Pathway for enhanced recovery after spinal surgery-a systematic review of evidence for use of individual components

Background Enhanced recovery in spinal surgery (ERSS) has shown promising improvements in clinical and economical outcomes. We have proposed an ERSS pathway based on available evidence. We aimed to delineate the clinical efficacy of individual pathway components in ERSS through a systematic narrative review. Methods We included systematic reviews and meta-analysis, randomized controlled trials, non-randomized controlled studies, and observational studies in adults and pediatric patients evaluating any one of the 22 pre-defined components. Our primary outcomes included all-cause mortality, morbidity outcomes (e.g., pulmonary, cardiac, renal, surgical complications), patient-reported outcomes and experiences (e.g., pain, quality of care experience), and health services outcomes (e.g., length of stay and costs). Following databases (1990 onwards) were searched: MEDLINE, EMBASE, and Cochrane Library (Cochrane Database of Systematic Reviews and CENTRAL). Two authors screened the citations, full-text articles, and extracted data. A narrative synthesis was provided. We constructed Evidence Profile (EP) tables for each component of the pathway, where appropriate information was available. Due to clinical and methodological heterogeneity, we did not conduct a meta-analyses. GRADE system was used to classify confidence in cumulative evidence for each component of the pathway. Results We identified 5423 relevant studies excluding duplicates as relating to the 22 pre-defined components of enhanced recovery in spinal surgery. We included 664 studies in the systematic review. We identified specific evidence within the context of spinal surgery for 14/22 proposed components. Evidence was summarized in EP tables where suitable. We performed thematic synthesis without EP for 6/22 elements. We identified appropriate societal guidelines for the remainder of the components. Conclusions We identified the following components with high quality of evidence as per GRADE system: pre-emptive analgesia, peri-operative blood conservation (antifibrinolytic use), surgical site preparation and antibiotic prophylaxis. There was moderate level of evidence for implementation of prehabilitation, minimally invasive surgery, multimodal perioperative analgesia, intravenous lignocaine and ketamine use as well as early mobilization. This review allows for the first formalized evidence-based unified protocol in the field of ERSS. Further studies validating the multimodal ERSS framework are essential to guide the future evolution of care in patients undergoing spinal surgery. Supplementary Information The online version contains supplementary material available at 10.1186/s12871-021-01281-1.


Background
Enhanced recovery after surgery (ERAS) programs have demonstrated improvements in outcomes. Improvements have been demonstrated in recovery, functional measures, lower morbidity, decreased length of stay with healthcare cost savings [1,2]. The disease burden of spinal pathologies is high [3]. Between 2004 to 2015, there has been an increase in volume of elective lumbar fusion accompanied by increased hospital costs [4]. Limited enhanced recovery pathways have been applied to spinal surgery. A consistent feature is a uniform finding of decreased length of stay [5][6][7][8]. There was a notable decrease in the adverse events during hospital stay [8,9].
There is a need to apply lessons learned from enhanced recovery programs in other surgical specialties to surgery of the spine [10]. Prior narrative qualitative reviews have delineated recommendations for the incorporation of individual components into an enhanced recovery after spinal surgery (ERSS) program. Several critical components of enhanced recovery in spinal surgery have been identified. These include: provision of comprehensive perioperative nutrition, multimodal analgesia, minimally invasive surgery where clinically feasible and early mobilization [10,11]. Individual ERSS programs differ substantially [12]. Our group of authors have identified and proposed the first comprehensive program of Enhanced Recovery in Spinal Surgery (Table 1), [11]. We defined the individual components based on the enhanced recovery protocols in other surgical subspecialties and prior qualitative reviews of ERAS in spinal surgery [1,[12][13][14][15][16][17][18][19][20]. . The aim of this study was to systematically evaluate pre-defined individual components of an ERSS pathway (program). We planed to create an evidence-based assessment of the available literature for each pre-defined component of an ERSS program [21]. Formulating the evidence base for each component, would strengthen the quality of ERSS programs. Consistency with regards to best practice in ERSS, would allow for standardization of care pathways. Greater standardization of care pathways results in improved external validity across comparative research.

Methods
This systematic review has been performed according the methodological standards for complex reviews [22][23][24][25][26][27][28][29]. Our findings have been reported according to the standards for the Preferred Reporting Items for Systematic Reviews and Meta-Analysis [30] (supplementary file 1). Protocol for this review was prospectively registered with the International Register of Systematic Reviews identification number CRD42019135289 [31]. The authors identified the essential components of enhanced recovery within the area of spinal surgery. The authors performed this process by reviewing the current enhanced recovery protocols as recommended by the ERAS Society. We identified and applied the relevant components to the field of spinal surgery [1,10,[12][13][14][15]32]. We have published this work through a peer reviewed protocol dissemination [11].

