American Society for Enhanced Recovery (ASER) and Perioperative Quality Initiative (POQI) joint consensus statement on optimal analgesia within an enhanced recovery pathway for colorectal surgery: part 1—from the preoperative period to PACU

Use of a single-shot spinal opioid (i.e., morphine or hydromorphone) is associated with significantly lower pain at rest and on movement, and reduced opioid requirements (Meylan et al. 2009). These benefits are more prominent in patients undergoing abdominal versus other types of surgery (e.g., cardiac). Although the dose range for this meta-analysis varies considerably (dose range, 100–4000 mcg), current practice tends toward using lower doses of intrathecal morphine (<0.3 mg) as higher dose of intrathecal morphine are associated with more episodes of respiratory depression (Gehling and Tryba 2009). Some centers use a spinal dose of bupivacaine as a carrier for the opioid to cover the incision although there is often a resulting sympathetic block.

Use of thoracic epidural anesthesia (TEA) for open CRS is associated with superior postoperative analgesia (Werawatganon and Charuluxanun 2005; Block et al. 2003), decreased pulmonary/cardiac morbidity (Popping et al. 2014), and earlier return of gastrointestinal function as compared to parenteral analgesia (Marret et al. 2007; Hughes et al. 2014). However, the overall benefits of TEA in improving recovery or decreasing length of stay in patients undergoing laparoscopic colorectal procedures are uncertain (Liu et al. 2014; Khan et al. 2013). Concern that prolonged sympathetic blockade in TEA requires patients to have further intravenous fluids to maintain intravascular volume in face of arterial hypotension were not observed in one meta-analysis of TEA versus patient-controlled analgesia (PCA) in laparoscopic colectomy (Liu et al. 2014).

Peripheral regional analgesia options for CRS patients include transversus abdominis plane (TAP), paravertebral, or wound and peritoneal infiltration blocks/catheters. All have been shown to some extent to improve perioperative analgesia while decreasing opioid use. Paravertebral blocks and catheters for surgical anesthesia at the level of the thoracic and lumbar vertebrae are associated with less pain during the immediate postoperative period (Thavaneswaran et al. 2010).

Multiple meta-analyses indicate that TAP blocks/catheters for abdominal surgical procedures are associated with superior analgesia and decreased postoperative opioid consumption compared to opioid analgesia alone (Baeriswyl et al. 2015; Zhao et al. 2014; Johns et al. 2012; Siddiqui et al. 2011; Charlton et al. 2010). Preoperative (vs. postoperative) TAP block administration appears to have greater effects on early pain and opioid consumption compared with postoperative administration although the effect of preoperative (vs. postoperative) TAP blocks on longer-term outcomes is unknown (De Oliveira et al. 2014). TAP blocks provide comparable short-term analgesia to wound infiltration but provide superior analgesia in longer term and in the setting of a multimodal analgesic regimen (Yu et al. 2014; Guo et al. 2015a). Concerning the choice of local anesthetic, ropivacaine, bupivacaine, and liposomal bupivacaine have all been used in ERPs with good results (Hamada et al. 2016; Cohen 2012). Liposomal bupivacaine shows some promise for longer-term postoperative analgesia either as infiltration or for TAP blocks (Hutchins et al. 2015); however, there are no large-scale randomized controlled trials available in intra-abdominal surgery to guide practice or to definitively demonstrate the analgesic efficacy of this intervention (Cohen 2012; Candiotti et al. 2014). In short, while there is insufficient evidence to recommend one medication over another at this point in time, it should be noted that additives, such as dexamethasone, are needed to prolong the duration of non-liposomal mixtures (Akkaya et al. 2014).

Intraperitoneal instillation of local anesthetics during major abdominal surgery, including open and laparoscopic colectomy, is associated with significantly lower pain scores postoperatively; although one study did continue TEA in addition to intraperitoneal instillation for 2 days after surgery (Marks et al. 2012; Kahokehr et al. 2011; Park et al. 2011). Additionally, wound infiltration has been shown to be associated with a decrease in morphine consumption and significantly lower pain scores in the early postoperative period in abdominal surgery patients, but none specifically in colorectal surgery (Bamigboye and Hofmeyr 2009). A meta-analysis suggested that the use of local anesthetic wound infiltration was associated with pain scores comparable to those obtained with epidural analgesia and a slight decrease in opioid use, although the data was noted to be quite heterogeneous and significance might exist in the patients receiving each treatment (Ventham et al. 2013). The analgesic efficacy of local anesthetic infusion through wound catheters is uncertain, and meta-analyses have been conflicting in CRS (Liu et al. 2006; Gupta et al. 2011).

