25.3.1 Nutrition

Nutritional evaluation in the preoperative period is important as patients who are undernourished or obese both present an increased risk for wound complications and may require closer monitoring in the postoperative period. Body mass index (BMI), typically represented within the patient’s percentile at given sex and age, has been often used as a general indicator of nutritional status. It must be recognized that although patients with AIS are typically underweight, ^{19 , 20 , 21 , 22 , 23} this does not necessarily correlate with postoperative wound healing complications or infection. Postoperative weight loss and thus malnutrition may, however, contribute significantly to wound healing complications. Tarrant et al found that 30.6% of patients in their series of 77 patients undergoing PSF for AIS lost a clinically significant amount of weight from admission to discharge as defined as greater than 10% of body weight. ²⁴ Such weight loss was associated with a higher rate of infection (13.6 vs. 2%, p = 0.047). Patients required nearly 15 weeks to reach their preoperative weight and over 25% were defined as undernourished (Z-score < 2.0) at discharge. Obesity carries similar risks, typically related to wound healing and infection. Newton et al reviewed 2,122 patients undergoing PSF for AIS and found that obesity was the only significant risk factor for developing a postoperative infection (OR: 7.6; p < 0.0001) with patients having a BMI > 95% for age demonstrating the greatest risk. ²⁵

25.3.2 Pulmonary

A formal assessment of pulmonary function by a pulmonologist is likely unnecessary in the majority of patients with AIS. Simple screening questions may help direct appropriate referrals to a pulmonologist. Persistent cough or congestion, especially with productive mucous or frequent viral illnesses or colds, may indicate difficulty in a patient’s ability to clear secretions. Asking about a history of snoring, positive sleep studies, daytime fatigue, and difficulty sleeping can point toward sleep apnea. It is also important to ensure that the patient has no history of pulmonary complications from anesthesia including a need for postsurgical oxygen supplementation.

25.3.3 Gastrointestinal

Preoperative management of chronic constipation through the use of laxatives or motility agents can be useful for both the patient postoperatively and the surgeon intraoperatively as the stool in the colon can impact fluoroscopic data acquisition. Smith and Smith performed a small prospective randomized controlled trial of 60 patients undergoing PSF for AIS treated with either a preoperative bowel regimen (NuLytely) or no bowel preparation. Whereas the bowel regimen group showed less weight gain, required fewer postoperative bowel medications, and had a shorter time to first bowel movement, these benefits were modest, and many of the patients found the bowel preparation to be quite unpleasant. The authors did not recommend this preparation as a routine intervention. We will typically manage patients with chronic constipation using a less aggressive bowel preparation such as polyethylene glycol, docusate, or senna glycoside for 2 to 3 days preoperatively. ²⁶

25.3.4 Patient/Parent Expectations

Perhaps the final but most critical component of the preoperative evaluation includes a careful discussion of the surgical procedure and postoperative expectations for recovery. Patients and their families are understandably nervous and easily confused regarding the postoperative period. Many families have the impression that their children will need to be bedridden for some time after surgery, require a brace, or miss a prolonged amount of time from school. Establishing guidelines for discharge with the family before surgery is important as it gives the entire care team finite, distinct goals that must be met in order to be discharged. An accelerated or expedited care pathway will likely fail if the family is not educated on the criteria needed for discharge or the expected length of stay. It is also equally crucial that the family does not see a longer hospital stay as a failure. It is the author’s preference to give patients and families an understanding of our average length of stay while acknowledging that some children may in fact be discharged earlier and others may require a longer hospital stay. We also recommend that time out of school, as well as extracurricular activities such as exercise or sports, be discussed preoperatively. Whereas we cannot advocate for a specific return-to-play process given the lack of strong available literature, each surgeon should discuss this with the family.

25.4.1 Overview

Postoperative management of patients undergoing PSF for AIS will depend on the preferences of the patient and surgeon but must be structured around the facilities and capabilities of the hospital system. Most centers have transitioned to surgical floor admissions for AIS patients. ^{10 , 13 , 18 , 27} Some centers will still mandate admission to an intensive care unit or transitional unit if nursing limitations exist. The benefits of early postoperative mobilization in AIS patients have been demonstrated in many studies and have become the standard for postoperative care. The benefits include earlier independent mobility, return of bowel function, return to a regular diet, and discharge. With appropriate support from physical therapy, the nursing staff can assume the responsibility for mobilization in most settings. Surgeons, nurses, therapists, and administrators may use these guidelines, as well as numerous published manuscripts, ^{10 , 13 , 14 , 27 , 28 , 29 , 30 , 31 , 32} to support obtaining the resources necessary to adopt a standardized discharge pathway.

