The choice of incision for laparotomy depends on the area that needs to be exposed, the elective or emergency nature of the operation and the personal preference. The transverse incision seems to be associated with fewer early postoperative complications (mainly pain and pulmonary morbidity) and lower incidence of late incisional hernia 55. However, some authors report abdominal or neural dysfunction after transverse access because of nerve, muscle or vessel interruption 56. A midline incision is still the incision of choice in conditions that require rapid intra-abdominal entry or where the preoperative diagnosis is uncertain, as it is quicker and can easily be extended 57.

Abdominal access in endoscopic surgery carries a finite risk of visceral injury. Bleeding, intestinal perforation, vascular injury, intraperitoneal adhesions and subcutaneous emphysema are the main potential complications. To date there are four techniques used to create a pneumoperitoneum: blind Veress needle, direct trocar insertion, optical trocar insertion and open laparoscopy. The first two entry mechanisms are blind. The described overall complication rates in these techniques are below 1%. Some studies revealed a higher rate of visceral injuries in the open-entry technique. In a survey 106 gynecologists (57%) used only the closed-entry technique. This group reported 31 complications (0.1%) in 31,532 procedures. Even in the case of patients who were at risk for entry-related complications (previous laparotomy, obesity), pneumoperitoneum was established by the closed-entry technique. The remaining 81 gynecologists used both entry techniques. However, the open-entry technique was used on special indications and in only 2.0% of cases (range: 1-20%). These special indications were suspected adhesions or previous laparotomy (90%) and obese (7%) or very thin patients (3%). These 81 gynecologists reported 20,027 closed-entry procedures and 579 open-entry procedures and complication rates of 0.12% and 1.38%, respectively (P < 001). Significantly more visceral lesions were found (P < .001) at open-entry technique 58. Therefore there is no evidence to prefer one technique in laparoscopic access.

To select the kind of access a recent study gives some useful recommendations: (1) use left upper quadrant entry in patients with suspected adhesions or umbilical hernia (2) limited movement of the inserted Veress needle (3) an intraperitoneal pressure less than 10 mmHg is a reliable indicator of correct placement of the Veress needle (4) the angle of the Veress needle should be at entry 45 degrees in non-obese patients and 90 degrees in obese patients. Direct insertion of the trocar without prior pneumoperitoneum is associated with less insufflation-related complications such as gas embolism, is faster to perform and is a safe alternative. The visual entry cannula system may provide advantages over the traditional techniques but has to be fully explored in the future 59.

The seriousness of vascular injury is high in comparison to visceral injuries during the abdominal access. These cases are rare and no evidence based recommendations of treatment can be given. Injuries to the main vascular structures need an immediate conversion and surgical repair. Small bowel injuries can be treated laparoscopically. Severe lesions sometimes require segment resections and conversion to open surgery. Injuries of the liver or spleen are manageable with laparoscopic devices. If severe bleeding continues a pre-emptive laparotomy is recommended (see also next section).

There is limited data regarding iatrogenic injuries in colorectal surgery. The main fears of the surgeon are vessel injury, damage to the spleen during colorectal surgery (incidence of 0.006%) 60, or intestinal perforation and uretric injuries (incidence < 0.01%). Injuries to the abdominal or pelvic veins occur mainly in patients undergoing oncologic resections, and those with difficult anatomic exposure, owing to previous operation, recurrent tumor or radiation therapy. Most of the injuries can be repaired by primary suture or end-to-end anastomosis. Few injuries need interposition grafts, patch venoplasty or venous ligation. Therefore a vascular surgeon should be available in hospitals where cancer resections are frequently performed 61.

Iatrogenic perforation of the bowel occurs either during adhesiolysis or inadvertently due to thermic lesion, the latter are often not recognized during the operation. The surgeon should prefer primary repair or resection with anastomosis. In laparoscopic cases the bowel injury should be sutured immediately as it might be difficult to localise later 6263.