Eligibility criteria
Our patient population included adult and paediatric patients undergoing spinal surgical procedure on any spinal anatomical site. These anatomical sites cervical (anterior or posterior cervical decompression and fusion), thoracic (e.g., thoracic decompression and fusion), lumbar (e.g., lumbar decompression and fusion, lumbar laminectomy, sacral or any one combination of these). The interventions of interest have been classified in 5 perioperative pillars: preadmission period, preoperative period, intraoperative period, postoperative period, and audit and compliance processes (Table 1). These interventions (22 individual pathway components) have been defined through previous published work [11]. We reviewed the evidence with regards to each component studied independently or in any one combination [33].
Comparator interventions consisted of standard of care, no treatment or placebo.
In line with other ERAS reviews, we defined our primary outcomes in the following groups [12,34]: Morbidity, including pulmonary, cardiac, and renal complication rates; surgical complication rates; and readmission rates. Mortality from all causes. Patient-reported experiences and outcomes (PREMs/PROMs), including pain-related outcomes. (pain score rating and/or opioid consumption, pain management satisfaction), readiness for surgery, quality of care patient scores, and quality of recovery outcomes. Health service-related outcomes, including length of stay and reported economic/financial outcomes (e.g., costs of the length of stay).
We included systematic reviews and meta-analysis, randomized controlled trials, non-randomized controlled studies, and observational studies (e.g., cohort studies, case-control studies, cross-sectional studies, and case series). We included human data studies published in the English language after 1990. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses of Complex Interventions series lays out PICOTS framework for systematic reviews of complex interventions [35].
Our full review eligibility criteria are listed in Table 2.

Information sources and literature searches
The following electronic databases (from 1990 onwards) were searched: MEDLINE via Ovid SP; EMBASE via.
Ovid SP; and Cochrane Library (Cochrane Database of Systematic Reviews and CENTRAL). We searched the grey literature through the available search engines: Google Scholar, OpenGrey and GreyNet [36][37][38]. We initiated the original search for studies in January 2020 and updated it in May 2020. For the search strategy, we combined keyword(s) and subject headings for all literature types in the pre-determined databases [29]. Keywords were related to spine surgery, enhanced recovery, pre-operative care, intra-operative care, post-operative care, analgesia, mobilization, fluids. The specific details are contained within the supplementary files (supplementary file 2). We handled study overlap by tracking the index primary studies. For some selected pre-defined pathway components, there was a paucity of identified studies as pertaining to spinal surgery. Under those circumstances, we sought to identify large studies, metaanalysis or societal recommendations of best practice.

Data extraction, management, analysis and presentation
Standardized data parameters were extracted from each study. These parameters included: publication details, study characteristics, participant characteristics, type of spinal surgery, intervention and comparator characteristics, and outcomes. The results of the data search were presented in a PRISMA flow diagram indicating the number of studies retrieved, screened and excluded as per exclusion criteria (see Fig. 1). We have presented our findings according to each individual predetermined element of the multimodal enhanced recovery pathway (Table 1). One author extracted appropriate information from randomized controlled trials on the methodological quality of studies. This information included random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and other bias [39,40]. For non-randomized studies data extraction on random sequence generation and allocation concealment was not applicable.

Risk of Bias and thematic synthesis
Risk of bias in randomized controlled studies was assessed using the Cochrane Risk of Bias tool [40]. ROBINS-I (Risk of Bias in Non-randomized Studies of Interventions) tool was used to evaluate the risk of bias in non-randomized studies [41]. We used the revised AMSTAR-2 tool to assess the risk of bias in systematic reviews [42]. We used the GRADE system (Grading of Recommendations, Assessment, Development and Evaluation) system to classify the evidence into one of four categories: high, moderate, low, and very low [43,44]. Evidence based on randomized controlled trials was considered as high quality unless confidence in the evidence was decreased due to study limitations, inconsistency of results, indirectness of evidence, imprecision, and reporting biases. Observational studies were considered low quality; however, they were graded higher if the treatment effect observed is very large or if there is evidence of a dose-response relationship [33,44,45].
Endpoint of the GRADE evidence summary consists of Evidence Profile (EP) tables across individual pathway components [43]. Risk of bias across outcomes for individual pathway components was presented in Evidence Profile Tables (supplementary file 3). We performed a thematic synthesis and narrative analysis for each proposed component [27]. Forest plots were generated for the following components: anaesthetic protocol, use of multimodal analgesia and intravenous lignocaine infusion (supplementary file 5). In line with our planned protocol, quantitative data synthesis was not attempted due to the inherent heterogeneity of the studies. This method of evidence synthesis is in line with other published enhanced recovery reviews [1,13,15,17,18,46,47]. We did not make recommendations on the utility of pathway components, in line with recommended practice for systematic reviews [48].

