FormalPara Key Points

Intravenous opioids are still the mainstay of postoperative analgesia but have several adverse effects; in comparison, the neuraxial route may offer advantages

Multimodal analgesia is an important component of opioid minimization and enhanced recovery protocols after spinal fusion.

Non-steroidal anti-inflammatory drugs and methadone are valuable adjuncts in a multimodal analgesic strategy after posterior spinal fusion.

Evidence for efficacy of gabapentinoids, low-dose ketamine, intravenous lidocaine, and regional techniques is currently insufficient.

Acute opioid tolerance, opioid-induced hyperalgesia, and chronic postsurgical pain are significant problems after major surgery. Additional research is needed to identify effective preventative strategies.

1 Introduction

Adolescent idiopathic scoliosis (AIS) is the most common spinal deformity in children. Some degree of spinal curvature is present in 1–3% of children 10–16 years of age [1]. Posterior spinal fusion (PSF) accounts for 90% of scoliosis surgery [2] with approximately 38,000 spinal fusion surgeries occurring annually in the United States [3]. Analgesia after PSF is challenging due to extensive dissection, inflammation, and ensuing central and peripheral nerve sensitization. Inadequate analgesia may lead to delays in recovery goals such as oral intake and ambulation, causing patient/family dissatisfaction and increased risk of chronic postsurgical pain (CPSP) [4,5,6].

Traditionally, opioids have been the mainstay of analgesia after scoliosis surgery, despite significant adverse effects. Minor adverse effects are common, including vomiting (40%), pruritus (20%), and constipation (15–90%) [7]. Severe adverse events such as respiratory depression are much less frequent (0.0013%) [7]. It is believed that opioid adverse effects are related to the total dose consumed. In addition to adverse effects, development of acute opioid tolerance and opioid-induced hyperalgesia may complicate and delay recovery and discharge. With a goal of decreasing opioid consumption, multimodal analgesia (MMA) has been adopted for pain management. MMA combines different modes/classes of analgesics to treat pain and supports enhanced recovery after surgery (ERAS) [8]. However, there are adverse effects to different medications used for MMA; for example, high-dose ketorolac and non-steroidal anti-inflammatory drugs pose potentially increased risk for non-fusion and bleeding. Therefore, it is necessary to evaluate risks and benefits of the different adjunct medications and safe dosing factors before deciding on the optimal MMA protocol.

This review article aims to summarize currently available evidence and knowledge gaps regarding postoperative analgesia after pediatric posterior spine fusion for adolescent idiopathic scoliosis, including multimodal analgesia and enhanced recovery protocols. With growing literature and several multimodal strategies being used and evaluated, we anticipate this timely topic will aid development of hospital-specific protocols and provide direction for future research.

2 Multimodal Analgesia and Enhanced Recovery After Surgery Protocols

Since their introduction in the 1990s, MMA [9] and ERAS [10] have been widely recognized and promoted by the medical community at large. ERAS is a multidisciplinary perioperative care model that aims to hasten the recovery of patients undergoing surgery without increasing complications or diminishing patient satisfaction. Many surgical subspecialties have developed ERAS protocols to accelerate discharge and improve perioperative care. Similarly, ERAS protocols for AIS patients undergoing PSF have been developed to decrease length of stay (LOS) and cost of care without increasing complication rates.

MMA is a key component of ERAS protocols. It consists of the combined administration of different medications targeting different mechanisms for providing analgesia [11]. This could be preemptive (administered before surgical insult) or preventive (where the timing of administration is not critical). Preventive analgesia is a wider concept and aims to minimize noxious sensations arising intra- or postoperatively. The goal of MMA is to decrease opioid requirements and opioid-related adverse effects, and combined with ERAS, the overall goals are to achieve accelerated recovery goals, including transition from epidural or patient-controlled analgesia (PCA) to enteral analgesia, and ambulation. Since current consensus indicates that use of preemptive analgesia does not result in consistent clinical benefits after surgery [12], we do not classify modalities as preemptive or preventive, but describe when they are used. Regardless of the timing of the intervention (preoperative or intraoperative or postoperative), we will discuss the influence on postoperative analgesia for each analgesic intervention.

In a quality improvement project, Muhly et al. standardized a rapid recovery pathway for AIS patients undergoing PSF. While receiving a standardized MMA regimen, patients were transitioned off PCA to enteral analgesia with intravenous (IV) intermittent opioids for breakthrough pain and ambulated on the first postoperative day (POD). After ERAS protocol initiation, investigators were able to accelerate functional recovery, decrease pain scores, and decrease LOS from 5.7 to 4 days [13]. Most protocols incorporate preoperative education, multimodal analgesia, early mobilization, early transition to oral medications, prompt discontinuation of drains/catheters, and involve multidisciplinary collaborations [14,15,16,17,18]. Salient features of the protocols and outcomes are described in Table 1. Some of the studies evaluated pain outcomes and readmission rates. Fletcher et al. also compared an accelerated pathway at one institution (N = 279) with a standard discharge pathway at a separate institution (N = 86), demonstrating that prioritizing early diet, frequent mobilization with physical therapy, and early removal of drains and urinary catheters resulted in shorter LOS (2.9 vs 4.3 days, p < 0.0001) and no significant increase in complications [19].

Table 1 Multimodal analgesia and enhanced recovery after surgery: protocols from literature

According to a national database study (N = 7349), there is wide variation in adjuvant therapies used and insufficient evidence for the effectiveness of several of the adjuvants [20]. Multimodal therapy may result in additive or synergistic effects and more effective pain relief compared with single-modality interventions [21]. Multimodal analgesia can potentially minimize adverse events by reducing opioid requirements as well [22, 23]. It is recommended that clinicians offer multimodal analgesia for treating postoperative pain in both children and adults. Providers must recognize that multiple potential combinations are possible and appropriate, depending on the type of surgery, patient preference, and individual factors [21]. We present currently available evidence for the different components of MMA regimens after surgeries for AIS, including opioids and opioid adjuvants.