In general, the incidence of iatrogenic splenic injury is underestimated because of poor documentation. Splenectomy is considered a poor prognostic factor 5564. Splenic injury results in increased blood loss, longer hospital stay and higher mortality and infection rates. Splenic injury can be reduced by achieving good exposure, avoiding undue traction and careful division of splenic ligaments and adhesions. If the spleen is injured preservation is desirable and often feasible 65.

To date, data available concerning studies which compare the safety of surgical devices are limited. Conventional monopolar electro-surgery has several short-comings in laparoscopic surgery including the risk of thermal injury, difficult hemostasis and disturbing smoke production, making the use of additional tools like bipolar graspers, sutures or clips necessary. To overcome these problems and to reduce instrument changes, number of trocars and operation time, several multifunctional tools have been developed. The most popular devices are electro thermal bipolar vessel sealers and ultrasonically coagulating shears. In a recent prospective randomized study we could show that bipolar vessel sealers and ultrasonic coagulation shears shorten dissection time in laparoscopic left-sided colectomy and are cost-effective compared to monopolar electro surgery. Other studies showed less operative blood loss and a decrease in operating time when the ultrasonic dissection device were used. For now it is still the preference of the surgeon as to which device is used 6667.

Preoperative anemia and intraoperative blood transfusion are independent risk factors for intra- and postoperative complications in colorectal surgery 5. In a case-matched study of 147 patients undergoing colectomy using either an open or laparoscopic approach the open colectomy group required significantly more units of blood (P = .003) to maintain similar hemoglobin levels after surgery. Estimated blood loss (P < .001) and the number of patients who received transfusions on the day of surgery (P = .002), during the first 48 hours after surgery (P = .005), and during the entire hospital stay (P = .003) were significantly higher in the open colectomy group 68.

To prevent intraoperative blood loss and postoperative complications some laparoscopic surgeons prefer ultrasonic dissection with produces significantly less blood loss and thereby iron supplementation in preoperative anemic patients two weeks prior to surgery 69.

One of the initial arguments to discredit laparoscopy was the index of conversion, which was interpreted as operative failure. However, today conversion is no longer considered a failure, but as result of good clinical judgment. The average conversion rate in laparoscopic colorectal surgery is approximately 10%. Independent predictive risk factors for conversion are BMI (odds ratio of 2.1 per 10 units increase in BMI), ASA grade (I 2.3%, II 9%, II-IV 13.8%), type of resection (low rectal resection 18.4% vs. left colorectal resection 15.3% vs. 8.1% in right colon resections), intraoperative abscess, surgeons seniority 70. The Laparoscopic Colorectal Surgery Study Group showed in a multicenter study with 1'658 patients a conversion rate of only 5.2% (n = 86). Converted patients were significantly heavier (body mass index 26.5 vs. 24.9) and rectal resections were converted more frequently (20.9 vs. 13 percent) 71.

Specific indications for conversion were technical problems, adhesions, bleeding, abscess, fistula, inflammatory mass and bowel perforation. Also prior abdominal surgery increased the conversion rate up to 41% in a recent study 1472. The effect of conversion on morbidity and mortality is discussed controversially in the literature. Recent studies describe similar outcome after conversion compared to the open access 7173. However, large randomized trials clearly demonstrate increased morbidity and loss of short-tem benefits in converted patients 4974. Obesity is associated with a higher conversion rate but the outcome of converted patients seems to be similar to the open cases 26. In another study obesitiy was not a risk factor for conversion 75.

The value of prophylactic drainage in colorectal surgery has been studied extensively. Currently available data from randomized controlled trials point out that a routine prophylactic drainage provides no benefit after uncomplicated major colon and rectal surgery 76. On the contrary, a no drain policy was associated with less wound infections and a fewer anastomotic leaks. These studies underscore the low sensitivity of drains in detecting leakage and bleeding, which questions the putative warning function of a prophylactic drain. In addition, neither acute/simple nor gangrenous or perforated appendicitis benefit from a prophylactic drainage. In summary, there is sufficient evidence showing that routine drainage after colorectal anastomoses does not prevent leaks or other complications 7778.

