Enhanced recovery protocols (ERPs) were developed primarily for the care of the colorectal surgical patient during the late 1990s and demonstrated early success in the 2000s.^1-7 These efforts represented consensus from dieticians, nurses, surgeons, and anesthesiologists at the time and ultimately grew into codified components of care with excellent outcomes.^8-11 This work truly represents the culmination of the best science that has assessed the surgical stress response and mitigating therapies, principally from the work of Henrik Kehlet who deserves the title of “Father of Enhanced Recovery After Surgery.”⁸ ERP implementation has evolved from a convoluted and often complex set of care plans to a true discipline for evidence-based care of the surgical patient. Although the early focus was on length of stay, the science has evolved into an approach for improved patient-centric care that deserves the more proper title of “enhanced recovery protocols” (ERPs). The concepts related to preoperative cardiovascular and pulmonary risk assessment and risk modification are well defined by surgical and anesthesia textbooks and preoperative anesthesia clinic processes, and although they are clearly an essential part of a strong perioperative care plan, they are not typically considered ERP components of care. The majority of a gastrointestinal (GI) surgery–related ERP consists of a variety of shared components that will be addressed in a later section. These strategies are designed to recognize and optimize preoperative physiologic adverse factors, the perioperative stress response, narcotic-sparing analgesia, evidence-based reduction of “potentially preventable complications,” early and aggressive ambulation, and early return to enteral intake.^3,8 The optimal use of these strategies has consistently demonstrated a significant reduction in hospital stay and costs, while significantly improving patient safety. The potential components of care include preoperative assessment and education, nutritional repletion, improvement in perioperative glycemic management, anesthesia/analgesia, goal-directed fluid therapy, prevention of nausea and ileus, thromboembolic prophylaxis, minimally invasive techniques, temperature monitoring, early postoperative nutrition, and early mobilization.² The net result of a well-developed program with a high degree of institutional compliance has been a universal improvement in clinical outcomes, reduced length of stay, reduced cost, and most importantly, significantly improved physical recovery of the patient.^5,9 An interesting institutional journey is reflected in the work by Bakker et al,¹⁰ who determined that the strongest predictors for a shorter duration of stay (5.7 days with high compliance vs 7.3 days with low compliance) were no nasogastric tube, early mobilization, early oral nutrition, early removal of epidural, early removal of catheter, and nonopioid oral analgesia. However, despite the institutional recognition of these fairly simple components of care and the benefits, the mean adherence rate was 73% in 2006 and 2007, 66% in 2008 and 2009, 63% in 2010 and 2011, and 82% in 2012 and 2013.¹⁰ This implies that constant monitoring of both process and outcome is essential for durable success with ERPs.

Despite the almost 20-year recognition of the benefits of ERP, the current adoption rate of programs remains disappointingly low.⁹ The low adoption rate is a function of a lack of understanding related to the relative impact of various components of the plan, the perceived complexity of delivering the components, and most significantly, clinicians’ unwillingness to change behavior in the face of incontrovertible evidence. The purpose of this chapter is (1) to define the mutually shared components across GI surgery ERPs and the rationale for these components; (2) to define procedure- and discipline-specific components of care; (3) to review the clinical and financial benefits in favor of ERPs; and (4) to review issues around adoption of ERPs at an institutional level. This chapter will review the shared components of GI surgery–related ERPs and the nuances related to esophageal, gastric, bariatric, liver/biliary, pancreatic, and colorectal surgery.

The Enhanced Recovery After Surgery (ERAS) Society (www.erassociety.org) has published multiple guidelines related to GI surgery, and the reader is referred to 2 major society websites focused on ERP that provide a variety of documents for open access to ERP plans (American Society of Enhanced Recovery at aserhq.org and Society of American Gastrointestinal and Endoscopic Surgeons at www.sages.org/smart-enhanced-recovery-program/). The organizations provide analyses and quality rankings of the various data elements that can be interpreted by individual healthcare systems for relative benefits for local implementation. It is intriguing that despite the large database available to the ERAS Society, little has been published on actual outcomes, unlike the ubiquitous National Surgical Quality Improvement Program (NSQIP)–related publications. Once again, the attempt to assess the long lists of multiple components on these sites can be intimidating early in the process development of an organization that requires sometimes cumbersome change management strategies. It is well beyond the scope of this chapter to discuss the complex science of change management, but it is clear that a strategy of rapidly assessing current outcomes and implementing an aligned strategy can lead to rapid process improvement. This chapter will attempt to organize components of an ERP strategy into the following components of the episode of care to allow systems to assess the requisite integration of care required: preadmission; day of surgery; and intraoperative, postoperative, and discharge planning.