In addition, recent advances in minimally invasive scoliosis (MIS) surgery may suggest need for future refinement of pain management protocols. Surgical techniques used are direct or extreme lateral interbody fusion, axial lumbar interbody fusion, and transforaminal lumbar interbody fusion [24]. However, there is insufficient literature regarding pain trajectories after MIS surgeries. One retrospective study did not find any differences in PCA opioid use or pain scores on comparing seven patients who underwent MIS with 15 adolescents who underwent standard open surgery [25], while another one found higher pain scores in the MIS group versus PSF [26]. Hence, it is too early to assess if any modifications will be needed. It seems more likely that MMA will be useful for MIS procedures too.

3 Opioids

Although there is a push for opioid-reduced and opioid-free anesthesia, opioids continue to be used for intraoperative and postoperative analgesia [27]. Commonly used opioids for pain management in pediatric spine fusion protocols are discussed below and classified by route of administration.

3.1 Intravenous Opioids

Both ultrashort (remifentanil) and long-acting (methadone) opioids are often used intraoperatively during spine fusion with an impact on postoperative analgesia and pain outcomes. Discussion of these opioids will be followed by commonly used postoperative use of opioids using patient-controlled analgesia modalities.

3.1.1 Methadone

Methadone is a μ-opioid agonist, a potent N-methyl-d-aspartate (NMDA) receptor antagonist [28] and serotonin/norepinephrine reuptake inhibitor, thus it potentially provides mood as well as analgesic benefits [29, 30]. Several studies in adults have shown a beneficial effect of methadone doses of 0.1–0.3 mg/kg or 10–20 mg doses prior to incision for its effects on pain scores and opioid consumption after complex spine surgery [31,32,33]. However, there are only two studies evaluating use of methadone in children undergoing spine fusion, both showing decreased opioid (hydromorphone) consumption postoperatively (Table 2). One may argue that methadone is also an opioid and total morphine equivalents need to be compared. Methadone is unique given its long elimination half-life (24–36 h) but quick onset (8 min) compared with other commonly used IV opioids [30]. Methadone pharmacokinetic studies have shown a linear increase in blood concentrations with doses; a pharmacokinetic study in adolescents undergoing spine fusion administered methadone 0.25 mg/kg IV before surgical incision found mean concentrations of 58 μg/L by the first hour, which was previously deemed the minimum effective analgesic concentration [34]. The authors in the aforementioned pharmacokinetic study recommend following the bolus (0.25 mg/kg) with an infusion (0.1–0.15 mg/kg/h for 4 h) during spinal surgery to ensure adequate plasma concentrations for 24 h. More pediatric studies are needed to establish appropriate safety parameters for methadone dosing. In adults, the incidence of respiratory depression was not different from controls, except in elderly patients, and although QTc on electrocardiogram was prolonged in 58.8% of patients after surgery, it did not lead to arrhythmias [35]. Additionally, in cases of respiratory depression, it is recommended to consider a naloxone infusion and not a one-time reversal dose due to methadone’s long half-life [36].

Table 2 Studies evaluating use of methadone in children undergoing spine fusion

3.1.2 Remifentanil

Remifentanil, on the other hand, has an extremely short half-life of 1 min and an elimination half-life of 0.5 h. Hence, it is often used in conjunction with propofol for total intravenous anesthesia intraoperatively during spine fusion to facilitate intraoperative neuromonitoring and enable ‘wake up’ tests when necessary. The intraoperative use of remifentanil may have ramifications for postoperative analgesia. There has been concern that remifentanil could theoretically trigger opioid-induced hyperalgesia (OIH). Its use has been associated with higher morphine consumption during the first 24 h after scoliosis surgery [39]. However, findings of acute opioid tolerance and OIH after intraoperative remifentanil have been inconsistent, as other clinical studies have suggested no difference [40,41,42,43]. Co-administration of anesthetics such as propofol, ketamine, and nitrous oxide have been suggested to attenuate development of OIH [44,45,46].

A retrospective chart review examined the influence of intraoperative remifentanil (N = 37) and fentanyl infusions (N = 25) on postoperative morphine equivalents use over 48 h after spine fusion. To their surprise, they found higher postoperative opioid usage in the fentanyl group [47].

Hence, although the effect of intraoperative remifentanil may theoretically have negative effects on postoperative analgesia, current evidence does not condone its use.

3.1.3 Patient-Controlled Analgesia

For severe acute postoperative pain, IV opioid administration via PCA continues to be widely used [48]. Most studies use weight-appropriate initial loading doses followed by morphine 20 μg/kg or hydromorphone 2–4 μg/kg PCA demand doses with lockout intervals of 7–10 min (± basal infusions of 10 or 2 μg/kg/h, respectively) [39, 49,50,51]. Intermittent as-needed boluses of morphine (0.05–0.1 mg/kg) or hydromorphone (10–20 μg/kg) or fentanyl (0.5–1 μg/kg) are often made available every 4 h for breakthrough severe pain. In addition, medications for counteracting opioid adverse effects are usually available through pre-designed order sets on most electronic prescribing platforms. These medications include nalbuphine, an opioid agonist–antagonist (0.05 mg/kg) for opioid-induced itching/urinary retention, ondansetron (0.1 mg/kg) for nausea/vomiting, and naloxone, an opioid antagonist (1–2 μg/kg for over-sedation; 10–20 μg/kg for respiratory depression, and sometimes 0.25 μg/kg/h IV infusion for itching).