Major non-opioid oral analgesic agents include non-steroidal anti-inflammatory agents (NSAIDs), acetaminophen (paracetamol), gabapentinoids (gabapentin and pregabalin), and tramadol. All except oral tramadol have been examined within the setting of intra-abdominal surgery and have been found to have significant effect on reducing the opioid burden postoperatively. As such, routine, scheduled use of these agents should be considered as part of a plan to achieve optimal analgesia after colorectal surgery.

Acetaminophen (paracetamol) when administered as part of a multimodal regimen is associated with a decrease in pain and decrease in opioid usage, which may result in a decrease in some opioid-related side effects (Doleman et al. 2015; De Oliveira et al. 2015; Wong et al. 2013; Apfel et al. 2013; McNicol et al. 2011; Toms et al. 2008; Remy et al. 2005). A single dose of IV acetaminophen (paracetamol), typically in a dose of 1 g, given prior to surgery (meta-analysis of 11 RCTs of 740 patients) was associated with significantly lower early pain at rest, early pain with movement, postoperative opioid consumption, and postoperative nausea and vomiting (De Oliveira et al. 2015). In a meta-analysis of 7 RCTs (n = 544 participants), 1 g or 15 mg/kg of IV acetaminophen (paracetamol) given 10–30 min before induction/incision (vs. the same dose given 10–30 min at the end of surgery/before skin closure) was associated with a reduction in 24-h opioid consumption and a lower incidence of postoperative vomiting in the preventive acetaminophen (paracetamol) group (Doleman et al. 2015). Most studies including pharmacokinetic outcomes reported higher postoperative plasma concentrations and larger proportions of patients achieving target plasma concentrations after IV dosing compared with oral dosing (Jibril et al. 2015). However, for patients who can take oral medications preoperatively, there does not appear to be evidence of a clear benefit of the intravenous formulation. Decision making should take into account of convenience and cost (Jibril et al. 2015).

NSAIDs, whether non-selective or cyclooxygenase-2 inhibitors (COX-2), when administered as part of a multimodal regimen, are associated with a decrease in pain and decrease in opioid usage which may result in a decrease in some opioid-related side effects (De Oliveira et al. 2012; Marret et al. 2005; Straube et al. 2005; Maund et al. 2011; Michelet et al. 2012; Elia et al. 2005). Use of COX-2 inhibitors has minimal effect on coagulation even at supra-therapeutic doses (Leese et al. 2000). A systematic review noted that preoperative COX-2 inhibitors significantly reduced postoperative pain, analgesic consumption, and antiemetic use, and improved patients satisfaction compared with preoperative placebo (Straube et al. 2005). In the studies examining celecoxib, the doses used were 200 or 400 mg PO, and for parecoxib, they were 40 mg PO [Pandazi, 2010].

It is uncertain whether the perioperative use of NSAIDs carries a risk of harm. While it is unlikely that NSAIDs increase the risk of renal injury in euvolemic patients who do not have contraindications to receiving these medications (Myles and Power 1998), caution should be undertaken in patients who are hypotensive or thought to be hypovolemic. Additionally, there is a concern for the potential for an association with increased anastomotic leak, but the literature surrounding this question is not conclusive. As such, insufficient evidence is available to recommend against routine use of NSAIDs, especially COX-2 inhibitors, as these medications are effective in treating pain and reducing opioid use in the perioperative period (Chou et al. 2016a; Bhangu et al. 2014).

Several meta-analyses including studies concerning intra-abdominal surgery suggest that gabapentinoids (gabapentin, pregabalin) when given as a single dose preoperatively are associated with a decrease in postoperative pain and opioid consumption at 24 h (Engelman and Cateloy 2011; Eipe et al. 2015; Hurley et al. 2006; Mishriky et al. 2015; Peng et al. 2007; Seib and Paul 2006; Zhang et al. 2011). For gabapentin, a preoperative dose of 300–1200 mg is associated with lower pain scores (both at rest and with movement) and reduced opioid consumption (Hurley et al. 2006; Peng et al. 2007; Seib and Paul 2006). It should be noted that one small RCT found that a single preoperative dose of gabapentin 600 mg PO did not significantly reduce opioid consumption or pain scores on POD 1 or 2 for patients presenting for colectomy (Siddiqui et al. 2014). However, opioid consumption and pain scores were lower at all time points in the gabapentin group compared to placebo, but there were only 36 patients per group and it was underpowered to detect any difference. As noted by the authors, continuing doses in the postoperative period may confer added benefit given the pharmacokinetics of gabapentin. This corresponds with the dosing reported in successful ERPs for CRS where gabapentin is used as one component to reduce opioid consumption in the perioperative period (Larson et al. 2014; McEvoy et al. 2016). However, the exact contribution of gabapentin to these positive outcomes is unknown. For pregabalin, a recent meta-analysis indicated that pain scores at rest were reduced with all doses of pregabalin (mostly 75–300 mg) but pain scores with movement were only reduced with the 300 mg dose and there were no significant differences in side effects between the three dose levels of pregabalin. The opioid-sparing effect of pregabalin appeared to be limited to doses 100–150 and 300 mg but not ≤75 mg at 2 h after surgery (Mishriky et al. 2015). While most of the studies in the meta-analysis involve abdominal hysterectomy and cholecystectomy, none were in CRS patients. Of note, there is a substantial cost difference at present between gabapentin and pregabalin. As such, if a gabapentinioid is to be considered as one component in an ERP, we recommend use of gabapentin as a first line agent unless the patient was prescribed pregabalin for a chronic pain condition prior to surgery or if gabapentin use resulted in significant sedation.