The major impediments to successful implementation are typically system based and due to fragmented care delivery and debates over the validity of one or another element.^10-15 The experience by Nelson et al¹⁶ and his colleagues¹⁷ in Alberta, Canada, which provided regular outcomes data, may be an effective means to overcome resistance. Particularly in the United States, the lack of economic integration and therefore structural integration can increase the difficulties of change management. The lack of economic integration is also challenging because some components of the plan need to be administered before admission and, therefore, are not bundled into hospital payments, resulting in patient out-of-pocket costs that are not typically reimbursed. This is a significant stumbling block to the successful compliance with components such as immunonutrition and carbohydrate loading. Another important obstacle is the perceived need for complex risk adjustment models. In fact, US hospitals can access the important cost drivers of care using the readily available and routinely submitted administrative data. Alternatively, many hospitals participate in national programs, such as NSQIP or the Vizient programs, which provide easily accessible and robust data related to both outcomes and cost. It should be a relatively easy discussion to review high-impact negative outcomes and, in the true spirit of quality improvement, implement a plausible solution and in rapid sequence assess change in outcome.^18-22 An example of how difficult this approach can be is exemplified by the work of Harbaugh et al²³ assessing the only US Food and Drug Administration–approved prophylactic agent for reduction postoperative ileus (POI), which demonstrated that even across members of an advanced quality collaborative, adoption of a component with proven benefits is highly variable.

Preadmission

ANEMIA

Preoperative anemia occurs frequently in GI surgical patients as a result of chronic blood loss, nutritional deficiencies (typically iron), and the impact of neoadjuvant oncologic treatment. Conversely, outside of emergency surgery, which has a separate set of strategies for blood conservation and repletion, the preoperative period provides a sufficient time frame to address preexisting anemia.^24-28 Preoperative anemia and blood transfusions are associated with a higher incidence of postoperative infections, a longer hospital stay, higher cost, and a worse overall outcome.^29,30 Iron deficiency anemia is by far the most frequent form of preoperative anemia, is particularly common among the elderly population, and is readily evaluated and treated in a time-compressed manner.^31,32 The opportunity for a system to include routine complete blood count and reflex to specific iron studies based on specific thresholds can lead to a cost-effective and efficient means of identifying at-risk patient populations.^33,34 Incorporation of this type of testing strategy can further lead to an integrated approach to effective treatment with iron infusions and a reduction in the need for perioperative transfusions.^35,36

SARCOPENIA

Sarcopenia is a common adverse risk factor for a variety of GI surgical procedures.^37-41 Once again, the ability to identify this risk factor is readily available as virtually all GI surgery patients are evaluated with abdominal computed tomography and the nomograms are well accepted. A typical algorithm is the use of a single slice at the level of the third lumbar vertebra (L3) or the measurement of total fat area (cm²), subcutaneous fat area (cm²), visceral fat area (cm²), and skeletal muscle area (cm²) using accepted Hounsfield unit (HU) thresholds (adipose tissue, −190 to −30; skeletal muscle, −29 to +150). These parameters are then normalized for patient stature and designated as total fat index (cm²/m²), subcutaneous fat index (cm²/m²), visceral fat index (cm²/m²), and skeletal muscle index (cm²/m²) in line with accepted methodology.^42,43 Sarcopenia is defined as a skeletal muscle index <43 cm²/m² for males and <41 cm²/m² for females using previously published cutoff values.⁴⁴ Unfortunately, despite the growing recognition of the frequency and significant impact of sarcopenia, there is surprisingly little information in the surgical literature regarding the appropriate method for repletion and/or the ability to reverse some percentage of the physiology in a time-compressed manner to influence surgical outcomes.