Most cited advantages of PCA include patients’ control over their own analgesia with improved relief, satisfaction, and psychological well-being [52]. While adult literature does not support use of basal infusions due to increased risk of respiratory depression [53], pediatric studies demonstrated that basal infusions had an inconsistent effect on opioid consumption [54,55,56,57] and adverse effects [54,55,56,57,58,59]. A meta-analysis of pediatric PCA use found that basal rates alone have no effect on outcomes, including pain scores, opioid consumption, and adverse effects [60]. Because of this conflicting and insufficient evidence, the Society of Pediatric Anesthesia recommends basal infusions only be used in select patients based on clinical situation, pain severity, and risk factors [61]. Despite widespread use, adverse events related to opioids remain a concern, and minimizing duration of PCA use is a priority [62]. A recent study by Fletcher et al. describes best practice guidelines for discontinuing PCA on postoperative day one and transitioning to oral opioids to facilitate quicker recovery [14].

3.2 Oral Opioids

Oral opioids are initiated when children tolerate oral intake after surgery. With ERAS protocols, there is a push for starting liquids and progressing to regular diet as early as possible. In general, the goal would be to transition from IV to oral opioids on POD1. While there are no studies comparing different oral opioids used after spine fusion, commonly used opioids include oxycodone 0.1 mg/kg/dose, hydromorphone 0.03–0.08 mg/kg/dose and hydrocodone 0.1–0.2 mg/kg/dose every 4 h as needed for pain [13]. We refer readers to the implications of pharmacogenetics for use of oral opioids such as codeine and tramadol, as ultrarapid metabolizers for the CYP2D6 enzyme may experience higher adverse effects with these oral opioids, which carry black box warnings by the FDA [63]. Of note, patients are often discharged home with prescriptions for oral opioids. Children who underwent spine fusion were among those prescribed on average 44.13 more doses than children who underwent other surgeries (95% CI 34.72–53.54; p < 0.001) [64]. While one study showed that patients were discharged home with an average of 61 pills (SD 14), of which 90.1% were utilized [65], other studies show that patients were dispensed 113 pills (80–115), of which only 39 pills (20–80) were actually used [64]. Given the risk for higher opioid use with availability and leftover pills, which can subsequently be diverted, these data suggest closer attention is warranted to avoid overprescribing and underline the need for education for proper disposal of leftover opioids.

3.3 Epidural Opioids

Epidural analgesia can be delivered as either continuous infusions, patient-controlled approaches with demand bolus and lockout interval (PCEA) or programmed intermittent boluses (PIB). The modes of delivery used in the studies are described in Table 3, with most of the studies using continuous infusions. Although there are studies showing superior analgesic efficacy of PCEA + PIB over continuous/PCEA in laboring parturients [66], there are no comparative studies in children undergoing scoliosis surgery. Most studies evaluating efficacy and safety of epidural analgesia after spine fusion commonly include mixtures of opioids and local anesthetic solutions for epidural use (Table 3). Only one retrospective study (N = 56) evaluated the use of an opioid-only regimen with hydromorphone 5 μg/kg (maximum 200 μg) + fentanyl 1 μg/kg (maximum 50 μg) through an epidural catheter before wound closure, followed by epidural infusion of hydromorphone 5 μg/mL at 12–16 mL/h with 2-mL boluses permitted every 30 min [67]. Though not a comparative study, the authors concluded that narcotic-only epidural infusion was a safe and effective mode of analgesia. Some studies use single while others have used double catheters; one study compared their use and found that double epidural catheter had a modest benefit over single catheter for analgesia [68]. One of the studies used epidural analgesia when the epidural space was violated during surgery and showed no negative consequences [69]. Epidural opioid doses range from morphine 30–50 μg/kg, hydromorphone 5–20 μg/kg, and fentanyl 1 μg/kg with or without local anesthetic followed by infusions (of local anesthetic + opioid) via catheter. A Cochrane review of 11 trials (7 trials analyzed, 249 participants) found that there was little evidence that epidural local anesthetic infusion alone accelerates return of gastrointestinal function, time to first mobilization, or hospital discharge [68, 70], and studies have demonstrated epidural failure rates as wide as 8–37% [49, 71]. Yet, the Cochrane Review found moderate- and low-quality evidence that epidural analgesia may have a small advantage in pain reduction in the first 72 h after surgery compared with systemic analgesia with no difference in complication rates (vomiting, respiratory depression, wound infection, epidural abscess, etc.). Hence, epidural analgesia is a potentially safe and effective postoperative modality after spine fusion, and has been used in conjunction with MMA protocols [18]. The comparative efficacy of modes of epidural delivery for spine surgery are yet to be determined.

Table 3 Studies evaluating epidural for postoperative analgesia after spine fusion for idiopathic scoliosis

3.4 Intraoperative Intrathecal Opioids

Intrathecal opioids, mainly morphine—typically after induction of anesthesia and prior to incision—have been evaluated in doses of 2–20 μg/kg for postoperative analgesia after spine surgery [80,81,82,83,84,85]. Studies demonstrated that this technique reduced intraoperative and postoperative opioid consumption and decreased pain scores (Table 4). The analgesic effect lasts for at least 12 h [86] to 18.8 h [80,81,82,83]. There is some compelling evidence that intrathecal opioids may significantly decrease intraoperative blood loss, though the mechanism of the blood-sparing effect remains unclear [80,81,82,83]. Some hypothesize that the diminished blood loss may be due to lower mean arterial pressures. Yet, other studies have demonstrated no difference in blood pressures [82]. Effects of sex and race on efficacy and adverse effects of this recommended dose in adolescents undergoing scoliosis surgery (N = 287) were evaluated by Son-Hing et al. [84]. They found that analgesic efficacy was similar in females/males (F/M) and White/African American (W/AA) groups. While there was no statistical difference in the incidences of nausea/vomiting and pruritis between females and males (31.7%/25.5%), there was a statistically significant difference between White and African American groups (34.4%/17.5%). However, it is important to note that there was no statistical difference in the incidences of respiratory depression for different sexes (F/M 4.1%/6.4%) or different ethnicities (W/AA 4%/6.3%). Thus, while most studies demonstrated no differences in major adverse effects, a study evaluating different doses concluded that higher doses of intrathecal morphine (≥ 20 μg/kg) may be associated with greater risk of significant respiratory depression and 9–19 μg/kg (mean 14 μg/kg) doses are safe and effective [83].