Results
Our search strategy retrieved a total of 5423 studies excluding duplicates using 22 different searches for the each relevant ERSS item as outlined in Table 1. During the review of full text articles, we excluded studies which did not pertain to the topic studied (surgery of the spine), which did not describe the intervention in sufficient detail or published articles which were not methodologically suited (case reports, opinions, comments, narrative reviews). Where studies were not available pertaining to pre-defined pathway component of spinal surgery, databases and grey literature were reviewed as relating to societal recommendations and major pertinent studies for perioperative patient management. This methodology yielded 148 further studies for inclusion. We included 664 studies in the final review. The results of our search have been presented in Fig. 1, PRISMA Diagram. We have grouped the evidence base according to the component of the pathway.
Evidence Profile tables were generated when a number of studies were identified investigating an intervention for one of the predetermined outcomes. We have generated Evidence Profile tables for the following pathway components: 2.4 Tobacco use, 3.Prehabilitation, 4.1Preoperative nutritional screening, 5.Management of anemia, 6.Peri-operative blood conservation strategies, 12. Standard anaesthetic protocol, 16. Perioperative analgesia including use of intravenous lignocaine and 21.Mobilization.
For the following elements we identified published meta-analysis: 6.Peri-operative blood conservation strategies, use of tranexamic acid, 8. Pre-emptive analgesia, 10.2 Antimicrobial prophylaxis, 11. Local anaesthetic infiltration, 13. Surgical access (open and minimally invasive spinal surgery), 16. Perioperative analgesia including use of NSAIDS, ketamine, gabapentinoids and intrathecal morphine and 17. Thromboprophylaxis. We incorporated the relevant meta-analysis findings into each pathway.
We were able to identify heterogenous studies pertaining to surgery of the spine for the following components: 10.1 Surgical site preparation,14. Maintenance of normothermia, 15. Intra-operative fluid and electrolyte therapy, 18. Urinary drainage, 19. Post-operative nutrition and fluid management and 20.Post-operative glycemic control. For these components, we were unable to construct evidence profile tables. As such, we performed a thematic synthesis of evidence.
Due to the paucity of evidence pertaining to spinal surgery, we identified societal recommendations for the following components: 1.Preadmission information, Risk assessment (2.1 Preoperative risk stratification, 2.2 Preoperative optimization and 2.3 Alcohol use), 4.2 Perioperative immuno-nutrition, 7. Pre-operative fasting and carbohydrate loading, 9. Prevention of post-operative nausea and vomiting and 22. Audit.

Presentation
We have presented our findings according to each individual element of the multimodal enhanced recovery pathway in line with other subspecialty ERAS pathways [13,14,49]. Please see Table 3 and supplementary file 4.

Preadmission period
The preadmission period is an opportunity for patient education, assessment of comorbidities, risk stratification and optimization of modifiable patient-related factors.

Preadmission information, education and counseling
Patient information provision has long been considered a key element of enhanced recovery pathways [50,51]. Patient perioperative experience and the psychological aspect may be improved with pre-admission counseling [52][53][54][55]. Psychopathology and patient expectations have been linked to poor results in spinal surgery with

Management of anaemia
Clinically guided use of intravenous or oral iron, vitamin B12, folic acid or erythropoietin for patients suffering from anaemia and/or low iron stores should be implemented in patients undergoing moderate and major spinal surgery; 6. Perioperative blood conservation strategies Tranexamic acid used at the higher dosage is effective in decreasing intraoperative blood loss. Cell saver techniques should be used in adolescents undergoing major corrective surgery. Cellsaver techniques may be beneficial when major blood loss is anticipated in adults.

