac risk for patients undergoing gastrointestinal surgery include:

Because both the assessment of these cardiac risk factors and their clinical interpretation are complicated, patients with diverse cardiac risk factors, acute symptoms of a cardiac disease or reduced physical fitness should be referred for consultation with an experienced cardiologist [13‐16].

Late postoperative pulmonary complications are the second-leading cause of morbidity and mortality after major surgery. For this reason, preoperative optimisation of the patient’s physical condition and medication is important. Clinical parameters that represent risk factors for pulmonary complications after gastrointestinal surgery are listed in Table 3 (modified from [18, 19]). Although most of these risk factors cannot be circumvented, they must be kept under consideration [20].

Beta-adrenergic blockers are frequently used in the perioperative management of patients with cardiac disease due to their favourable effect on the supply and demand ratio of myocardial oxygen. Although still under debate, it is currently recommended that all patients who are already receiving beta-adrenergic blockers continue them perioperatively [14, 21]. Abrupt discontinuation can cause unstable angina, tachyarrhythmia, myocardial infarction and sudden death. If a patient who is scheduled for elective gastrointestinal surgery requires a new prescription, it should be started at least 1 month before the procedure to allow for dose adjustment [14, 21].

Diuretics should not be used on the day of surgery because this may increase the risk of intraoperative hypovolaemia. However, it is strongly recommended that their intake be continued postoperatively, especially for patients who have heart failure.

The relevance of the oral anti-diabetic drug metformin for inducing lactic acidosis has been controversially discussed in the literature [22]. Regardless, it is recommended that its intake be stopped 48 h prior to the surgery.

Anti-platelet therapy (usually 100 mg of acetylsalicylic acid daily) is standard for most patients with coronary artery disease. The 2009 European Society of Cardiology guidelines suggest that to reduce the risk of stent thrombosis and myocardial infarction, patients with a coronary bare metal stent (BMS) or a drug-eluting stent (DES) should receive anti-platelet therapy with both acetylsalicylic acid and a thienopyridine derivative (i.e., clopidogrel or ticlopidine) for 1 month (BMS) or 12 months (DES) after stent placement [14].

For patients who currently receive anti-platelet therapy and are scheduled for gastrointestinal surgery, the following wait times until surgery are recommended:

For high-risk cardiac patients (i.e. patients with recent acute coronary syndrome, recurrent angina pectoris or recent surgical and conservative coronary intervention) who require major surgery that cannot be postponed, thienopyridine derivatives should be stopped 7–10 days before the surgery, whereas acetylsalicylic acid should be continued during the entire perioperative period [14, 23‐25]. This recommendation also applies to patients who require an epidural catheter.

l-Dihydroxyphenylalanine is the most frequently prescribed drug for Parkinson’s disease. Because of its relatively short half-life, it should be continued during the entire perioperative period, as interrupting the medication can result in a life-threatening complication known as neuroleptic malignant-like syndrome, which is associated with fever, confusion and elevated concentrations of muscle enzymes [26].

Because it creates cardiac stress, smoking increases the risk for perioperative complications, particularly for elderly patients with impaired cardiac function [41]. For this reason, systematic smoking intervention is highly recommended. To be of benefit, however, smoking cessation needs to occur several weeks prior to the surgery [42, 43].

Although the exact clinical definition of malnutrition has not been established, it is well documented that poorly nourished patients are more likely to develop postoperative complications than well-nourished patients [44‐46]. Accordingly, preoperative enteral and parenteral nutritional support has been recommended for malnourished patients [47, 48]. For the general patient population, there is no evidence for preventive nutritional support.

Obesity increases the rate of perioperative complications, increases morbidity and mortality after colectomy [49] and is a risk factor for postoperative intra-abdominal infection after gastrectomy [50]. In a recent study of obese patients who underwent Roux-en-Y bypass or laparoscopic adjustable gastric banding, 4.3% developed a major adverse outcome within 30 days [51]. For those patients, a history of deep-vein thrombosis or pulmonary embolus, a diagnosis of obstructive sleep apnoea and impaired functional status represented independent risk factors for perioperative complications.