Table 4 Studies evaluating intrathecal opioids for postoperative analgesia after spine fusion

3.5 Comparison of Opioid Routes for Postoperative Analgesia

One retrospective study compared the efficacy of PCA morphine with single preoperative intrathecal morphine injection (7 μg/kg) and PCA (IT/PCA), and epidural catheter infusion (a bolus dose of hydromorphone [10–20 g/kg] followed by a continuous infusion of hydromorphone [20 g/mL] and bupivacaine 0.1% at an initial rate of 0.1–0.2 mL/kg/h) without PCA (EPI) for postoperative pain control after PSF [87]. They found that while both EPI and IT groups had superior pain control compared with PCA, the EPI group had longer postoperative analgesia due to the infusion, while IT analgesia lasted about 24 h. The difference in analgesia duration allowed for quicker return to diet. Another double-blinded randomized, controlled trial (RCT) compared use of IT morphine (7.5 μg/kg) (N = 37) with extended-release epidural morphine (EREM) (150 μg/kg) (N = 31) and found no significant differences for 48-h opioid consumption but lower pain scores over 28–36 h post-surgery in the EREM group, which also had lower incidence of pruritis [88]. In summary, regional opioids are superior to PCA for pain control and recovery; IT and epidural opioids are comparable in analgesia for about 24 h, beyond which opioid supplementation will be needed for the IT opioid regimens.

4 Acetaminophen

Acetaminophen (APAP) is a widely used, centrally acting analgesic used as an opioid adjuvant for postoperative pain. Its central analgesic effect is mediated through activation of descending serotonergic pathways [89]. Proposed primary mechanisms of action include COX-3 enzyme inhibition and acting as a cannabinoid agonist and NMDA antagonist in the spinal cord [90]. It has demonstrated opioid-sparing potential across numerous studies [91]. Oral, intravenous, and rectal formulations are available [92]. Since IV APAP was approved for use in the US in 2010, it became an important component of perioperative multimodal analgesia. In a placebo-controlled RCT in 36 adolescents, patients in the IV APAP experienced fewer hours (8.7%) in severe pain (visual analog scale pain score ≥ 6) versus the placebo group (17.8%) within 24 h after surgery [93]. However, there was no difference in cumulative oxycodone dose required during the first 24 postoperative hours. It is notable that the study used higher APAP daily and individual doses (30 vs 15 mg/kg/dose and 90 vs 75 mg/kg/day). The recommended maximum daily dose of APAP is 75 mg/kg/day with hepatotoxicity occurring at 150 mg/kg/day. In a prospective, observational study in adolescents undergoing PSF, Olbrecht and colleagues demonstrated that conventional doses of IV APAP significantly decreased opioid consumption on postoperative days 1 and 2, decreased LOS by 0.6 days and accelerated oral intake by approximately 1 day by mediation of opioid-related adverse effects [94]. However, with the cost of IV APAP increasing in 2014 (from US$14.60 to US$35.05 for a 1-gm bottle), its cost effectiveness over the much cheaper alternative oral acetaminophen has been called into question. Recent adult studies comparing IV and oral APAP for patients demonstrated no significant difference for pain scores and opioid consumption at 12, 24, and 48 h, incidence of nausea/vomiting, or LOS [95], with oral APAP showing more beneficial outcomes [96]. There are currently no prospective RCTs demonstrating clinically significant benefits of IV over oral APAP for pediatric spine fusion or dose-dependent effectiveness for IV APAP. So, while IV APAP may still be a useful adjunct for use in patients who are unable to tolerate oral pain medications, transition to the oral form as early as possible postoperatively may be the prudent option.

5 Nonsteroidal Anti-Inflammatory Drugs

Nonsteroidal anti-inflammatory drugs (NSAIDs) are analgesics that inhibit cyclooxygenase (COX), which decreases the production of prostaglandins and, in turn, the inflammatory response. The most commonly studied NSAID in children is ketorolac. In a national cohort of children undergoing scoliosis surgery (N = 7349), ketorolac was independently associated with significantly lower odds of prolonged LOS and prolonged duration of IV opioid use [20]. Munro et al. [50] showed that conventional use of ketorolac (0.5 mg/kg/dose IV every 6 h for a total of six doses) reduced pain scores, morphine consumption, and improved activity on PODs 1 and 2.

While NSAIDs decrease inflammation and pain via inhibition of COX and prostaglandin synthesis, there has been concern regarding their effects on platelet function and bone formation/healing—critical processes for successful spinal fusion. Concern about bone formation/healing has been fueled by delayed bone healing in animal models and studies performed in certain adult populations [97]. A study in a rabbit arthrodesis model comparing effects of large doses of celecoxib, indomethacin, and saline solution found that while indomethacin significantly inhibited the rate of spinal fusion, celecoxib did not. The authors postulated this was because bone healing was inhibited by COX-1 [98]. A systematic review and meta-analysis evaluating effect of postoperative ketorolac administration (dosage, duration of use) on pseudarthrosis in adults following PSF [99] concluded ketorolac (administered for > 2 days and/or at doses ≥ 120 mg/day) was associated with pseudarthrosis. Sucato and colleagues compared AIS patients who received ketorolac and those who did not after PSF. While overall pseudarthrosis rate was 2.5%, ketorolac did not increase probability of pseudarthrosis [100]. Another retrospective study in 434 adults undergoing spine surgery reported the incidence of non-fusion rates after use of perioperative ketorolac (< 110 mg/day) was 6%, high-dose ketorolac (120–240 mg/day) was 29%, celecoxib (200–600 mg/day) was 8.3%, and rofecoxib (50 mg/day) was 7.3% compared with 8.5% in the group that did not receive NSAIDs [101]. Therefore, it appears that short-term exposure to NSAIDs including low-dose ketorolac is not associated with nonunion after spinal fusion [102].