Current treatment strategies are designed to address nutritional supplementation of proteins, essential amino acids, and fatty acids, combined with focused resistance training and physical activity. Skeletal muscle possesses an inherent capacity for regeneration due to activation of resident satellite cells and is regulated in part by host innate immune responses, especially the macrophage response.⁴⁵ In addition, muscle wasting in the surgical patient can also be associated with chronic inflammatory diseases and the related pathophysiologic impact of proinflammatory cytokines including interferon-γ, interleukin (IL)-1, tumor necrosis factor-α, IL-6, IL-18, and IL-8.⁴⁶ This information is essential to understand the current gap in the ERP literature regarding the potential benefits of a specific regimen.^47,48 A number of meta-analyses regarding the assessment of immunonutrition have been published; however, virtually all of the data are from a period of time prior to adoption of ERPs, leading to conflicting outcomes when recent data are included.^49-54 This gap is highlighted by the meta-analysis performed by Hegazi et al,⁵² which assessed 8 randomized controlled trials (RCTs) of preoperative immunonutrition versus standard enteral therapy and 9 RCTs of immunonutrition versus no supplements.⁵² The authors found no advantage with immunonutrition over standard protein supplementation. Similarly, the comparison of 3 recent studies demonstrates the same conundrum regarding the benefits of supplement components within an ERP. Moya et al⁵⁵ randomized patients to receive either no supplement or immunonutrition and reported a reduction in surgical site infection (SSI) in the absence of a mechanical bowel preparation (MBP) or antibiotic bowel prep. Hübner et al⁵⁶ randomized patients undergoing elective major GI surgery to either immunonutrition or isocaloric isonitrogenous nutrition given for 5 days preoperatively within an ERAS pathway and demonstrated no improvement in outcomes. Finally, Thornblade et al⁵⁷ conducted a retrospective assessment of a quality database that assessed a population that received recommendations for usage of a preoperative supplement but provided no data on degree of compliance and suggested outcomes were better with the recommendation. These data are further complicated by recent data suggesting that glutamine, arginine (vs citrulline), and omega-3 fatty acids are associated with either increased risk or no benefit in many stressed patient populations.^58-65 The current data suggest that the commonly recommended supplements seeking to address benefits related to immunonutrition likely do not provide the mix of low carbohydrate, high leucine, and vitamin D components that appear optimal for sarcopenia.^66-69 Therefore, at this time, the data suggest that patients should be investigated for the presence of sarcopenia and that patients with sarcopenia will benefit from a supplement versus no supplement. It remains unclear what the refinements in supplementation will be over time as well as what the relative benefits of immunonutrition are for nonsarcopenic and normally nourished patients.

CARBOHYDRATE LOADING

At least 15% to 35% of patients undergoing major general, gynecologic, urologic, cardiothoracic, and orthopedic surgery will experience significant hyperglycemia in the immediate preoperative period, even if they are not diabetic. Postoperative hyperglycemia in nondiabetics is associated with at least a 2-fold increase in the risk of both surgical site infection (SSI) and mortality.^70-74 The risk of postoperative hyperglycemia is exacerbated by the commonly used and unnecessarily prolonged preoperative period of no oral intake (NPO), which creates a starvation-induced insulin resistance and gluconeogenesis response. Most ERPs recognize the risk associated with a prolonged NPO period and recommend the provision of a maltodextrin-containing beverage both the evening before and just 2 hours prior to surgery.^75,76 The concept of perioperative “carbohydrate loading” is frequently misunderstood, and practitioners may rely on sports drinks designed to support athletic-induced carbohydrate consumption from muscle activation and the associated dehydration. The administration of simple sugars (especially fructose) using fruit juices or sports drinks delivers an excessive glycemic index load, resulting in rapid and early glucose and insulin spikes followed by compensatory glucagon secretion, which does not improve insulin sensitivity. As a result, there are no published data assessing the impact of these products on “perioperative carbohydrate loading.”

