Evidence-Based Decision Making in Colon and Rectal Surgery

Enhanced Recovery Pathways

During the past decade, there has been much interest in postoperative recovery pathways designed to streamline and codify postoperative care following a variety of procedures. Although these protocols differ from hospital to hospital, there are basic elements that are included in most enhanced recovery pathways (ERPs) ( Table 181.1 ). The most common elements include preoperative counseling, avoidance of bowel preparation (see discussion later), no preoperative fasting, opioid-sparing analgesia and midthoracic epidurals, antibiotic prophylaxis, short incisions, no nasogastric tubes, normothermia, operative and postoperative fluid restrictions, no abdominal drains, oral diet at will, and early mobilization.

An early review by Wind et al., published in 2006, included six studies, three randomized controlled trials (RCTs), and three single-arm controlled clinical trials, published between 1998 and 2005. These were single-institution studies, and the number of ERP elements included ranged from 4 to 12, although all studies included early mobilization and diet. In five of six studies hospital stay was significantly shorter in the ERP patients, and in pooled analysis the ERP patients had a hospital stay almost 2 days shorter than patients in a traditional pathway (TP). There was no difference seen in the rate of readmissions. One study reported significantly lower morbidity in the ERP group, especially cardiovascular and pulmonary complications. In pooled analysis, this trend was also observed. There was no difference seen in rates of anastomotic leakage or mortality. Postoperative ileus (POI), measured by time to first bowel movement (BM) and tolerance of a solid diet, was reduced in the ERP group. There were mixed results regarding the outcomes of pain and fatigue, with some studies reporting no difference between ERP and TP groups, whereas others reported increased pain and fatigue in the TP group compared with the ERP group. These authors concluded that ERP programs result in improved recovery after surgery, with a reduction in morbidity rates and hospital stay. These findings were confirmed by a review published in 2009, by Gouvas et al., which evaluated 11 studies—four RCTs and seven controlled clinical trials—comparing ERP with TP. These authors conclude that ERPs contribute to a quicker recovery of patients after colorectal surgery and result in lower morbidity and shorter hospital stays.

Two more recent meta-analyses have further examined the impact of ERPs in colorectal surgery. In 2013 Zhuang et al. analyzed 13 studies (1910 patients) comparing ERP with TP. The mean number of enhanced recovery after surgery (ERAS) elements incorporated in each study was 11. The ERPs were associated with significantly decreased length of primary stay (−2.4 days; P < .001), total days in hospital (including readmission, −2.39 days; P < .001), and overall complications (relative risk [RR] = 0.68; P = .0006). There were no differences noted in readmission rates, surgery-specific complications, or mortality.

In 2014 Greco et al. performed a meta-analysis of 16 RCTs that included 2376 patients. In 11 of the 16 studies, at least 10 ERP elements were included in the ERPs; the most common elements included early postoperative feeding and mobilization, no postoperative nasogastric tube, epidural analgesia, and no preoperative fasting. Their analysis demonstrated a reduction in overall morbidity (RR = 0.60, 95% confidence interval [CI], 0.46 to 0.70) and length of stay (−2.28 days; 95% CI, −3.09 to −1.47 days) associated with ERP.

Although the individual elements differ among studies, the existing evidence is robust that a codified ERP can reduce length of stay and morbidity following colorectal surgery. Interestingly, many of the early studies in ERP were performed when open surgery was more common. The benefit associated with ERPs has been questioned in the setting of laparoscopic procedures, which are becoming increasingly common. Several studies have addressed this specific question. In 2011, Vlug et al. randomized 427 patients into four treatment arms—open colectomy with TP, open colectomy with ERP, laparoscopic colectomy with TP, and laparoscopic colectomy with ERP. The shortest length of primary hospital stay (median, 5 days) was noted in the laparoscopic/ERP group. In the laparoscopic/TP group, median length of stay was 6 days ( P < .001). A similar and significant difference was noted for total hospital stay (including readmission days). These authors concluded that optimal treatment for colorectal patients is laparoscopy in conjunction with ERP.

In 2012 Haverkamp et al. compared ERP and TP in 186 patients undergoing only laparoscopic colectomies. The median length of stay in the ERP cohort was 4 days compared with 6 days for the TP patients ( P = .007). Return to bowel function was noted 1 day earlier in the ERP group (2 vs. 3 days; P < .001). No differences were noted in postoperative complications, readmission, or mortality. Again, these authors conclude that ERPs are beneficial even in the setting of laparoscopic approaches to resection.

In 2014 Kennedy et al. reported the results of the EnRol (ENhanced Recovery Open versus Laparoscopic) trial, an RCT of 204 patients randomized to either open surgery or laparoscopic resection within an ERP. There was no difference in the primary outcome, physical fatigue at 1-month postoperatively, between the two groups, nor was there any difference in complications or other patient-reported outcomes. The total hospital stay was significantly shorter in the laparoscopy cohort (median, 5 days vs. 7 days; P = .033). Based on these results, the authors conclude that, within an ERP, laparoscopy can significantly reduce length of hospital stay.