25.4.2 Pain Management

Perhaps no area of the postoperative management of patients undergoing PSF for AIS is more controversial than management of pain in the perioperative period. One of the central tenets of any AD pathway is the early transition to oral pain medications. As noted previously, a wide range of postoperative pain protocols exists in the literature. It is likely that the presence of an accepted protocol for managing pain before, during, and after surgery is more important than the specifics of the regimen. A variety of initial approaches are reasonable. Pain management with a single catheter, double catheter, or combined epidural with PCA all have support in the literature. ^{33 , 34 , 35 , 36} Epidurals may be placed with a continuous rate or with a patient-controlled bolus with similar efficacy. ³⁷ Li et al recently compared a single shot of intrathecal morphine (ITM) to an intraoperatively placed epidural in patients undergoing PSF for AIS. Patients with ITM demonstrated earlier Foley removal and ambulation in addition to lower pain scores during the first POD (visual analog scale [VAS] 2.9 vs. 4.2; p = 0.034) with scores equilibrating between groups over the remainder of the stay. Length of stay was slightly shorter in the ITM group (3.1 vs. 3.5 days, p = 0.043). ³⁸ The AD pathway described by Fletcher et al ^{10 , 14} and the rapid recovery pathways introduced by Gornitzky et al ¹⁸ and Muhly et al ²⁹ favor a narcotic-based PCA with early transition to oral opioids over an epidural. The Harms Study Group, using a modified Delphi approach, reviewed postoperative pain management and reached consensus that intravenous narcotics using PCA was preferred over an epidural catheter because of the risk profile of the latter including pruritus and respiratory depression (Fig. 25‑1). ²⁹ A single dose of intraoperative methadone, a long-acting opioid, can also be used to minimize patient demand dose of narcotics after surgery. ^{18 , 30 , 31} Early transition to oral narcotics may be initiated once the patient is tolerating oral intake; however, this does not require the presence of bowel sounds or other clinical parameters. ²⁸

Adjuncts to narcotics are a cornerstone of any multimodal drug therapy and include antispasmodics, anti-inflammatories, and neuroleptics. Antispasmodics (i.e., diazepam) have little published support for postoperative pain management despite their universal use within accelerated or rapid recovery pathways. ^{10 , 14 , 18 , 29 , 38} Antispasmodic use was also supported by the Harms Study Group in their publication of clinical practice guidelines for postoperative management following PSF for AIS. ²⁸ Ketorolac and gabapentin use has been associated with shorter duration of postoperative opioid use in a large national database study, with the former also being associated with shorter hospital stay. ³⁹ Controversy still exists with regard to risk of pseudarthrosis in patients who are given ketorolac for extended periods of time. ^{32 , 40} Much of the concern stems from adult literature and has not been corroborated in the setting of PSF for AIS. ^{32 , 41} Neither Sucato et al ⁴² nor Vitale et al ⁴³ found an increased risk of pseudarthrosis with the use of ketorolac. Munro et al ⁴⁴ studied the impact of 36 hours of scheduled ketorolac in patients undergoing PSF for AIS and found that it allowed for earlier mobilization with no documented pseudarthrosis. Gabapentin, a gamma-aminobutyric acid (GABA) analogue, can decrease perioperative opioid requirements when administered preoperatively and continued in the postoperative period. Rusy et al performed a randomized controlled trial comparing perioperative gabapentin to placebo. Patients were administered a preoperative loading dose of gabapentin (15 mg/kg), which was continued in the postoperative period at 5 mg/kg for 5 days. Opioid use and first pain scores were both significantly lower when compared to placebo in the gabapentin group. ⁴⁵ Mayell et al, who used only a single preoperative dose of gabapentin, found no benefit. ⁴⁶ At a more granular level, gabapentin use has been shown to decrease the time to reaching key physical therapy goals such as time to ascending stairs, suggesting that it assists with pain during early mobilization. ⁴⁷

More recently, the addition of dexamethasone has also shown benefit at reducing opioid usage while enhancing mobilization. Fletcher and colleagues administered an intraoperative dose of dexamethasone in addition to three postoperative doses given every 8 hours for the first 24 hours postoperatively in 48 patients with AIS and compared opioid usage, time to mobility, and complications with 65 patients who were not given a steroid. The patients who received dexamethasone used 40% less morphine milligram equivalents and were more likely to ambulate at the first physical therapy evaluation than those who did not receive a steroid. There were no differences in complications between groups including wound-healing difficulties; however, the authors were unable to assess the possibility of long-term pseudarthrosis. ⁴⁹