Postoperative ketorolac administration was not associated with bleeding-related adverse events, such as increased likelihood of transfusion or increased reoperation rates [103]. A retrospective study of 208 children undergoing spine surgery found postoperative ketorolac use did not significantly increase complications, including transfusion and reoperation [103]. Ketorolac may cause prolonged bleeding time, but large-scale prospective RCTs and meta-analyses failed to establish an association with increased perioperative blood loss [104].

Given recent evidence that a low dose (10 mg) of IV ketorolac is as effective as higher doses, it is recommended to cap maximum doses of ketorolac at 10 mg for postoperative analgesia after spine fusion [105].

6 Gabapentinoids

Gabapentin and pregabalin are structural analogs of gamma-aminobutyric acid (GABA). Gabapentinoids modulate excitatory neurotransmitter release by binding to voltage-gated calcium channels, and may exert their effect by decreasing spontaneous sensory nerve firing [106].

There are two RCTs evaluating pregabalin within MMA for pediatric scoliosis, with opposite findings [107, 108] (Table 5). Findings vary for gabapentin use depending on timing of use. Only one study evaluated preoperative-only gabapentin and found no analgesic advantage to its use before surgery [109]. Studies where gabapentin was continued for 5 days postoperatively [110], until discharge, or one dose on POD1 in addition to a preoperative dose [111, 112] reported positive findings with respect to opioid consumption. One of the retrospective studies mentioned above also included a group receiving a combination of gabapentin and transdermal clonidine 0.5 mg/day [111]. They found that the addition of clonidine + gabapentin further decreased PCA usage and had faster time to ambulation compared with the gabapentin group. None of the above studies reported an increase in opioid adverse effects from the addition of gabapentin. However, sedation, dizziness, and visual disturbances are reported adverse effects, which have prompted a call for moderating the use of gabapentinoids in general [113]. In fact, current American Pain Society and European Society of Regional Anaesthesia and Pain Therapy guidelines offer conflicting recommendations for the use of gabapentinoids in the perioperative period [21, 114]. It is important to point out that a recent meta-analysis (281 trials, N = 24,682) did not support the preoperative use of gabapentinoids as they did not find a clinically significant analgesic effect. Although no pediatric studies were included, they did exploratory sub-group analyses by surgery, which showed that gabapentinoids may be favored (summary estimate -8 [-12, -5]) for spine surgeries. In summary, there is insufficient evidence to currently recommend the use of preoperative gabapentinoids for opioid-sparing effects after surgery.

Table 5 Studies evaluating gabapentinoids for pediatric scoliosis

7 Ketamine

Ketamine is an NMDA receptor antagonist that also activates μ-, δ-, and κ-opioid receptors, γ-aminobutyric acid-mediated central nervous system inhibition, and monoaminergic inhibitory pathways [115]. Tissue trauma during major surgery causes central nervous system sensitization by release of excitatory amino acids working through the NMDA receptor [116, 117]. Thus, many have theorized that ketamine may decrease central sensitization and peri-incisional hyperalgesia [118,119,120,121,122].

In children undergoing spine fusion, there have only been five trials evaluating ketamine as an adjunct analgesic (Table 6). Two of these trials utilized low-dose ketamine infusions only during the intraoperative phase of care [123, 124]. However, only one of those two studies found a significant difference in postoperative opioid consumption. The other three RCTs examined the benefits of low-dose ketamine infusions during both the intraoperative and postoperative phases of care [51, 125, 126]. Only one of those three studies found a significant difference in postoperative opioid consumption, and the authors of the one positive study noted that the difference might not be clinically relevant. All these studies had < 50 subjects and followed different protocols with respect to non-opioid analgesia, intraoperative inhalation versus total intravenous anesthesia, etc., which could influence results. While they were not powered to do so, none of the studies demonstrated a difference in time to oral intake, time to ambulation, LOS, or CPSP. Lastly and most importantly, the optimal regimen for ketamine infusions for children and adolescents remains unknown.

Table 6 Studies evaluating ketamine for AIS PSF in pediatric subjects: randomized controlled trials

Thus, we will cite some relevant adult literature to support our conclusions. A meta-analysis evaluating the benefits of perioperative low-dose ketamine for postoperative analgesia in adult spine surgery patients showed that ketamine significantly reduced opioid consumption and pain scores for 24 h. However, differences in dosing, infusion protocols (e.g., continuous infusions vs IV PCA), and patient populations (e.g., inclusion or exclusion of patients with chronic pain or opioid tolerance) were theorized to impact the results after the first 24 h [127]. Based on best available evidence for adults, the American Society of Regional Anesthesia and Pain Medicine (ASRA), American Society of Anesthesiologists, and the American Academy of Pain Medicine’s recent guideline for the use of ketamine in acute pain management recommends a low-dose IV ketamine bolus with a maximum dose of 0.35 mg/kg followed by a subanesthetic infusion of 0.15–1 mg/kg/h (titrated to the lowest effective dose) [128]. Higher doses of ketamine may be required in pediatric populations to maintain a steady-state concentration due to age-related pharmacokinetics [129]. Analgesic concentrations are deemed to be 70–160 ng/mL [130]. While dissociation and other psychoactive effects are adverse effects of ketamine infusions, little to no psychoactive effects were observed at the dose of 0.1 mg/kg IV (resulting in ∼ 25–50 ng/mL concentrations) in adults [131].

Although current evidence for the routine use of ketamine to reduce pain and opioid consumption for children and adolescents undergoing AIS surgery is not strong, there may be a role for low-dose ketamine postoperatively in patients who are opioid tolerant, have chronic pain, or are at high risk from adverse events due to high-dose opioids (ASRA guideline) [128]. Future studies are needed to determine effective and safe dosing regimens and duration of ketamine infusions to decrease opioid consumption, pain intensity, recovery, and LOS as well as CPSP and chronic opioid use after AIS surgeries.