Finally, two meta-analyses published in 2015 attempted to clarify overlapping benefits of laparoscopy and ERP. Zhuang et al. analyzed five RCTs, including 598 patients, to look at the benefit of laparoscopy when all patients are enrolled in an ERP. The authors noted that the overall quality of existing evidence was low to moderate, with several of the included trials using suboptimal ERPs. They concluded that total hospital stay following laparoscopic resection in the setting of an ERP was reduced compared with open resection but that more robust evidence is needed to truly prove that laparoscopy provides other benefits in the setting of optimal ERPs.

Spanjersberg et al., analyzed three RCTs and six controlled clinical trials in an attempt to answer two questions: (1) does laparoscopy offer benefit within an ERP, and (2) does ERP offer an advantage when all patients get laparoscopic resection. In the laparoscopic patients, the length of stay was shorter in patients enrolled in an ERP (−2.3 days; P = .001). In the ERP patients, postoperative morbidity was lower in the laparoscopic group than the open ([odds ratio] OR = 0.42; P = .006). As with the previously mentioned review, the quality of the included studies was graded to be moderate to poor. Despite this, the authors conclude that both ERP and laparoscopy are associated with independent benefit but that better designed trials are needed to more definitively answer these questions.

Overall, there has been a great deal of effort put into designing ERPs based on the best evidence available. In general, there are elements supported by extremely strong evidence, such as early initiation of diet and mobilization, and antibiotic prophylaxis (see discussion later), whereas other elements are less well supported. In 2009 the ERAS Group published a consensus review of optimal perioperative care in colorectal surgery. They reviewed the evidence for and made recommendations about 20 ERP elements. Again, although the evidence is not robust for all elements, this remains a good summary of the most common elements of standard ERPs for colorectal surgery. A more recent set of guidelines drew from these recommendations and was reviewed in 2013 by Gustafsson et al., as part of the ERAS Society. The strength of recommendations ranged from low to high for individual elements of the pathway. Although adherence to all elements is difficult and requires multidisciplinary coordination in the perioperative period, there is evidence to suggest that increasing compliance with ERPs is associated with reduced hospital stays and possibly, fewer complications (ERAS Compliance Group). In the 2013 review the authors concluded that there was high-quality evidence that ERPs result in shorter length of hospital stay following colorectal resections. However, the existing evidence suggesting that ERPs result in fewer complications and hospital readmission was deemed to be low.

Mechanical Bowel Preparation

Mechanical bowel preparation before elective colorectal resection remains a common practice among general and colorectal surgeons. However, its use over the past decade has been decreasing, primarily in response to many RCTs and meta-analyses that have not only failed to show a benefit to mechanical bowel preparation but also have demonstrated an increase in complications following bowel preparation.

Two of the earliest RCTs to examine this issue were performed in 1994 by Burke et al. and Santos et al. In both of these studies the authors concluded that bowel preparation does not influence outcome after elective colorectal surgery. Since that time, continued controversy over the use of bowel preparation has spawned several more RCTs. In 2007 Pena-Soria et al. examined the relationship between bowel preparation and surgical-site infection and anastomotic leak in 97 patients. They found no difference in surgical site infection between the two groups, but a higher rate of anastomotic dehiscence in the nonprepped group (8.3% vs. 4.1%; P = .05). The largest RCT examining this question was published in 2007 by Contant et al. and included more than 1400 patients at 13 hospitals. Patients were consented to receive either no bowel preparation, which included a regular diet the day before surgery versus a bowel preparation of either polyethylene glycol or sodium phosphate and a clear liquid diet the day before surgery. In this study the rate of anastomotic leak, 4.8% in patients who received bowel prep and 5.4% in patients who did not, did not differ significantly between groups ( P = .69). Patients who had mechanical bowel preparation did have fewer abscesses after anastomotic leak than those who did not (0.3% vs. 2.5%; P = .001). Other complications, such as fascial dehiscence, superficial infection, and mortality, did not differ between groups. These authors concluded that mechanical bowel preparation before elective colorectal surgery can safely be abandoned. Several studies supported these conclusions for left-sided colon and rectal resections as well.

Further buttressing the argument against mechanical bowel prep were multiple large meta-analyses synthesizing the results from the almost 20 years of trials examining this issue. In 2004, Slim et al. analyzed the results of seven randomized trials, including 1454 patients, comparing bowel preparation with no preparation in colorectal surgery. They reported significantly higher rates of anastomotic leak after bowel preparation (5.6% vs. 3.2%; P = .032). All other end points (wound infection, other septic complications, and nonseptic complications) also favored the no-preparation regimen. In 2010 Zhu et al. specifically analyzed five RCTs that compared mechanical bowel preparation with polyethylene glycol with no preparation. They found no significant differences in rates of surgical site infection, organ/space infection, mortality, or anastomotic leak between the groups. Finally, the largest and most thorough meta-analysis was published by Guenaga et al. in 2009. These authors analyzed 13 RCTs, including 4777 patients, comparing bowel preparation with no bowel preparation. They found that rates of anastomotic leakage, although slightly higher in the bowel preparation groups, were not significantly different following either low anterior rectal resections or colonic resections. Rates of secondary complications, such as wound infection and extraabdominal complications, were not different between the two groups. They concluded that there was no statistically significant evidence that patients benefit from mechanical bowel preparation.