A stoma may be a temporary solution when there is a dysfunction of a colostomy or ileostomy however the advantages are still under debate. One group argues that a protective stoma is only indicated in low rectal resections in patients with significant comorbidities, neoadjuvant radiochemotherapy and feculent peritonitis 79. If adjuvant radiochemotherapy is considered postoperatively in patients with a colorectal carcinoma the closure of the temporary loop ileostomy should be performed before the chemotherapy to minimize complications 80. Other groups do not recommend a stoma at all or only when the colorectal anastomosis is in the lower third of the rectum 8182. In emergency situations with peritonitis and perforation of the left colon primary anastomosis and protective ileostomy should be performed rather than a Hartman's procedure 8384.

The duration of the operation is influenced by many factors such as; surgical technique (open or laparoscopic), intraoperative complications, prior abdominal surgery, surgeon's experience and the operating team. Many studies showed that prolonged operating time correlated with higher intra- and postoperative complications. In a series of 541 colorectal anastomoses between 1999 and 2004 at a single colorectal unit, univariate analysis showed that a prolonged operating time had an odds ratio of 2.8 for developing an anastomotic leakage 13.

On the other hand, nearly all prospective randomized studies comparing a laparoscopic with the open approach reported longer operations in the laparoscopic group, but surprisingly without an increase of intra- and/or postoperative complications and with similar morbidity and mortality rates 8586. Probably, the negative effect of the prolonged operating time in laparoscopic surgery is overrun by advantages such as decreases in-hospital stay, wound infection, postoperative ileus and postoperative pain. However, there is a lack of well designed studies evaluating the influence of the operating time on postoperative outcome as a primary endpoint.

The decreased length of hospital stay due to more cost effective outpatient procedures necessitates good postoperative pain management. It has been demonstrated that well managed pain control supports respiratory function and lowers the risk of complications 87. In colorectal surgery the major modalities of postoperative pain control are patient-controlled anaesthesia, opioids, nonsteriodal anti-inflammatory drugs, and epidural anaesthesia.

Some studies show that pain control, patient satisfaction and bowel function are improved after abdominal surgery under epidural analgesia 88. Carli et al. showed in a prospective randomized study that epidural anaesthesia significantly shortened the duration of postoperative ileus and improved postoperative pain control. Postoperative complication rates and length of hospital stay was not shown to be improved in this study 89. Another study demonstrated that continuous epidural analgesia is superior to patient controlled opioid analgesia in relieving postoperative pain for up to 72 hours, but was associated with a higher incidence of pruritus 90. Epidural anesthesia has a low complication rate, however, if complications occur they are mostly severe. The risk of a symptomatic spinal mass lesion after patient-controlled epidural analgesia was 1:2857 (0.04%), including epidural haematoma (0.02%; 1:4741) or epidural abscess (0.014%; 1:7142). Another recent study demonstrated that epidural analgesia reduced the need for prolonged ventilation or reintubation, improved lung function, increased blood oxygenation, reduced risk for pneumonia, to the contrary increases the risk of hypotension, urinary retention, and pruritus. Technical failures occurred in 7% 9192. Despite the advantages of epidural anesthesia its use alone cannot prevent postoperative morbidity and mortality. It is therefore necessary to address its use in the context of multimodal intervention.

The resumption of a diet is critical to the recovery. Before discharge patients should demonstrate return of intestinal tract function based on oral food intake, flatus and/or bowel movements. Traditionally, patients received a nasogastric tube decompression and were set on a "nil per os diet" postoperatively. Different trials failed to show that a nasogastric tube has any postoperative benefits for the patient, causing most surgeons to abandon its routine use 93. There is much variability in regards to restarting enteral nutrition in patients undergoing colorectal surgery. Several trials demonstrated that the majority of patients tolerated oral intake in the immediate postoperative period, regardless of the presence or absence of traditional markers of normal gastrointestinal function. In a metaanalysis of 837 patients it was seen that reduced postoperative infections, reduced anastomotic complications and shorter length of stay was shown in patients who received immediate postoperative normal diet compared to patients who were fasted until gastrointestinal functions were resumed 94. In a recently published analysis the advantages of early enteral feeding were not significant but showed a trend towards fewer postoperative complications 95. Another metaanalysis of 13 trials (1'173 patients) came to the conclusion that there is no obvious advantage in keeping patients 'nil per os' following gastrointestinal surgery. Early enteral nutrition was associated with reduced mortality. This review supports the notion that early commencement of enteral feeding may be of benefit compared to the nothing by mouth policy 96.