8 Local Anesthetics

Local anesthetics act by blocking voltage-gated sodium channels of afferent neurons. For AIS surgery, local anesthetics have primarily been used as part of epidural analgesia regimens, usually together with opioids, as discussed in the epidural opioid section. However, recently, other modes of local anesthetic delivery have been explored, including IV continuous infusion, local wound infiltration, and, most recently, erector spinae blocks.

8.1 IV Lidocaine Infusions

The theoretical analgesic effects of lidocaine are diverse with peripheral and central actions including altered conduction in the dorsal horn and dorsal root ganglion, decreased firing of sodium channels after increased expression of sodium channels with peripheral nerve injury, reduced neurogenic inflammation at the site of injury, anti-hyperalgesia with NMDA receptor inhibition, and glycinergic system modulation. A study compared lidocaine versus control groups for proinflammatory mediators before, immediately after, 6 h after, and 12–15 h after spine surgery in children [132]. They observed reduced pain intensity until 6 h after surgery as well as negative correlations between pro-inflammatory mediators (neuron growth factor [NGF], high mobility group box 1 [HMGB1], interleukin 6 [IL-6]) and lidocaine concentrations after surgery.

The utility of perioperative lidocaine for children and adolescents undergoing major spine surgery has not been studied extensively. Batko et al. evaluated the efficacy of an addition of lidocaine to a standardized multimodal analgesia regimen (preoperative gabapentin; intraoperative acetaminophen, dexamethasone, sevoflurane, metimazole, and morphine; postoperative morphine IV-PCA, scheduled non-opioids—acetaminophen, metimazole, and gabapentin) [133]. Lidocaine was administered as a 1.5-mg/kg bolus over 30 min prior to incision followed by a 1 mg/kg/h infusion intraoperatively and then postoperatively up to 6 h after surgery finished. The control group received an equal volume and rate of a placebo. They demonstrated decreased morphine consumption at 24 h, 48 h (> 30% reduction), and the entire hospitalization, compared with the control group. Additionally, first oral intake, sitting, and walking were all positively influenced. Also, they conducted 2-month and 4-year follow-ups to determine if the two groups experienced any significant difference in quality of life but found no difference. Another recent retrospective study demonstrated in a small cohort of 50 pediatric patients that IV lidocaine infusions were generally well tolerated [134]. The mean ± SD infusion dose was 15 ± 6.3 μg/kg/min with 24% of infusions associated with adverse effects, primarily neurologic ones, including paresthesias (10%) and visual disturbances (4%).

Due to the paucity of pediatric literature in spine surgery, we mention relevant adult literature regarding perioperative IV lidocaine infusions. The most recent Cochrane review in 2019 included 68 clinical trials (two involving spine surgery) with over 4500 participants and found no significant effect on postoperative pain intensity, opioid consumption, return to bowel function, or postoperative nausea [135]. Since then, there have been two RCTs exploring the use of perioperative lidocaine infusions for adult patients undergoing multilevel spine surgery, both of which found no benefit with regard to pain intensity or opioid consumption [136, 137]. Of these, Dewinter et al. enrolled 70 patients (of which 28 were adolescent idiopathic scoliosis patients undergoing posterior spine fusion) in a prospective, double-blind RCT in which patients received either a lidocaine bolus of 1.5 mg/kg at induction of a standardized total intravenous anesthetic followed by an infusion of 1.5 mg/kg/h intraoperatively and continued until 6 h after arrival in the post-anesthesia care unit, or placebo. There was no difference between groups with respect to pain intensity, opioid consumption at 48 h and 72 h after surgery, incidence of postoperative nausea and vomiting, recovery of bowel function, LOS, or quality of life [136].

In short, while the evidence for perioperative IV lidocaine infusions (with a loading dose of 1.5 mg/kg followed by an infusion of 1–2 mg/kg/h) is compelling, it is not consistent enough to warrant routine use in a multimodal analgesia regimen at this time. Future studies should examine the dosing, duration, pain intensity reduction, and opioid consumption.

8.2 Local Wound Infiltration

Some adult studies show that continuous infusion of local anesthetics via wound catheters with an elastomeric pain pump significantly improves pain, including after spine surgeries in adults [138]. However, the available evidence for the use of wound catheters for PSF in AIS is limited. Two retrospective studies evaluated bupivacaine infusions through bilateral wound catheters [139, 140] (Table 7). Overall, the group of patients that received the wound catheters consumed 28–38% less opioids than the control group over 24 h postoperatively. Both the studies reported no differences in frequency of adverse effects, although possible additional risks include infection, dislodgement, etc. However, there were many uncontrolled variables that depended on the individual care teams, such as the analgesia regimens the patients received. In short, there is at most weak evidence that continuous local anesthetic infusions via wound catheters may lower postoperative opioid requirements but more prospective studies are needed.

Table 7 Studies evaluating bupivacaine infusions through bilateral wound catheters

8.3 Erector Spinae Blocks

A recent case report demonstrated proof-of-concept in two healthy AIS patients that pre-incisional bilateral erector spinae (ES) single shot blocks at two levels (T4 and T10) with 0.25% bupivacaine and epinephrine 5 μg/mL for PSF could enhance a multimodal perioperative anesthesia/analgesia regimen. The MMA regimen also included acetaminophen, dexamethasone, dexmedetomidine, and ketamine infusions. The first patient received minimal to no opioids intraoperatively and importantly they were then successfully transitioned to oral analgesia on POD1. Further research is needed to determine if this proof of concept may be translated to a larger population of AIS patients undergoing PSF [141]. In a letter to the editor, Tsui et al. described use of percutaneously placed ES catheters (bilaterally above incision site with tips above T6 transverse processes) for a T4–T12 PSF [142]. They used 0.5% lidocaine 20 mL boluses through ES catheters every 60 min via an automated pump (increased to 22 mL on POD1). They removed the catheters on POD2. They measured lidocaine concentrations in serum (0.9–1.1 μg/mL). Although the patient experienced peak pain (5/10 pain score) the first night, they conclude that ES catheters may aid mobilization. While these anecdotal reports suggest ES catheters may present a novel analgesic modality for postoperative pain management after spine fusion, optimal local anesthetic (LA) dosages and efficacy are yet to be elucidated.