Intravenous (IV) lidocaine infusion is indicated as part of a multimodal analgesic approach for visceral surgery when other local anesthetic approaches such as regional analgesia are not possible. In open and laparoscopic abdominal surgery, IV lidocaine infusions have been shown to result in significant reduction in postoperative pain intensity at rest and with cough and movement and opioid consumption for up to 48 h postoperatively, as well as being associated with earlier return of bowel function allowing for earlier recovery and shorter length of stay (Marret et al. 2008; Vigneault et al. 2011). Lidocaine infusions are contraindicated in patients with cardiovascular instability and concomitant use of alpha agonists or beta-blockers and in patients with allergies to other amide local anesthetics (bupivacaine). Side effects are more pronounced in patients with liver dysfunction, pulmonary diseases when the predominant problem is carbon dioxide retention, and congestive heart failure. Lidocaine is typically administered as a bolus (100–150 mg or 1.5–2.0 mg/kg) followed by an infusion of 1 to 3 mg/kg/h through the end of surgery. Several meta-analyses suggest that perioperative administration of IV lidocaine is associated with a decrease in postoperative pain and opioid consumption and possibly faster return of bowel function and decreased length of hospital stay (Marret et al. 2008; Vigneault et al. 2011; Khan et al. 2016; McCarthy et al. 2010).

N -methyl- d -aspartate antagonists

Ketamine. Perioperative inhibition of N-methyl-d-aspartate (NMDA) receptors with clinically available NMDA antagonists such as ketamine may be associated with improved perioperative pain and decreased opioid use (Wang et al. 2016; Ding et al. 2014; Dahmani et al. 2011; Bell et al. 2006). Perioperative ketamine, including boluses as well as intraoperative and postoperative low-dose infusions for up to 48 h, has been shown to result in significant reductions in pain, opioid consumption, and PONV with no significant side effect profile (Zakine et al. 2008; Laskowski et al. 2011; Sami Mebazaa et al. 2008). The intraoperative boluses ranged from 0.15 to 1 mg/kg, and perioperative infusions ranged from 1 to 5 mcg/kg/min, with a postoperative infusion rate of 2 mcg/kg/min. Ketamine has also been shown to be of particular benefit in patients on chronic opioid, but this has not been specifically tested in chronic pain patients undergoing CRS (Loftus et al. 2010).

Magnesium. Systemic infusions of perioperative magnesium may reduce postoperative pain and opioid consumption (De Oliveira et al. 2013; Guo et al. 2015b; Murphy et al. 2013). The optimal dosing is uncertain as some studies include both a bolus followed by an infusion whereas others only utilize an infusion without a loading bolus. Typical boluses are 30–50 mg/kg, and the infusion rates range from 4 to 15 mg/kg/h. None of the studies in a systematic review reported clinical toxicity related to toxic serum levels of magnesium (De Oliveira et al. 2013).

Glucocorticoid steroids may have analgesic properties possibly related to anti-inflammatory properties and should be considered as part of a multimodal perioperative pain regimen. Several meta-analyses examined perioperative dexamethasone and indicated that patients who received dexamethasone (4–10 mg or >0.1 mg/kg) had lower pain scores, used less opioids, and required less rescue analgesia (Waldron et al. 2013; Allen et al. 2012; De Oliveira et al. 2011). The concern for significant hyperglycemia (>180 mg/dL) has not been confirmed, even in bariatric patients receiving these doses of dexamethasone (Hans et al. 2006).