In the last few years Kehlet et al. favored a multi-modal rehabilitation with an emphasis on preoperative information, reduction of surgical stress responses, optimized dynamic pain relief with continuous epidural analgesia, early mobilization and oral nutrition (fast-track surgery) 97. Current results from fast-track colonic surgery suggest that postoperative pulmonary, cardiovascular, and muscle function are improved and body composition preserved as well as a normal oral intake of energy and protein can be achieved. Consequently, hospital stay is reduced to about 2-4 days, with decreased fatigue and need for sleep in the convalescence period. Despite a higher risk for readmissions, overall costs and morbidity seem to be reduced 979899. A recent randomized study by our group compared the 30-day complication rate of patients who underwent a fast track protocol or standard care after open colonic surgery. The fast-track protocol significantly decreased the number of complications (16 of 76 in the fast-track group vs. 37 of 75 in the standard care group; P = .0014), resulting in shorter hospital stay (median, 5 days; range, 2-30 vs. 9 days, respectively; range, 6-30; P < .0001). Fluid restriction and effective epidural analgesia were the key factors that determine outcome in the fast-track program 100.

In summary, there is a growing body of evidence that early enteral nutrition improves outcome and reduces postoperative complications. Despite proven advantages of fast track surgery the implementation of a standardized and multidisciplinary care is difficult since resistance is still enormous.

Colorectal operations are, at best, clean-contaminated procedures, and at times there is contamination of both the peritoneal cavity and the surfaces of the surgical wound. In addition, the diseases of the large bowel that require surgery tend to afflict elderly patients. Collectively, the combination of an unclean environment, major surgery and debilitated patients creates a situation that is associated with a very high incidence of wound infection. In open colorectal surgery the incidence of SSI varies from 2-25% and is associated with BMI ≥ 30, creation/revision/reversal of an ostomy, perioperative transfusion, male gender, ASA Score ≥ III and wound contamination 102103. The incisional SSI rates in colon (n = 339) and rectal (n = 217) resections were 9.4% and 18.0%, respectively (P = 0.0033). Risk factors for SSI in colon surgery were ostomy closure (OR = 7.3) and lack of oral antibiotics (OR = 3.3), while in rectal surgery, risk factors were preoperative steroids (OR = 3.7), preoperative radiation (OR = 2.8), and ostomy creation (OR = 4.9) 104. Some studies showed that perioperative oxygen supply and preoperative immunonutrition decreased SSI significantly 105106. It is widely accepted that a laparoscopic approach lowers the rate of SSI 36107. As for laparoscopic appendectomies 108, most surgeons use plastic wound protectors during specimen removal after laparoscopic resection. This certainly facilitates extraction through a small incision, but there are no randomized controlled trials demonstrating a reduction in wound infection. The role of antibiotic prophylaxis in preventing postoperative complications in colorectal surgery is well established through many studies. However, there is still a debate about the duration of the antibiotic treatment and the kind of antibiotic which should be used. In summary, most studies favour one to three intravenous doses of a second generation cephalosporine with or without metronidazole with the first dose being administered before skin incision 109110.

Anastomotic leakage is the most serious complication specific to intestinal surgery and ranges from 2.9% to as high as 15.3%. At least one third of the mortality after colorectal surgery is attributed to leaks. Within this context, knowledge of factors influencing anastomotic healing appear even more important 81111. However, there is lack of a clear definition for what constitutes an anastomotic leak (radiological proven, clinically relevant, with or without abscess).

In general, the leakage rate for intraperitoneal anastomoses is significantly lower than for extraperitoneal anastomoses. Anterior rectal resections have the highest leakage rate of up to 24% 112113. The main risk factors for anastomotic leakage using univariate analysis were male gender (OR = 3.5), previous abdominal surgery (OR = 2.4), Crohn's disease (OR = 3.3), rectal cancer < or = 12 cm from the anal verge (OR = 5.4) and prolonged operating time (P = 0.05 as a continuous variable and P = 0.01 when prolonged operative time was >120 min). Male gender, a history of previous abdominal surgery and the presence of a low cancer remained significant after multivariate analysis 13.