Based on this robust body of evidence, many surgeons began to reduce their use of bowel preparation prior to colorectal surgery. However, interestingly, new evidence is emerging that mechanical bowel preparation with oral antibiotic administration is beneficial prior to elective colorectal surgery. In almost all of the trials mentioned previously, oral antibiotics were not included as part of the mechanical bowel preparation pathway. Many investigators believe that the benefit from bowel preparation stems from the delivery of the oral antibiotics to the colon lumen and mucosa, a process that is enhanced by the mechanical colon cleanse. In light of these concerns regarding the existing bowel preparation literature, a new series of studies have been published evaluating the efficacy of bowel preparations that include oral antibiotics. The results of these studies, which are discussed in more detail later, indicate that, although mechanical preparation alone may not be of benefit, mechanical preparation with oral antibiotics is beneficial in reducing surgical site infection and anastomotic leak following colorectal surgery.

In 2012 Cannon et al. evaluated almost 10,000 patients undergoing elective colorectal surgery within the Veterans Administration Health System. They compared patients receiving no bowel prep to those receiving mechanical-only bowel prep, mechanical bowel prep plus oral antibiotics, or oral antibiotics alone. They reported that oral antibiotics plus mechanical bowel preparation was associated with a 57% decrease in surgical site infection occurrence compared with no bowel prep (OR = 0.43; 95% CI, 0.34 to 0.55).

Following that study, in 2013 Toneva et al. reported on the association between oral antibiotic bowel preparation and length of stay and readmissions in a similar Veterans Administration Health System cohort of 8140 patients. They report that oral antibiotic bowel preparation was associated with a significantly reduced length of stay, as well as a significant reduction in the number of readmissions, due mostly to a reduction in readmission for infection.

In 2014 Kim et al. used the Michigan Surgical Quality Collaborative data to examine almost 1000 pairs of patients undergoing elective colectomy who differed only by administration of bowel preparation. The bowel preparation group received mechanical bowel preparation with nonabsorbable oral antibiotics, and the control group received no bowel prep. These authors found that patients receiving full preparation were less likely to have any surgical site infection (5.0% vs. 9.7%; P = .0001), organ/space infection (1.6% vs. 3.1%; P = .024), and superficial surgical site infection (3.0% vs. 6.0%; P = .001). They were also less likely to develop postoperative Clostridium difficile colitis (0.5% vs. 1.8%; P = .01).

In 2015 four retrospective studies using American College of Surgeons National Surgical Quality Improvement Program–targeted colectomy data were published. Moghadamyeghaneh et al. reported on just more than 5000 patients undergoing elective colorectal resections between 2012–2013. They reported no difference in postoperative morbidity between patients receiving no preparation and either mechanical preparation alone or oral antibiotic preparation alone. Multivariable analysis revealed that the combination of oral antibiotics and mechanical bowel preparation significantly reduced the risk of overall morbidity (OR = 0.63; P < .01), surgical site infection (OR = 0.31; P < .01), and anastomotic disruption (OR = 0.44; P < .01), especially following left-sided resections. Morris et al. examined 8145 patients undergoing elective colon and rectal resections. They found that patients receiving oral antibiotics had a significantly lower risk of surgical site infection than either those patients receiving no bowel preparation or those receiving mechanical preparation only. This was consistent for both open and minimally invasive approaches and for both colon and rectal resections. Scarborough et al. reported on the outcome of almost 5000 patients undergoing elective colorectal resections. Again, they found that patients receiving oral antibiotics combined with mechanical preparation had the lowest rate of surgical site infection, anastomotic leak, and procedure-related readmission. There was no difference noted among the no preparation, oral antibiotic alone preparation, or mechanical preparation alone groups. Finally, Kiran et al. reported on 8442 patients undergoing elective colorectal procedures. After their multivariable analysis, mechanical bowel preparation with oral antibiotics was independently associated with reduced surgical site infection (OR = 0.40; 95% CI, 0.31 to 0.53), anastomotic leak (OR = 0.57; 95% CI, 0.35 to 0.94), and ileus (OR = 0.71, 95% CI, 0.56 to 0.90).

All of these studies have countered the increasingly held belief that bowel preparation prior to elective colorectal surgery is not necessary and may be harmful. Each of these provides retrospective evidence that oral antibiotic administration in combination with a mechanical bowel preparation can have significant beneficial effects for colorectal surgery patients, including decreased risks of wound infection, anastomotic leak, ileus, and readmission. Based on this body of evidence, many providers are routinely using the combination of oral antibiotics and mechanical bowel preparation for their colorectal surgery patients. Randomized controlled data would add to this ample body of retrospective data as the debate around the appropriate use of preoperative bowel preparation continues to evolve.