25.4.3 Nutrition and Bowel Management

Patients with AIS are typically healthy without preexisting gastrointestinal maladies. PSF is not physically disruptive to the digestive tract in the way that anterior spinal surgery may be. Traditional management dicta have suggested that reinstitution of liquid or solid intake should be delayed until the presence of some clinical parameter such as the return of bowel sounds. Specific risks associated with early reintroduction of oral intake after PSF have not been specifically addressed in the literature. Numerous other interventions, many seemingly more traumatic to the gastrointestinal system such as colectomy or major gynecologic surgery, have shown no increase in complications after early reinstitution of enteral feeding. Huang et al performed a meta-analysis of 1,800 patients undergoing cesarian section and found that early oral feeding was associated with earlier return of bowel function and was not associated with an increase in nausea, abdominal distention, or mild ileus. ⁴⁹ Miller et al ⁵⁰ demonstrated the benefits of an enhanced recovery protocol in colorectal surgery with patients showing shorter time to first bowel movement (2.4 vs. 3.4 days, p < 0.0001), earlier discharge with both open and laparoscopic surgery, and a lower readmission rate when compared to a traditional protocol. One of the central tenets of this protocol is early reinstitution of liquids. There does not appear to be any increase in return to the hospital or postoperative complications in patients undergoing PSF for AIS who are discharged prior to the return of bowel function. ^{10 , 13 , 18 , 28 , 29} The authors recommend beginning clear liquid intake immediately after surgery and a gradual return to a regular diet as the patient tolerates, often on the first postoperative morning. ²⁸

Bowel management in the perioperative period is predicated on the assumption that a postoperative ileus is expected following spinal surgery due to the combination of anesthesia, pain, and narcotics, among other factors. National databases suggest that gastrointestinal complications, the most common of which include ileus, occur in approximately 3% of cases. Carreon et al reviewed 702 patients included in the prospectively collected Spinal Deformity Study Group database and found a single patient with superior mesenteric artery syndrome and only 3 patients (0.43%) with symptomatic ileus. These numbers may be limited by the nature of reporting in a large database; however, they do underscore the fact that whereas many patients may have constipation postoperatively, the risk of symptomatic ileus is low.

A number of approaches have been used to expedite the first bowel movement. Smith and Smith evaluated the use of NuLytely preoperatively and found a slight improvement in time to first bowel movement (0.68 days) with less abdominal distention. Nonetheless, the authors felt that these benefits were minimal and that the inconvenience and discomfort of a bowel regimen outweighed a slightly earlier return of bowel function. ²⁶ A multitude of postoperative bowel regimens have been advocated, most of which include a promotility agent such as Miralax with a stool softener. We know of no literature supporting one regimen over another. The use of chewing gum to hasten a return of bowel motility has also been investigated. Jennings et al first reported on the potential benefit of chewing gum in a prospective randomized trial of 83 patients. Patients who chewed gum were found to have a bowel movement 30.9 hours sooner after surgery than those who did not. ⁵¹ Chan et al performed a similar randomized trial of 60 patients and failed to see any benefit with regard to return of flatus or time to first bowel movement. ⁵² Meng et al also failed to see any benefit in patients who chewed gum postoperatively, noting instead that patients who underwent selective thoracic fusion had a shorter time to first flatus than those undergoing a longer fusion. ⁵³ It is the authors’ preference to discuss the potential benefits with the family before surgery, as the downside risk of chewing gum is minimal.

Standard across all of the published early recovery pathways has been discharge prior to a first bowel movement. Return to the hospital for symptomatic ileus or constipation is rare. Fletcher et al reported only a single patient returning to the system for ileus after discharge prior to the first bowel movement. ¹⁰ Pre- and postoperative education about the time to first bowel movement is mandatory in order to minimize family concerns about a delay that may span several days. Patients should be discharged with instructions to ambulate regularly, to maintain their bowel regimen of a promotility agent until regular bowel movements are occurring, and to minimize narcotics whenever possible.

25.4.4 Mobilization

Early mobilization following PSF has gained favor over many years. Postoperative management of patients historically required prolonged recumbence due to lack of fixation or less stable implants and concern for the loss of implant fixation. Indeed, Hibbs’ initial reports on uninstrumented spinal fusion required the patient to be recumbent for 6 weeks in a plaster jacket prior to mobilization in a brace. ⁵⁴ A paradigm shift in postoperative mobility occurred as segmental spinal fixation grew more popular. As previously noted, the benefits of mobilization were espoused as early as 1973; however, the adoption of this concept took much longer. ⁷ Erwin et al published perhaps the first true case-control study on early mobilization after spinal fusion in 1976. This group compared 187 patients treated with Harrington instrumentation and 3 months of recumbency in a plaster cast to 177 patients who were treated instead with early mobilization at 7 days postoperatively in a cast that was worn for 6 months. No difference was seen in complications or loss of correction between the groups. There were fewer pseudarthroses in the prolonged recumbency group, likely related to the absence of bone graft, compared to the early mobilization group. ⁵⁵ McMaster noted in 1980 that the early mobilization of patients “7–10 days after operation” in a plaster jacket did not impact final fusion and allowed earlier return to normalcy. ⁵⁶