Due to the increasing complexity of modern concepts on cancer treatment, an increasing number of patients are receiving preoperative chemotherapy, either alone or with irradiation. Clinicians should be aware of four common side effects of frequently used chemotherapeutic agents and of radiotherapy. First, the anthracycline epirubicin, which is used to treat locally advanced gastric cancer, increases the risk for cardiomyopathy and heart failure, particularly among elderly patients with pre-existing cardiac disease [52]. Consequently, cancer patients with cardiomyopathy who received preoperative chemotherapy with epirubicin should be considered for a routine echocardiography, in addition to conventional electrocardiography, prior to any surgical procedure [53]. Second, the pyrimidine analogue 5-fluorouracil, the most frequently used chemotherapeutic agent for gastrointestinal cancer, can induce coronary vasospams, myocardial ischemia and subsequent infarction [52]. Third, patients with gastrointestinal malignancy who receive the topoisomerase I inhibitor irinotecan are at risk for diarrhoea-induced malnutrition [52]. Fourth, radiotherapy is a cardiac risk factor for patients with cancer of the lower oesophagus or the gastroesophageal junction, and irradiation of rectal cancer can cause severe enteritis, malabsorption and diarrhoea [54, 55].

The goal of preoperative antimicrobial prophylaxis is to reduce the intraoperative bacterial load to a degree that can be controlled by the patient’s innate immune system. The importance of adequate antimicrobial prophylaxis in preventing surgical site infection (SSI) is illustrated by its inclusion in World Health Organisation’s surgical safety checklist and by the results of two trials of the use of such a list [56, 57]. Even so, there are many questions to be answered and improvements to be realised before the optimal effect of this intervention can be achieved.

For example, although the time to reach the plasma peak concentration is known for most antibiotic agents, very little is known about the time to achieve an adequate concentration on the skin or in other organs. In addition, the changes in haemodynamics caused by induction of anaesthesia and by the effects of any underlying disease might influence this timing. Intravenously applied beta-lactam antibiotics, which are commonly used as SSI prophylaxis, normally reach peak plasma concentration soon after their application, but their distribution to tissues depends on organ perfusion, plasma binding rate, molecular size and other factors. Studies by Stone and Classen have suggested that application should occur between 2 and 1 h before the skin incision [58, 59], and the World Health Organisation checklist advocates a 1-h interval (the commonly used 30-min interval has never been rigorously investigated). Other studies have demonstrated that extension beyond 24 h actually leads to the emergence of resistant bacteria and higher rates of SSI [60].

Other issues in optimising the use of preoperative prophylactic antibiotics include the choice of drug in relation to the site and type of surgery, adequate dosing, early identification of potential additional patient risk factors and local characteristics of bacterial resistance. The surgical site will determine whether both Gram-negative and Gram-positive bacteria should be covered. The frequently used cephalosporins, for example, lose their Gram-positive potency from generation to generation, whereas their Gram-negative potency increases (although this is not true for fifth-generation cephalosporins such as ceftobiprole, which recover their effectiveness against Gram-positive germs). Cefazolin is suitable if preventing infection of the skin and deeper soft tissue (i.e. fascial and muscular layers) is the main goal. Third-generation cephalosporins are appropriate for inhibiting infection of the abdominal cavity organs or space. Third-generation ceftriaxon has a high permeability into tissues and a high plasma protein binding rate, which lead to a long plasma half-life of up to 8 h. In contrast, second-generation cefotiam has a plasma half-life of 30 min and, therefore, is not adequate as a single-dose perioperative prophylaxis if the time to wound closure will exceed 1 h. In cases where the surgical intervention exceeds the plasma half-life of the chosen antibiotic drug, a second dose after 3 h may be considered [61].