7.Preoperative fasting and carbohydrate loading
In patients without delayed gastric emptying standard societal fasting implementations can be made; 8. Pre-emptive analgesia Multimodal pre-emptive analgesia utilizing individual gabapentinoids and non-steroidal antiinflammatory agents improves pain scores and functional measures in the immediate postoperative period;

Prevention of postoperative nausea and vomiting
Risk assessment of patients according to the anaesthetic and procedural factors is recommended.
Step-wise non-pharmacological and pharmacological PONV prophylaxis according to the guidelines is recommended. Use of anaesthetic techniques which minimize risk of PONV in high-risk patients should be considered; 16. Peri-operative analgesia Simple analgesics such as acetaminophen and NSAIDs are safe and efficacious, particularly in increased pain and decreased function. This has led to an increased reliance on pre-surgical psychological screening (PPS) as part of the surgical diagnostic process in spinal surgery [55][56][57][58]. Studies do not show any evidence of harm from preoperative information provision or psychological intervention. There may be utility from information provision, balanced against no known harmful effects. There is limited evidence available pertaining to the intervention specifically in patients undergoing surgery of the spine.
2. Risk assessment, preoperative assessment, optimization and lifestyle factor modification 2.1. Pre-operative risk stratification Perioperative period offers an opportunity for risk stratification [59]. Nearly 80 % of patient deaths come from the high-risk patient group [60]. In a major retrospective study in the USA, it was found occurrence of a major complication within 30 days of surgery was associated with reduced median survival by 69% at 8 years [61]. Multiple diverse risk scoring systems are currently in use for major surgery, including spinal surgery. Assessing cardiovascular risk can be undertaken whilst utilizing ACC/AHA guidelines on perioperative cardiovascular evaluation and care for non-cardiac surgery [62]. Lee's cardiac risk index, Revised Cardiac Risk Index (RCRI), POSSUM (Physiological and Operative Severity Scoring for the enumeration of Mortality and morbidity), the Portsmouth POSSUM (P-POSSUM) and Duke Activity Status Index (DASI) scores have been validated for patients undergoing major non-cardiac surgery [63][64][65][66].
The American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) has been deemed to have high internal validity although external validation has been inconsistent [67,68]. A retrospective cohort study of the NSQIP database consisting of patients who had undergone elective posterior lumbar fusion was undertaken. The variables associated with greater risk and extended length of stay included increasing age, morbid obesity, operative time, multilevel procedure and intraoperative transfusion [69]. In a retrospective review of NSQIP database, patients undergoing elective spinal surgery had the expected risk factors for cardiovascular complications consistent with those demonstrated by the Revised Cardiac Risk Index [70]. Individual preoperative risk assessment tools tests can be used to identify patients at risk of complications and perform a risk-based stratification. There is moderate quality evidence in support individualized risk stratification utilizing the most suitable risk-stratification assessment tool.

Preoperative assessment and optimization
Patient preoperative assessment allows for an opportunity for examination of comorbidities with subsequent identification of fixed and optimization of modifiable conditions . Obese patients having spinal surgery were found to have increased blood loss, prolonged hospital stay and were more likely to develop infection [76,77]. Patients with diabetes were found to have greater disability and more likely to have failed spinal fusion as compared to patients without diabetes [78][79][80]. Frailty is an emerging risk assessment tool, requiring further studies [81]. The degree to which preoperative optimization and modification of multimorbidity, affects healthcare outcomes is unclear [59,81]. Modifiable co-morbidities should be optimized using the preoperative process. Evidence base is of low quality due to a limited number of heterogenous studies. 2.3. Alcohol use Postoperative morbidity is increased by two-to threefold in alcohol abusers [82]. Preoperative alcohol consumption is associated with an increased risk of postoperative morbidity, general infections, wound complications, pulmonary complications, prolonged stay at the hospital, and admission to intensive care unit [83].
In a subset of patients without clinical or historical evidence of alcohol-related illness, 1 month of preoperative abstinence has been shown to significantly improve outcome [83][84][85]. Significant alcohol consumption has been shown to be associated with increased perioperative morbidity. For alcohol abusers, 1 month of abstinence before surgery is beneficial. Evidence is considered to be of moderate quality due to heterogenous endpoints in available studies.

Tobacco use
Smoking is an independent risk factor for non-union in spinal fusion procedures [86][87][88][89]. Postoperative infection and wound complications are significantly increased by tobacco consumption [88]. Decreased risk of infection, perioperative respiratory problems, and wound complications have been demonstrated 1 month after cessation of smoking [90]. Longer periods of cessation of smoking appear to be more effective in reducing the incidence/risk of postoperative complications [91,92]. There is translational high quality evidence for cessation of smoking at least 4 weeks preoperatively.

Prehabilitation
Prehabilitation can be defined as "the process of enhancing the functional capability of an individual in preparation for the surgical intervention". This process consists of: functional preoperative prehabilitation, nutritional and psychological intervention [93]. Pre-operative functional capacity is closely related to post-operative morbidity [66]. Whether improving the post-operative outcomes through prehabilitation has beneficial effects on mortality is not yet clear [94]. Multimodal prehabilitation in spinal surgery has been associated with improved recovery milestones, earlier discharge and appreciable improvement in patient satisfaction scores in the study group [95,96]. Health-economic benefits have been greater in patients having prehabilitation [96,97]. Patient reported outcomes such as readiness for surgery and perceived quality of life, were found to be improved by pre-operative neuroscience education and physiotherapy [98][99][100]. Overall evidence quality for prehabilitation is moderate.
There is moderate quality evidence available for performance of risk assessment and screening of nutritional status in patients undergoing spinal surgery.