Modern postoperative management of scoliosis patients takes advantage of the stability afforded by segmental spinal instrumentation and rigid fixation. Clinically significant pseudarthrosis is rare and is unlikely related to early mobilization. Sanders et al attempted to have all patients treated at the Children’s Hospital of Los Angeles sit on the side of the bed or stand with nursing assistance as early as POD 0, ambulating by noon of POD 1 with physical therapy, and then twice daily with nursing. The authors noted that this was done without any need for the traditional “spine precautions” or “clearance from PT.” ¹³ A similar approach was taken at the Children’s Hospital of Philadelphia ¹⁸ and Children’s Hospital of Atlanta, ^{10 , 11 , 14} where mobility and ambulation were begun on POD 1 and encouraged three times per day thereafter. Early mobilization theoretically promotes return to normal function. Willimon et al surveyed patients about the return to school habits after a number of orthopaedic procedures and found that most patients undergoing PSF for AIS missed 4 to 6 weeks of school (27.6 in-school days and 42.3 days total). ⁵⁷ Sarwahi et al used a validated survey to show that 77% of patients returned to school by 3 months with one-half to two-thirds of patients returning to noncontact and contact sports by 6 months. ⁵⁸

It is our thought that the most important factor regarding return to school is the establishment of expectations preoperatively. Patients and families want to follow the “typical” postoperative course. If the expectation is set with the family as well as the school that a student will be able to return to school between 2 and 4 weeks postoperatively, it will be more likely that the family will be comfortable with this concept as the time for return approaches. On the contrary, if the decision is left entirely to the family, it is not uncommon for parents to fear for their children getting jostled in the school halls and to have concern over the potential for instrumentation dislodgement. Education about the security of the implants, the rigidity of the system, and the minimal risk associated with return to the school environment are critical to allowing for the resumption of normal activities. The authors have had parents who were counseled by well-meaning but perhaps less well-informed primary care providers or therapists to have the patient remain out of school for months and even entire semesters due to these fears. Careful, consistent, and repeated education of the family by the entire care team will ensure a quicker return and resumption of normal activities for patients and parents alike. Our team begins the discussion about early mobilization before surgery and reinforces it during the immediate postoperative meeting with the family, during which we establish walking by the morning after surgery as the expectation. This consistency helps reinforce the safety of mobilization with the family and other providers who may interact with the patient and have less familiarity with the care plan. Orders for early mobilization are also included in our standardized order set, discharge instructions, and follow-up.

Although discharge instructions are likely individualized, even at centers with a high level of standardization, some commonalities should be highlighted. A bandage change in the hospital does not appear necessary unless saturation of the dressing is noted. In a large meta-analysis of patients undergoing a range of surgical procedures, the Cochrane group was unable to determine the benefit of any one dressing for the prevention of surgical site infection. ^{59 , 60 , 61} Whereas no literature has investigated dressing changes in surgery for AIS, the use of 2-octyl cyanoacrylate (Dermabond, Ethicon Inc., Cincinnati, OH, USA) and an impervious dressing such as an OpSite (Smith and Nephew, Watford, England, UK) has been shown to decrease postoperative drainage and need for dressing changes in patients undergoing total knee arthroplasty. ⁶² In another meta-analysis, Toon et al with the Cochrane group found no evidence that earlier removal of a bandage increases wound complications. They noted that low-quality evidence does suggest that this practice may shorten hospital stay and decrease the cost associated with multiple dressing changes. ^{63 , 64} The Harms Study Group was unable to reach consensus on this matter, with 48% of members still recommending that the bandage be removed prior to discharge. ²⁸

Time to showering and bathing is another controversial topic. No data exist examining this topic with regard to PSF. Keeping the wound completely dry for long durations is likely unnecessary with modern multilayered closure, topical skin-glue sealants, and impervious dressings. Yu et al found no difference in wound infections after total knee arthroplasty when patients were allowed to shower at 2 days versus 2 weeks postoperatively. ⁶⁵ Our center has recommended that bandages remain on for 1 week from the day of surgery to simplify instructions for the family. The patient may shower immediately but should make an effort to avoid significant saturation of the bandage with water, instead focusing on washing the hair, front of the body, and legs. The bandage may be removed by the family at home and normal showering may commence at that time. We typically recommend that baths, soaks, Jacuzzis, and pools wait until the wound is completely healed, which is around the 1-month time point.

As indicated previously, discharge prior to the first bowel movement is typical of most accelerated pathways. Patients and families should be counseled that it may take several days in order to resume bowel function and that constipation is related to the duration of narcotics when combined with the physiological stresses of a large surgery. Our recommendation has been for patients to stay on a stool softener or promotility agent such as polyethylene glycol until regular narcotic use has subsided. Finally, 91% of members in the Harms Study Group recommended that the family be contacted by a provider or care team within the first week to ensure that all questions are answered and that care is being administered appropriately. ²⁸