Several studies using euglycemic clamp have demonstrated improved perioperative insulin sensitivity with administration of maltodextrin, as well as a reduction in postoperative gluconeogenesis, which together improve postoperative glucose metabolism for as long as 72 hours postoperatively.^77-79 However, the recent data from the PROCY trial suggest that the delivery of the recommended 3 doses of 40+ g of maltodextrin per dose does not sufficiently reduce the population rate of hyperglycemia (25% range) to decrease perioperative complications.⁸⁰ However, a strategy of administering 3 doses of 25 g provided similar benefits, with a perioperative hyperglycemia rate of 7%.⁸¹

The preoperative loading period also allows for the opportunity to support a recently documented impact of surgical stress on the reduction of arginine bioavailability and an associated increase in asymmetric dimethyl arginine (ADMA), which is a natural inhibitor of arginine-associated nitric oxide function.^82-84 The net result is a lowering of the arginine/ADMA ratio in the early postoperative period, which is associated with increased SSI rates, cardiovascular complications, and acute kidney injury.^85,86 Both Ekeloef et al⁸⁷ and Ragina et al⁸⁸ recently demonstrated a significant reduction (20%-25%) of both arginine and endothelial function after colectomy. L-Citrulline has recently and consistently been demonstrated to safely and effectively restore systemic arginine levels and reduce ADMA in a variety of clinical scenarios. The major reason for reliance upon citrulline is that surgical stress increases the function of constitutively active hepatic arginase, which degrades a significant component of enterally administered arginine, rendering it inactive.^89-91 Conversely, virtually all enterally absorbed citrulline passes through the liver to subsequently be converted virtually completely to arginine in the kidney. The net result is that citrulline directly supports systemic access to arginine for use by all end organs and immune response cells (ie, macrophages and lymphocytes).^92-94 The higher degree of bioavailability of systemic arginine is also important because of its ultimate conversion to ornithine and then proline, which supports wound healing via collagen formation. Higher doses of arginine required to support similar systemic levels are limited by the GI side effects on small bowel secretion of fluid and electrolytes.^92-94 Therefore, the preoperative “loading” period may need to be further investigated to allow support of important aspects of both glycemic management and endothelial function.

BOWEL PREPARATION (COLON RESECTION ONLY)

The classic article by Condon et al⁹⁵ assessed bowel preparation strategies in a 3-arm study comparing intravenous cephalothin alone versus oral neomycin and erythromycin alone versus both intravenous and oral regimens. This 3-arm trial showed superior outcome in the dual regimen; however, the intravenous medication was limited in bacterial coverage, which may have impacted the results. This issue was addressed by Coppa and Eng,⁹⁶ with 350 patients randomized to intravenous cefoxitin with or without oral neomycin and erythromycin. They found significant improvement as well with dual regimens for wound infection (11% vs 5%). As a result of that work and other work, including the seminal work of Nichols and Condon, MBP including oral antibiotics supplemented with preoperative intravenous antibiotics has been a mainstay of colon surgery for decades.^97-101 MBP has been primarily associated with reductions in SSI, especially superficial wound infection, although more recently, it has been associated with a reduction in anastomotic leak.^102,103 However, in the ERP era, the utility of MBP has been questioned by 2 meta-analyses evaluating recent data.^104,105 The meta-analysis by Bucher et al¹⁰⁵ included 7 RCTs available in the literature and suggested a higher incidence of anastomotic dehiscence in patients receiving MBP (5.6%, 36/642 patients) versus no MBP (2.8%, 18/655 patients; P = .03; odds ratio [OR], 1.85; 95% confidence interval [CI], 1.06-3.22). The rate of intra-abdominal infection (peritonitis or abscess) was similar in the MBP group (3.7%, 17/458 patients) compared with the no-MBP group (2.0%, 9/461 patients; OR, 1.69; 95% CI, 0.76-3.75; P = .18).¹⁰⁵ The rate of wound infection was not significantly different in patients receiving MBP (7.5%, 48/642 patients) versus no MBP (5.5%, 36/655 patients; OR, 1.38; 95% CI, 0.89-2.15; P = .15).¹⁰⁵ The meta-analysis is significantly impacted by 2 studies. The first is by Contant et al¹⁰⁶ that studied 1431 patients undergoing open colorectal resection randomized to intravenous antibiotics (aerobic and anaerobic coverage) with or without MBP. The data demonstrated a significant increase in the rate of intra-abdominal abscess (2.5% vs 0.3%); however, there was no significant difference in superficial wound infection (no MBP vs MBP, 14% vs 13.8%) or anastomotic leak (no MBP vs MBP, 5.4% vs 4.8%). The authors concluded that MBP can be safely avoided.¹⁰⁶ However, the increase in pelvic abscess rate and a fairly high superficial wound infection raise some concern over this recommendation and possibly the negative impact of no oral antibiotics. Jung et al¹⁰⁷ performed a similarly designed study of 1505 open colectomy patients and also concluded that there was no significant difference in wound infection (MBP vs no MBP, 7.8% vs 6.4%) or anastomotic leak (MBP vs no MBP, 2% vs 2.6%). A major pitfall of these combined data is the absence of the putative effective treatment that incorporates oral antibiotics with the MBP.