9 Muscle Relaxants

Since scoliosis is a musculoskeletal condition, correction of the deformity is associated with muscle tightness. Diazepam (benzodiazepine and GABA agonist), methocarbamol (a centrally acting muscle relaxant), baclofen (possibly GABA-B agonist), and tizanidine (α2 adrenergic agonist) are often used to treat muscle spasm after spine fusion (see Table 1 for MMA protocols). Diazepam is often used in doses of 0.05–0.1 mg/kg every 4–6 h as needed IV and methocarbamol in doses of 15 mg/kg (maximum 1000 mg) every 8 h IV and then transitioned to oral formulations at similar doses for diazepam and 500–1000 mg every 8 h for methocarbamol. However, there are not many systematic studies assessing these adjuvants on pain relief. A double-blinded RCT comparing chlorzoxazone, a centrally acting muscle relaxant, demonstrated no immediate analgesic effects compared with placebo in patients experiencing moderate-to-severe acute post-operative pain following spine surgery [143]. In a randomized prospective study of 50 consecutive patients comparing an opioid only (meperidine hydrochloride) with adjunctive use of diazepam/baclofen, the regimens with muscle relaxants successfully relieved postoperative spasm, but did not change pain severity or opioid requirement [144]. It is important to note the adverse effects of this adjunctive group of medications, which include drowsiness and withdrawal after prolonged use [145]. Although there is evidence for their benefits in chronic low back pain and other surgeries including joint surgery and Chiari decompression, further studies are warranted to understand their efficacy after spine fusion [145].

10 Dexmedetomidine

Dexmedetomidine’s action is mediated via postsynaptic α2 adrenergic receptors. It has been shown to have analgesic and opioid-sparing properties, but the literature in pediatric scoliosis is limited. One retrospective study compared use of PCA with opioids ± dexmedetomidine infusion 0.4 μg/kg/h (over 24 h postoperatively) and concluded that it may have opioid-sparing effects, as opioid use increased after discontinuing dexmedetomidine (N = 37) compared with the PCA opioid-only group (N = 94) [146]. Another similar study using fentanyl PCA recommends use of dexmedetomidine (0.25 μg/kg/h with fentanyl 0.5 μg/kg/h to decrease opioid consumption after surgery [147]. In comparison, retrospective chart review of 106 children receiving PCA + dexmedetomidine, and 57 who received PCA opioids only, failed to demonstrate any difference in opioid use on any postoperative day [148]. There are only conference abstracts studying the effect of intraoperative dexmedetomidine on postoperative analgesia following scoliosis surgery. Thus, further research is warranted to study dose, timing, and opioid-sparing effects of dexmedetomidine for spine fusion.

11 Non-Pharmacological Methods

Anxiety and pain catastrophizing enhances pain perception after surgery in children [149, 150]. Several non-pharmacological methods that target anxiety have been shown to decrease post-surgical pain [151]. These include education (setting expectations preoperatively), psychological methods (guided imagery, hypnosis, distraction, cognitive behavioral therapy/counseling, mindfulness), physical methods (cold, heat, massage, acupuncture, transcutaneous electrical nerve stimulation), and distraction (virtual reality, play, videos). Some of these may be helpful both before and after surgery to help children cope with pain [152]. In one RCT, children were taught hypnosis with guided imagery at 1 week before admission, lasting no longer than 30 min (N = 26). Imagery interventions decreased self-reported pain after major surgery (including scoliosis surgery) compared with usual care (N = 26) [153]. However, very few studies evaluate non-pharmacological therapies perioperatively for spine fusion surgeries, and this is an emerging field of research that may hold a lot of promise, to enable ERAS and minimize opioids after surgery.

12 Chronic Postsurgical Pain (CPSP)

CPSP is defined as pain of at least 3–6 months duration, that develops after a surgical procedure, increases in intensity or has different characteristics after the surgical procedure, and significantly affects function. Other conditions like infection and malignancy should be excluded before a diagnosis of CPSP is made [154, 155]. CPSP is a significant clinical as well as socioeconomic problem in children, with a prevalence of ≈ 20% at 12 months after surgery, and important negative behavioral and physical consequences [156]. Spine fusion for AIS has been studied in several pediatric cohorts, with differences in CPSP incidence ranging from 11% to 53.6%, depending on the definition of CPSP used (Table 8).

Table 8 Studies describing chronic post-surgical pain incidence and predisposing factors after scoliosis surgery in children

12.1 Factors Affecting CPSP after Spine Fusion

Occurrence of CPSP after spine fusion in children is multifactorial. Psychosocial, perioperative, and genomic factors have been proposed. In addition, several preventive perioperative measures have been evaluated, mostly in adult cohorts, with conflicting results [165,166,167,168].

12.1.1 Psychosocial Factors and CPSP

The psychosocial factors that have been identified to be associated with risk of CPSP are anxiety sensitivity, self-image perception, pain unpleasantness, and pain catastrophizing [3, 156, 158, 160]. A longitudinal, prospective study identified that membership in a high symptom cluster including higher depression, fatigue, pain interference, catastrophizing, and painDETECT scores, predicted pain interference at 1 year after spine fusion [169]. In addition, parental factors including parent pain catastrophizing as well as anxiety has also been shown to influence child’s risk of CPSP [161, 170]. Understanding psychological risk is crucial to developing interventional treatments preoperatively. A recent systematic review provided preliminary evidence that cognitive behavioral therapy-based psychological interventions reduce CPSP intensity and disability in adults, which will also likely be true for children [151].