The risk of developing SSI in developed countries is very low, around 1% to 3% for clean surgery, but there is no question that this rate can be significantly reduced [62]. The drugs used should be defined in advance for each intervention, including alternatives should the patient have any contraindication against the first-choice antibiotics. Although it can be applied in the operating room, the drug is ideally applied beforehand. Application should be started immediately after intravenous access has been established (between 2 and 0.5 h before skin incision, as most antibiotics should reach a relevant tissue level in that time frame). If the institution carries a holding area, application might take place there; the “time out” of the surgical checklist should ensure the antibiotic was given within the appropriate time frame.

Any application of antimicrobial drugs after closure of the surgical wound should not be considered perioperative prophylaxis, as this categorisation thwarts the rationale behind the approach. Application after wound closure actually increases the risk of SSI fivefold and promotes the development of resistant bacterial strains [59, 63]. Even correctly performed “single-shot” perioperative antibiotic prophylaxis can induce severe Clostridium difficile infections and diarrhoea with highly virulent strains [64, 65]. Besides their negative and sometimes life-threatening consequences, SSIs significantly increase the length of the hospital stay and the associated costs. Because studies providing a high level of evidence for the how and when of perioperative antimicrobial prophylaxis are lacking, much research is still needed on the optimal and adequate use of well-established measures, including perioperative antimicrobial therapy, to reduce the incidence of SSI. Until then, the practical approach should be oriented along the theoretical rationale behind this intervention.

There is an ongoing discussion about whether antibiotic prophylaxis should cover resistant bacterial strains, such as methicillin-resistant Staphylococcus aureus (MRSA), if the patient had been colonised before the surgery [66]. The data do not support the general use of perioperative prophylaxis with agents active against MRSA if the patient is colonised by resistant bacteria. Two small studies of hospitals with a high prevalence of MRSA demonstrated conflicting results for a cohort of cardiac surgery and neurosurgery patients [67, 68]. The study in cardiac surgery patients did not demonstrate a difference in the rate of surgical wound infections when comparing vancomycin with cefazolin, whereas the same approach significantly reduced shunt infections and mortality when neurosurgical patients were studied. However, for patients at high risk of SSI, who should be identified in advance of surgery, the extension of antibiotic prophylaxis to agents against MRSA and other resistant bacterial strains might be considered.

At our University Medical Center, vancomycin is not the drug of choice for such cases because it has a high molecular weight and penetrates poorly into tissues. Clindamycin, rifampicin or fosfomycin might be used as long as the MRSA strain is sensitive to these agents; linezolid and daptomycin are additional options.

The determination of the most effective use of preoperative and perioperative antimicrobial therapy for a patient requires efforts by a multidisciplinary team, including pivotally the anaesthesiologist and the surgeon but also the microbiologist and the nursing staff. These efforts will often require the use of a standard protocol, quality management and change. The change includes altered behaviour by surgeons and the other team members involved in the patient’s care pathway through the surgical procedure and recovery [69, 70]. Quality improvement measures from industry have been successfully applied to hospital settings to achieve adequate, if not optimal, prophylactic antibiotic use [71].

Out of 10,000 episodes of anaesthesia, documented severe aspiration occurs in 1–5 cases; aspiration might also occur unobserved during all phases of anaesthesia, from induction to recovery [72]. In an otherwise unstressed individual, this unobserved aspiration (“microaspiration”) might cause no serious complications, but because the surgical intervention challenges the immune system, the lungs may be more susceptible to exposure to minor amounts of bacteria from the oral cavity. Abdominal surgery clearly carries a higher risk of postoperative pulmonary complications than other types of surgery do; measures to reduce the rate of these complications will clearly influence patient outcome [73, 74].