In recent years, many papers have reviewed large quality databases, and the consistent theme seems to be a significant reduction in both SSIs and anastomotic leaks.^108-112 The Michigan Surgical Quality Consortium analyzed 2062 elective colectomies between January 2008 and June 2009; 49.6% of patients were administered MBP and 36.4% received MBP and oral antibiotics. Patients receiving oral antibiotics were less likely to have any SSI (4.5% vs 11.8%; P = .0001), to have an organ space infection (1.8% vs 4.2%; P = .044), and to have a superficial SSI (2.6% vs 7.6%; P = .001).¹⁰² Patients receiving bowel preparation with oral antibiotics were also less likely to have a prolonged ileus (3.9% vs 8.6%; P = .011). Similarly, Kiran et al¹⁰³ reviewed the National Surgical Quality Improvement Program–targeted colectomy data initiated in 2012 to capture information on the use and type of bowel preparation and colorectal-specific complications. They found that in 8442 patients, MBP with antibiotics, but not without, was independently associated with reduced anastomotic leak (OR, 0.57; 95% CI, 0.35-0.94), SSI (OR, 0.40; 95% CI, 0.31-0.53), and POI (OR, 0.71; 95% CI, 0.56-0.90).¹⁰³ A recent meta-analysis and a review of the older prospective randomized trials came to the same conclusion based on high-quality studies.^108,113 Finally, Wick et al¹⁰⁹ described the incremental benefit of adding an MBP with oral antibiotics to their ERP with significant reductions in SSI. Therefore, despite the recommendations of the ERAS Society and in the absence of any convincing data from their data set or a well-powered study comparing no preparation to MBP with oral antibiotics, this author recommends the latter strategy as part of an effective ERP.

PREOPERATIVE EDUCATION

The importance of providing consistent, precise, and easily understood information regarding the episode is key to developing an effective ERP.^114-116 It is beyond the scope of this chapter to define specifics because the goal of the educational program needs to be both patient centric and system specific. It goes without saying that in order to provide high-quality care and allow a pathway for the patient to be an effective member of the team it is essential that everyone agrees to the components of care. Patients are highly sensitive to variations in messaging, and failure in consistently messaging the processes and goals can render an ERP ineffective.

Day of Surgery

SURGICAL CARE IMPROVEMENT PROJECT PROCESSES

The Surgical Care Improvement Project (SCIP) program campaign began in August 2005 as a mandated national initiative with public reporting of compliance designed to primarily reduce the risk of SSI. There has been considerable debate on the relative benefits of various components, or even the degree of compliance on outcomes related to institutional adoption of the SCIP, and ultimately, the program was retired in December 2015. With respect to ERP, the important process measures that seem to be highly effective and that should be adhered to include selection of an appropriate parenteral antibiotic; administration of that antibiotic within 1 hour preoperatively; termination of the antibiotic prophylaxis within 24 hours of surgery; removal of the urinary catheter within 24 hours; and appropriate deep venous thrombosis prophylaxis.^117-122