Another multivariate analysis showed that American Society of Anaesthesiologists Grade III to V (P = 0.04; odds ratio, 5.6; 95 percent confidence interval, 1.6-15.3) and emergency operation (P = 0.03; odds ratio, 4.6; 95 percent confidence interval, 1.9-9.8) were independent factors associated with anastomotic leakage. The risk of anastomotic leakage was 8.1% (odds ratio, 10.5; 95% confidence interval, 2.7-26.8) if both factors were present 114.

Most studies comparing high and low anterior resections have shown that the level of anastomosis is the most important predictive factor for leakage. The high-risk level for leakage varies between anastomoses from <10 to <5 cm from the anal verge depending on the cited study 115116.

There seems to be no significant difference in leakage when comparing a handsewn and a stapled technique regardless of the level of anastomosis 117. Intraoperative problems and postoperative strictures seem to be more frequent in stapled anastomosis 118. However, in a recent Cochrane review ileocolic stapler anastomoses were associated with fewer leaks than handsewn anastomoses 119.

The available data comparing the anastomotic leakage rate in laparoscopic or open operated patients showed no difference regardless of the level of the anastomosis 120. In cancer patients anastomotic leakage (regardless of open or laparoscopic technique) is associated with poor survival and a higher recurrence rate after curative resection 121122.

Because of the severity of the complications associated with an anastomotic leak, it is imperative to identify the problem and act as early as possible. Most groups base the diagnosis on clinical symptomatic leakage, manifested as gas, purulent or fecal discharge from the drain, purulent discharge from the rectum, pelvic abscess or peritonitis. It is usually necessary to obtain objective tests of anastomotic integrity because of the non-specific clinical signs. Water soluble enemas or CT scans are widely used for diagnosis of anastomotic leak. Interestingly, in two recent studies anastomotic leaks were more often diagnosed late in the postoperative period and more often after hospital discharge, or 12 days postoperatively 123124.

Anastomotic leaks may be divided into those which are clinically significant and those which are not. Subclinical leaks are more benign in their natural history compared with clinical leaks although quality of life and bowel function does not differ in these groups 125. In pelvic abscess formation after colorectal surgery CT scan-guided percutaneous drainage should be performed in hemodynamically stable non-septic patients and has a success rate of up to 80% 126127. With signs of free anastomotic leckage in the abdominal cavity by CT scan the indication for surgery is mostly given. Despite the good results with conservative therapy (including antibiotics), the indication for surgical repair of anastomotic leakage should be made as early as possible to improve patient outcome. Re-laparoscopy and lavage after laparoscopic operation is feasible and safe and has less postoperative complications than an open re-intervention 128.

In general postoperative bleeding after colorectal procedures is a rare complication. The risk depends on the performed surgical procedure, the co-morbidities of the patient and in individual cases on an impaired clotting system. In the initially postoperative phase abnormal heart rate and low blood pressure should be reported and interpreted by the surgeon. Haemoglobin and hematocrit measurements can help to determine a blood loss.

Postoperative ileus has long been considered an inevitable consequence of gastrointestinal surgery. It prolongs hospital stay, increases morbidity, and adds to treatment costs. The pathophysiology of postoperative ileus is multifactorial. The operating time and intraoperative blood loss are independent risk factors for a postoperative ileus 129.

Postoperative ileus can develop after all types of surgery including extraperitoneal surgery. A variety of treatment options have been reported. However, it is difficult to compare these studies because of of the different anesthesia protocols used and patient comorbidities differed significantly.

Paralytic postoperative ileus is usually treated with a combination of different approaches. These include limitation of narcotic use by substituting alternative medications such as nonsteroidals and the placement of a thoracic epidural with local anesthetic. The selective use of nasogastric decompression and correction of electrolyte imbalances also are important factors to consider.