There is an ongoing debate on the extent to which oral hygiene measures, types of endotracheal tubing cuffs, cuff pressure control and continuous supraglottic suctioning might ameliorate postoperative pulmonary complications [75]. In a study of 86 patients who underwent oesophagectomy, the incidence of postoperative pneumonia was significantly lower among patients who brushed their teeth five times a day than among patients who used standard oral care as usual [76]. With a large cohort of post-cardiac surgery patients, Bouza et al. demonstrated the cost-effectiveness of supraglottic suctioning; however, the effectiveness of this measure as an intraoperative intervention and during other types of surgery has still to be proven [77]. In 2000, the Acute Respiratory Distress Syndrome Clinical Network trial demonstrated for the first time a lower mortality in a large cohort of patients with acute lung injury or acute respiratory distress when a protective ventilatory strategy was applied using small tidal volumes and predefined positive end-expiratory pressure settings [78]. Other recent studies clearly hint that the use of protective ventilatory settings during anaesthesia in the operating room—especially during large abdominal surgery—reduce the inflammatory reaction of the human body and very likely improve patient outcome [79‐81].

Induction of anaesthesia leads to atelectasis formation, predominately in the caudal-dependent parts of the lungs. Prevention and reversal of atelectasis increases functional residual capacity and improves gas exchange in the postoperative period [82, 83]. Measures to reduce atelectasis might, therefore, reduce postoperative pulmonary complications. The usefulness of positive end-expiratory pressure intraoperatively is controversial, but Squadrone et al. have demonstrated the positive effects of applying continuous positive pressure in the recovery room after major elective abdominal surgery [84, 85].

In conclusion, oral hygiene measures and, at least in high-risk patients, the use of special endotracheal tubes clearly influence the outcome of patients who undergo major abdominal surgery. In addition, protective intraoperative ventilatory settings should be used because they influence the systemic inflammatory response to surgery and ventilation. Finally, there is good rationale to prevent or reverse atelectasis, but the number of studies supporting this approach on an evidence-based level is limited.

The fact that anaesthetic agents modulate the response of an organism to surgical stress and hence can influence patient outcome after major abdominal surgery represents a double-edged sword. These drugs can either enhance or lower the ability of the immune system to sufficiently react to sources of infection during the postoperative period, and they can increase or decrease mesenteric perfusion [86‐88]. The effect on mesenteric blood flow does not depend on whether balanced general anaesthesia (i.e. opioids combined with volatile anaesthetic agents), pure inhaled anaesthesia or pure intravenous anaesthesia is used, but on the degree to which the applied technique influences haemodynamics in general.

The combination of general anaesthesia and thoracic epidural anaesthesia (TEA) has become the technique of choice at many institutions for major abdominal surgery. TEA improves mesenteric blood flow, increases oxygen supply to the abdominal cavity and allows sufficient pain control after surgery (the latter represents one of the cornerstones of the fast-track surgery concept) [89‐91]. Very recent studies have suggested that for some types of cancer TEA might also reduce the rate of recurrence after surgical resection. The possibility of reducing tumour recurrence makes the combination of general anaesthesia and TEA even more appealing, even if some contraindications for its use might exist [92‐95].

Numerous studies have demonstrated the negative effects of intra-hospital hyperglycaemic phases on the rate of postoperative complications (including wound healing), nosocomial infections, length of hospital stay and mortality [96‐100]. An influential study by Van de Berghe introduced the concept of intensive glucose control for critical care patients, which has been promoted and quickly applied to a more general population of intensive care patients [101]. However, studies of this approach that were performed with patients with septicaemia and in a general population of intensive care patients raised serious concerns about subsequent severe hypoglycaemia and about the effects on outcome that were originally reported [102, 103].

In consequence, this concept was modified toward less extreme blood glucose levels, but glucose control per se can still be regarded as a golden standard for reducing perioperative complications. The optimal glucose level in the perioperative setting has not yet been prospectively investigated. There is some evidence that in daily practice the blood glucose level should be kept between 110 and 180 mg/dl to simultaneously reduce the incidence of hyperglycaemia-related complications, such as compromise of the immune system, and to avoid hypoglycaemia [99]. Intensified glucose control requires well-trained, motivated health care staff. Accurate measurement of the actual glucose level at the bedside, or the actual value being available without delay, is also necessary.