ILEUS PROPHYLAXIS

POI had traditionally been perceived as an unavoidable outcome of major abdominal surgery, primarily due to poorly understood multifactorial pathophysiology.¹²³ Although POI is frequently blamed on factors out of the control of the surgeon, including neurogenic stimuli, release of inflammatory mediators, or requisite surgical manipulation, it has become clear that the majority of the cause is related to narcotic analgesics.¹²³ Surgery-related mediators also cause the release of endogenous opioid peptides that further exacerbate the effects of exogenous opioid analgesics (administered for analgesia) on the inhibition of bowel function.^124-129 POI occurs at a lower rate following minimally invasive surgical procedures due to a reduction in surgical trauma and postoperative pain but may still occur due to the effects of opioid analgesics.^129,130 Although not life-threatening, POI prolongs postoperative hospital stay and healthcare resource utilization and costs.^130-133 Therefore, without a reduction in POI, an ERP will be unsuccessful in safely reducing the hospital stay and potentially the readmission rate. Although a narcotic-sparing analgesic regimen (see later in this chapter) can minimize the risk of POI, the availability of alvimopan, a first-in-class oral μ-opioid receptor antagonist, offers the only prophylactic treatment that reduces the rate of POI.¹³⁴ A pooled analysis of the phase III prospective, randomized, and blinded alvimopan trials confirmed that a 12-mg dose provided optimal reduction in GI morbidity and return of GI function.¹³⁵ Subsequent to the prospective randomized trials, several large quality databases have been interrogated and confirmed the system-level benefits of a strategy of POI prophylaxis that incorporates alvimopan in the care plan.^23,136-139 Gum chewing has been advocated as another option to reduce the rate of POI; however, in a program employing early feeding strategies, the relative benefits of chewing gum remain unclear, but it is inexpensive and apparently safe.^140,141

NAUSEA AND VOMITING PROPHYLAXIS

Postoperative nausea and vomiting (PONV) are common and unpleasant side effects associated with anesthesia and surgery, with an incidence of approximately 30%.¹ High-risk patients may have a considerably higher incidence, especially females, nonsmokers, patients with a history of motion sickness or migraines, and patients exposed to narcotics or volatile anesthetics.^142,143 Current therapeutic options include a combination of antiemetics acting at different receptors.¹⁴⁴ The major receptor systems are involved in PONV including the cholinergic (muscarinic), dopaminergic (D₂), and histaminergic systems. Ondansetron, granisetron, dolasetron, and tropisetron have shown efficacy for PONV prevention and are associated with a low incidence of side effects. Metoclopramide acts on both central dopamine and serotonin receptors and has both prokinetic and antiemetic effects but may be limited due to extrapyramidal side effects. Dexamethasone is an effective antiemetic, although its mechanism of action remains uncertain.¹⁴⁴ Based on current evidence, a multimodal approach to PONV should include the following strategy: (1) preoperative anxiolysis; (2) aggressive hydration (25 mL/kg) in outpatients unclear of impact in guided fluid management for major surgery; (3) oxygen; (4) prophylactic antiemetics (dexamethasone 10 mg at induction and ondansetron 1 mg at end of surgery); (5) total intravenous anesthesia with propofol, remifentanil, and a nonsteroidal anti-inflammatory drug; and (6) avoidance of nitrous oxide.^145,146

TRANSVERSUS ABDOMINIS PLANE BLOCK FOR ANALGESIA

The initial description of the landmark technique for performing transversus abdominis plane (TAP) block advocated a single entry point, the triangle of Petit, to access a number of abdominal wall nerves, hence providing more widespread analgesia.^147,148 Ultrasound guidance was subsequently recommended to improve localization and deposition of the local anesthetic and was associated with a sensory block from T7 to L1.¹⁴⁹ Radiologic evidence suggests that 20 mL of dye in the TAP 20 to 240 minutes after injection migrated from the superior margin of the iliac crest to the level of the costal margin and posteriorly to the quadratus lumborum.¹⁴⁹ The “4-quadrant TAP block” is a further enhancement of analgesia and is beneficial due to analgesic impact to both the intercostal (upper TAP) plexi and the deep circumflex iliac artery plexi (lower TAP plexus).^150,151 Although data are pending from several ongoing trials, a 4-quadrant block appears to be a safe and inexpensive adjunct to an effective narcotic-sparing analgesia. This is supported by a variety of large data set analyses as well as a single-institution experience within an otherwise unchanged ERP.^152-158