Alpha-2 agonists

A Cochrane review of dexmedetomidine infusions for pain found reduced opioid consumption but no significant difference in pain scores compared to placebo, and there was more hypotension in the dexmedetomidine group (Jessen Lundorf et al. 2016). Dexmedetomidine has been added both perineurally to nerve blocks and intravenously to prolong nerve block (Abdallah et al. 2016; Das et al. 2016). There is limited data, but dexmedetomidine does appear to prolong nerve blocks. However, the extended duration is not as long as that provided by perineural dexamethasone provides. In a recent extensive review of perioperative alpha agonists, dexmedetomidine and clonidine were compared (Blaudszun et al. 2012). Similar to dexmedetomidine, clonidine can also reduce opioid consumption. In addition, both alpha agonists appear to have a weak antiemetic effect, but as expected, both drugs had adverse effects on hemodynamics. In summary, both dexmedetomidine and clonidine when administered perioperatively can reduce morphine consumption up to 24 h and to a similar extent as acetaminophen (paracetamol), but not as much as other NSAIDs. Both clonidine and dexmedetomidine have other side effects such as sedation and hypotension that have to be considered (Garg et al. 2014). Typical doses of clonidine range from 1 to 5 mcg/kg PO, IV, or perineurally and for dexmedetomidine from 0.5 mcg/kg IV bolus followed by an infusion of 0.2–0.7 mcg/kg/h or 0.5 mcg/kg perineurally.

If oral acetaminophen (paracetamol) is administered preoperatively as part of a multimodal analgesia pathway, there is typically no need to administer it again until the next scheduled dose. If the next scheduled dose is possible orally, then administration should be oral. However, if the preoperative dose has been missed or oral administration is not possible when the next dose is due, then intravenous acetaminophen (paracetamol) can be administered. Currently, there is no evidence that intravenous acetaminophen (paracetamol) is superior to oral formulations as an analgesic (Jibril et al. 2015; Fenlon et al. 2013). If a patient is not able to take oral medications, then intravenous acetaminophen (paracetamol) has been shown to be an effective opioid-sparing analgesic compared to placebo (O’Neal 2013; Smith 2011).

Non-steroidal anti-inflammatory drugs

As noted above, NSAIDs should be prescribed orally as part of routine preoperative medications for colorectal surgery. However, if a preoperative dose is not given for concerns of bleeding, parenteral options are available. Ketorolac has been used extensively, and newer preparations including intravenous ibuprofen and diclofenac are now available (De Oliveira et al. 2012; Kroll 2012). These medications can improve pain scores, reduce opioid requirements, and reduce opioid-related side effects as discussed in the section above. The optimal dose and timing of each of these medications is not yet known. There is currently no evidence of a better risk benefit profile for one intravenous NSAID over another.

Overall, there is limited evidence for the use of tramadol in colorectal surgery. However, one study in particular compared postoperative pain control with an opioid IV-PCA to IV tramadol and found that the patients in the non-PCA (tramadol) group needed less rescue analgesia and none of them needed to have an IV-PCA started (Choi et al. 2015). Additionally, another study included scheduled tramadol as part of an ERP (Lloyd et al. 2010). While this was only one component of the ERP, overall pain scores and other outcomes were improved. If tramadol is to be used, one study would suggest caution with its use in patients undergoing major abdominal surgery who are over 75 years, ASA 3 or 4, and have impaired mobility or frailty, as use in this setting was associated with delirium (Brouquet et al. 2010). Based on the evidence that exists in colorectal surgery, we recommend considering tramadol as an analgesic adjunct as one part of an ERP.

The primary objective of the practitioner should be to minimize the use of opioids wherever possible. Under general anesthesia, patients are unconscious and do not perceive pain. Although there is no pain perception, surgery-induced activation of nociceptive reflex arcs has hemodynamic consequences. Using opioids to reduce the hemodynamic changes should be resisted as much as possible, as hyperalgesia can develop acutely and opioid-related side effects can occur at very low doses (Hayhurst and Durieux 2016; Rathmell et al. 2006). Control of blood pressure and heart rate at intubation and during surgery by beta-blockers, calcium channel blockers, or lidocaine should be encouraged rather than using opioids without a specific indication other than intraoperative tachycardia or hypertension. As noted above, replacing opioids with any non-opioid analgesic or esmolol intraoperatively leads to less opioid being required in the PACU and often out to 24 h postoperatively in some laparoscopic surgery (Lee and Lee 2010; Collard et al. 2007; Pranevicius and Pranevicius 2009). If an opioid must be used, we recommend using small doses of short-acting opioids such as alfentanil or fentanyl.