Traditionally large amounts of fluids, either in the form of crystalloid solutions or as a combination of crystalloids and colloids, have been applied during major abdominal surgery [41]. However, studies on the positive effects of restrictive fluid management on postoperative complications and the concept of fast-track surgery have challenged this traditional approach and have opened up discussion on how liberal or restrictive perioperative volume therapy should be [104, 105]. Today, a more differentiated and individualised approach is suggested.

Studies have shown that in healthy individuals, restrictive (versus moderately liberal) volume replacement does not provide any beneficial effect on patient outcome. A recent review of seven prospective studies on restrictive (998–2,740 ml) and liberal (2,750–5,388 ml) volume replacement demonstrated inconsistent results, and the authors stated that “evidence-based guidelines for optimal procedure-specific peri-operative fixed-volume regimens cannot be formulated” [105]. However, there is evidence that high-risk surgical patients might profit from early intervention and goal-directed therapy in the perioperative phase, including the optimisation of volume status, to achieve the best rate of oxygen delivery to the cells of the body and in particular to cells involved in the mechanical and biological stress caused by the surgical intervention [106‐108]. Because the rate of oxygen delivery is determined by cardiac output, volume status, contractility, peripheral resistance and oxygen content, it becomes evident that its optimisation requires extended monitoring and a multi-factorial approach.

In the monitoring of patients who receive volume therapy, there has been a clear paradigmatic shift from pressure- to volume-targeted parameters and the use of monitoring devices that are less invasive than the traditional pulmonary artery catheter. Early intervention might even include preoperative transfer to an intensive care unit (ICU) and transport of an already “optimised” patient to the operating room [109‐111].

There is also ongoing debate on the best solution to use for volume replacement and volume optimisation. Studies of the use of colloidal volume replacement solutions such as hydroxyethylstarch and gelatin in intensive care medicine demonstrated that they may negatively affect patient outcome, especially renal function [102, 112‐114]. However, those studies were performed in the intensive care setting, mostly with patients who had septicaemia and extended use of these substances. Even so, colloidal volume replacement acts faster than cristalloids in restoring plasma volume, a characteristic that is particularly important if a large volume is lost during the course of surgery [115]. The use of colloidal volume replacement may be contraindicated for patients with pre-existing kidney problems: one study clearly demonstrated that the risk of a negative impact on kidney function increases with the preoperative renal Sequential Organ Failure Assessment score and, therefore, with the pre-existing degree of renal impairment.

In summary, a moderately restrictive volume replacement strategy for uncompromised patients seems adequate; extreme volume loading definitely should be avoided. High-risk surgical patients should be identified early, monitored for plasma volume and ideally already be optimised with regard to oxygen delivery when they arrive at the operating room. Colloids should be made available for volume replacement during surgery because they act faster than cristalloids in restoring plasma volume and thus in ensuring oxygen supply. The amount given should not exceed 20 ml/kg body weight and day. In cases of pre-existing renal impairment, the use of artificial colloids should be used more cautiously.

Numerous prospective randomised studies have demonstrated the negative effects of perioperative hypothermia on the duration of muscle relaxants, intraoperative blood loss, transfusion requirements, shivering, discomfort, postanaesthetic recovery, morbid cardiac events, surgical wound infections and duration of hospitalisation. Adequate control of body temperature, with warm forced-air blankets or warm fluids, is thus critical for patient outcome [116].

The efficacy of a forced-air warming system is determined mainly by the design of the blanket. During abdominal surgery, small blankets (i.e. upper body blankets and paediatric blankets) are usually used, which reduces the requirement on the airflow from the power unit. Nevertheless, the largest blanket that is feasible should be used. Forced-air warming blankets alone might not be sufficient to keep a patient normothermic throughout surgery, in cases in which large amounts of fluids are infused during the procedure, fluid warming should be employed as well.

Another approach is to pre-warm the patient. In a recently published pilot study, warming of patients before they reach the operating room reduced the postoperative degree of hypothermia, as pre-warming effectively reduces the redistribution of heat after induction of anaesthesia [117]. Effective pre-warming requires 30–60 min in a holding area and also the willingness of the care team to master the organizational challenges of this approach.