EPIDURAL ANESTHESIA AND ANALGESIA

Epidural anesthesia as a component of ERP originated from the early work of Henrik Kehlet and his team, who investigated the potential benefits of reduction of the perioperative stress response.^159,160 Interestingly, unlike many of the subsequent analyses of ERP components by this team, it remained untested in randomized studies. As a result, it remains a component of recommendations of the ERAS Society (http://erassociety.org/guidelines/list-of-guidelines/). However, the growing understanding and adoption of various narcotic-sparing strategies have generally reduced the impact of epidural analgesia within an ERP. The majority of recent studies, especially in laparoscopic colorectal surgery, have identified limited adjunctive analgesia benefits.^161-166

The title of this section was purposely changed to “guided” rather than “goal-directed” to attempt to refine the direction the therapeutic management of perioperative fluid appears to be heading. The current discussion includes liberal versus restrictive and goal directed as the 2 main approaches to particularly intraoperative fluid management. Conversely, the concept of “guided” offers a middle ground of fluid administration based on additional assessments identifying patients who may or may not be fluid responsive. The latter may benefit from a strategy of judicious pressor therapy for hypotension or inotropic support for impaired myocardial contractility.^167,168

A report from the United Kingdom in 1999 identified that fluid imbalance led to serious postoperative morbidity and mortality and was associated with a high frequency of poor documentation of fluid balance.^169,170 That same report suggested that overhydration was a contributory cause of postoperative morbidity and mortality.¹⁶⁹ Both hypervolemia and hypovolemia impair cardiac function, pulmonary function, tissue oxygenation, wound healing, POI, renal function, and coagulation, which may all be affected by perioperative fluid administration.¹⁷¹ Therefore, the true relationship between postoperative complications and volume loading is a U-shaped curve with the goal being lower on either arm of the U.¹⁷²

The traditional approach to fluid therapy includes replacement of the fluid lost (by basal fluid requirements, perspiration through the surgical wound, loss to the third space and blood loss, and exudation through the surgical wound) and maintenance of physiologic functions (preloading of neuraxial blockade).¹⁷³ This approach has been associated with high volumes manifested by postoperative weight gain. Alternatively, restricted fluid therapy is based on a mL/kg/h strategy that seeks to achieve a zero balance.¹⁷³ The data have clearly demonstrated that a liberal versus restrictive fluid management is consistently associated with a greater incidence of major postoperative complications.¹⁷⁴ However, there is an often underappreciated complication rate associated with underresuscitation, primarily manifested by acute kidney injury.¹⁷⁴ Therefore, a recent critical analysis of the available studies reported that 3 of the trials showed improved outcome after restrictive fluid regimens; 2 trials showed no difference in the outcome.¹⁷⁵

Goal-directed fluid management strategies have become popular and rely on one or another means of optimizing cardiac output; however, the data on relative benefits of supranormal cardiac output remain elusive.¹⁷⁶ Therefore, this approach is rapidly morphing into a concept of guided or individualized fluid administration, primarily based on a an assessment of fluid responsiveness.¹⁷⁷ The appropriate identification of fluid responsiveness under general anesthesia and mechanical ventilation requires a dynamic parameter of cardiac function. The current methodologies include indicators derived from pulse power analysis, pulse contour analysis, esophageal Doppler monitoring, and others.^178-181 Esophageal Doppler uses a thin plastic tube placed in the esophagus to calculate cardiac output based on the amount of blood that moves past the probe over a given time (stroke distance) and estimates the cross-sectional area of the aorta determined from nomograms.¹⁷⁸ Fluid responsiveness is then implied by changes in stroke volume. A typical strategy uses an increase in stroke volume of at least 10% by a fluid bolus of 3 mL/kg, as consistent with fluid responsive, and the boluses continue until that 10% increase is reached.¹⁷⁸ Alternatively, arterial pulse contour analysis measures the stroke volume on a beat-to-beat basis from an arterial pulse waveform, but the main drawback of this method is that it is an invasive procedure.¹⁸¹ Respiratory variations in the arterial pulse pressure in patients on positive-pressure ventilation can inform clinicians about the status of a patient on the Frank-Starling relationship. High respiratory variations (>15%) mean that the patient is on the steep portion of the curve, and low respiratory variations (<10%) indicate that the patient is on the plateau (ie, not fluid responsive).¹⁸² A similar strategy of boluses can be administered until the plateau is reached. The specific fluid remains controversial, with recommendations for colloid or balanced crystalloid as the predominant solutions.^183-188