12.1.2 Perioperative Factors and CPSP

Preoperative as well as acute postoperative pain, and higher postoperative opioid consumption, have been found to be associated with higher pain trajectories after spine surgery [3, 159, 171]. Similarly, higher surgical duration, but not scoliosis curve or number of vertebral levels to be fused, has been predictive of CPSP [3].

12.1.3 Perioperative Medications and CPSP

While intraoperative remifentanil has been suspected to cause OIH, there are no studies implicating use of remifentanil in CPSP. Interestingly, intraoperative intrathecal morphine was found to predict membership in high opioid use trajectories after spine fusion, which was hypothesized to be due to innate genetic resistance to opioid actions [172]. There is evidence that higher opioid use in the postoperative period may lead to CPSP after spine fusion in adolescents [3] as well as after other surgeries [173, 174]. While this might be a proxy for intense postoperative pain, OIH and acute opioid tolerance (AOT) may also play a role [175, 176]. Thus, minimizing opioid use in the perioperative period using MMA is expected to decrease the incidence of CPSP. However, a meta-analysis evaluating MMA and CPSP in adults showed that available evidence does not support the efficacy of gabapentin, pregabalin, NSAIDs, intravenous steroids, oral NMDA blockers, oral mexiletine, intravenous fentanyl, intravenous lidocaine, oral venlafaxine, or inhaled nitrous oxide for the prevention of CPSP [177]. They did find that that IV ketamine (bolus doses in the range of 0.2–0.75 mg/kg, followed by infusions of 2–7 μg/kg/min) may decrease incidence of CPSP. Recent meta-analysis using limited data supported use of IV lidocaine infusions to prevent CPSP, though the difference in pain intensity was not significantly decreased [178].

Given current evidence, the most promising strategies to prevent CPSP in children undergoing surgery would be preoperative setting of expectations, psychological optimization to help with pain coping, and close monitoring of those who have high risk factors for CPSP. Besides, there are other individual genomic factors that may be involved that are beyond the scope of this manuscript. We refer to the following reviews for further reading on this topic [165,166,167,168].

13 Prolonged Opioid Use After Surgery

Importantly, spine fusion and CPSP may also pose a risk for prolonged opioid use. In fact, one of the surgeries predictive of higher opioid use after hospital discharge was spine fusion, with 25.42 (95% CI 19.16–31.68; p < 0.001) more doses than those who underwent other types of surgery [64]. A search of a large insurance database revealed that prolonged opioid use (receiving new prescriptions for an opioid medication > 6 weeks following the date of surgery, up to 8 months postoperatively) after PSF for AIS was 9.78% [179]. Besides preoperative opioid use (odds ratio, 2.93; p < 0.001), which was the most significant predictor, female sex, obesity, preoperative anxiety, and preoperative muscle relaxer use were also significant risk factors for prolonged postoperative opioid use. Fewer total fusion levels and preoperative anxiolytic and antidepressant use decreased risk for prolonged opioid use after PSF. Thus, use of behavioral non-pharmacological and pharmacological therapies may be useful in decreasing opioid use after surgery. However, we would caution against routine use of pharmacologic agents as there are potential safety concerns with their use. A meta-analysis of antidepressants for acute and chronic postoperative pain evaluated several studies using amitriptyline, bicifadine, desipramine, duloxetine, fluoxetine, fluradoline, tryptophan, and venlafaxine but concluded that the evidence was insufficient to recommend use of these medications routinely in adults [180].

14 Conclusion

The goal of perioperative analgesic regimens is to enhance recovery while minimizing opioid use. While MMA is a critical component of rapid recovery, an optimal regimen for scoliosis surgery has not been established. We present a menu of MMA/ERAS components, suggested dosing regimens, and recovery pathways based on our literature review (Tables 9 and 10). Studies have evaluated other medications including magnesium, dexmedetomidine, dexamethasone, and esmolol as opioid-sparing adjuncts [181,182,– 184]. However, these interventions have not been evaluated systematically in children undergoing spine fusion.

Table 9 Evidence based postoperative enhanced recovery pathway after spine fusion for idiopathic scoliosis
Table 10 Summary of evidence based recommendations for use of multimodal analgesia components

Large-scale, multi-institutional studies are required to establish optimal regimens as spinal fusion is associated with considerable healthcare charges (estimated at US$1.1 billion in 2012), mostly determined by the cost of the implant and partly by in-patient hospital stay) [185, 186]. This is especially relevant given that health care providers are incentivized to deliver more efficient and cost-effective care with outcome driven goals [187]. PCA and epidural analgesia provide excellent pain relief, but additional research is needed to determine best practices for each and to decrease the incidence of adverse effects. NSAIDs have proven to be excellent adjuvants that decrease opioid-related adverse effects, accelerate mobilization, and shorten LOS without introducing additional risk. The addition of IV APAP to an opioid‐only strategy with or without ketorolac saves at least US$510 per spine surgery patient and decreases opioid adverse effects [188]. However, the optimal cost-effective IV versus oral APAP dosage regimens have not been determined. Although implementation of an accelerated discharge program for the surgical treatment of AIS significantly reduced average LOS by 21%, this accounted for only a 9% decrease in the average cost per episode of care [186]. It is important to factor in cost savings associated with CPSP and prolonged opioid use, which are expensive problems. Additional research is needed for individualization of analgesia to prevent CPSP. Improved screening for at-risk patients, preoperative targeted risk optimization (based on psychosocial factors, setting of expectations, genetic and epigenetic factors) [165, 189], individualized multimodal regimens guided by pharmacogenomics [190, 191], early mobilization, and targeted follow-up for opioid tapering and functional rehabilitation is essential [192]. This is imperative in light of the present opioid crisis as the risk for new persistent opioid use after discharge is higher in children undergoing spine surgery [64, 193] and increases to 30% with continued need for opioids at 30 days [194].