Postadmission

Narcotic-Sparing Analgesia

Effective analgesia with minimal side effects is one of the most important components of ERPs and, when fully adopted, results in high-quality and high-value surgical care. The same care plans can be adopted for a broad range of GI surgeries, but a transition in a conceptual approach to analgesia is required. We need to continue to separate ourselves from the concepts of the 1990s when regulatory agencies such as The Joint Commission identified pain as the “fifth vital sign” and the Centers for Medicare and Medicaid Services included satisfaction with analgesia as a quality reporting measure.¹⁸⁹ The result was a tremendous increase in the use of narcotics, and unfortunately, a narcotic epidemic has occurred, with unintentional drug overdose now the second leading cause of accidental death.¹⁹⁰ Evolving acute pain management from the various forms of the World Health Organization “analgesic ladder” to the concepts of (1) TAP block/neuraxial block; (2) scheduled narcotic-sparing multimodal oral analgesia (nonsteroidal anti-inflammatory drugs [NSAIDs], acetaminophen, and gabalins); (3) oral narcotic for initial breakthrough pain; and (4) reservation of parenteral narcotics for severe residual pain.¹⁹¹ The increasingly limited use of epidurals for the majority of ERP GI surgery indications has been mentioned earlier. In addition, the field of pharmacogenetics has yielded preliminary data regarding the ability to significantly reduce the amount of narcotics and treatment-related pain scores by using patient-centric optimally metabolized medications.¹⁹²

As mentioned earlier, an oral, scheduled, multimodal analgesic plan is a key construct of ERP analgesia.^193-196 NSAIDs are potent analgesics (600 mg of ibuprofen is as efficacious as 15 mg of oxycodone hydrochloride) and act through inhibition of cyclooxygenase and prostaglandin synthesis.¹⁹⁷ The addition of NSAIDs (including nonselective and cyclooxygenase-2 inhibitors) has clearly and consistently been tied to superior analgesia and opioid-sparing effects.^194,195 NSAID administration has been associated with platelet dysfunction, GI tract irritation or bleeding, and renal dysfunction. A Cochrane review that examined 23 trials (comprising 1459 patients) noted that NSAIDs caused a clinically unimportant transient reduction in renal function in the early postoperative period in patients with normal preoperative renal function and should not be withheld from adults with normal preoperative renal function because of concerns about postoperative renal impairment.¹⁹⁸ Conversely, the perceived risk of NSAID-related increases in postoperative bleeding has been refuted by a significant amount of data.^199-202There has been growing concern regarding NSAID-related anastomotic leak, which has been suggested in several reports. Some data suggest an association between NSAID use and an increase in anastomotic leakage; however, further studies are needed to determine the validity of this association.^203-207 There is some literature suggesting that variable systemic levels can occur in patients with mutations in the CYP2C8 or CYP2C9 genes, which can result in delayed metabolism and therefore supratherapeutic drug levels.^208-210 Therefore, utilization of naproxen, which is one of the few NSAIDs (selective or nonselective) that is excreted unchanged in the urine and therefore not impacted by genetic variation, may offer a safer strategy.

Acetaminophen is available in both oral and parenteral forms and should also be provided in a scheduled fashion within an ERP. The recurring theme is improved analgesia, narcotic reduction, and reduced opioid side effects.²⁰² Although theoretically tied to hepatic toxicity, the risk is minimal when acetaminophen is administered in the usually recommended dosage range, which yields an additive analgesic benefit to concomitant NSAID use.