Perioperative immuno-nutrition
Overall systematic evidence on immuno-nutrition (IN) in surgery has been contradictory [106,107]. Clinical studies demonstrating benefit of IN are heterogenous with nonstandardized primary solutions, controls or timing of administration of supplements [108]. There is no evidence for use of IN in patients undergoing surgery of the spine.

Management of anaemia
Preoperative anemia is an independent risk factor for increased 30-day mortality and morbidity in surgical patients [109][110][111][112]. In patients undergoing surgery of the spine, preoperative anaemia was associated with increased length of stay [113,114]. Intraoperative blood transfusion in spinal surgery has been associated with increased postoperative complications, length of hospital stay and 30-day re-admission rates [115]. It is however unclear whether correcting iron deficiency anaemia improves reported outcomes, other than decreasing the risks associated with perioperative blood transfusion [116,117]. It is unknown whether correcting nonanaemic iron deficiency (NAID) decreases the risk of perioperative complications [118]. The association of iron replacement therapy, in particular intravenous format, with infection, is currently contentious [119,120].
Clinically guided appropriate pre-operative use of intravenous or oral iron, vitamin B12, folic acid or erythropoietin for patients suffering from anaemia and/ or low iron stores should be implemented in patients undergoing moderate and major spinal surgery. There is moderate quality translational evidence for correcting the iron deficiency anaemia, in order to decrease the perioperative risk of complications secondary to blood transfusion.

Perioperative blood conservation strategies
Patients undergoing moderate and major spinal surgery are at risk of significant blood loss necessitating fluid and blood product replacement [121,122]. Recent systematic analysis of tranexamic acid use in spinal surgery patients concluded that the use of tranexamic acid (TXA) was effective in reducing intra-operative blood loss and decreased the volume of blood transfusion [123]. It is likely that the higher bolus doses employed (greater or equal to 15 mg/kg followed by intraoperative infusion) were more effective in attenuating blood loss and transfusion requirements [124,125]. Lower perioperative blood loss has consistently been demonstrated with the use of -amin-caproic acid (EACA) in spinal surgery [124]. Point of care testing devices allow for standardization of transfusion practices and early identification and treatment of hypofibrinogenemia [122]. In a study of patients undergoing major spinal surgery ROTEM (ROtational ThromboElastoMetry) device used with TXA was found to lead to a significantly lower blood loss and lower transfusion of packed red blood cells as compared to the TXA alone [126]. Studies have demonstrated the utility of ROTEM in decreasing the rate of blood product transfusion [127,128].
Blood conservation options include preoperative autologous blood donation and intraoperative cell saver use.
Preoperative autologous blood donation in elective major spine surgery has been effective in reducing allogenic transfusion, however inclusion in the program resulted in increased risk of transfusion [129][130][131]. A Cochrane meta-analysis assessing the use of cell saver in major surgery demonstrated significantly decreased rate of allogenic blood transfusion [132]. The use of the cell saver in posterior spinal instrumentation and fusion surgery in school-aged children and adolescents was able to decrease the amount of intraoperative allogeneic RBC transfusion but failed to decrease total perioperative allogeneic RBC transfusion [133,134]. In contrast, use of cell-saver was found to be associated with increased risk of bleeding in patients having spinal fusion surgery [135]. There is no evidence supporting the use of controlled hypotension to minimize the risk of bleeding, particularly in prone patients [136]. Perioperative blood conservation is aided through simple clinical measures such as temperature regulation, optimal patient positioning and meticulous surgical techniques.
There is high quality evidence for antifibrinolytic use in surgery of the spine when significant blood loss is anticipated. There is moderate quality evidence for cell saver use when significant blood loss is anticipated. There is low quality evidence for use of point of care testing to decrease the number of red blood cell units transfused.
Pre-operative period 7. Preoperative fasting and carbohydrate loading In a randomized controlled trial examining patient population having co spinal surgery, preoperative carbohydrate loading did not attenuate postoperative insulin sensitivity [137]. This is contradictory to other general surgical trials which demonstrate improved insulin postoperative sensitivity with CHO loading [138][139][140][141]. The clinical relevance of administering preoperative CHO loading in patients with diabetes remains to be established [47]. Permitting patients to drink water or clear fluid preoperatively results in significantly lower gastric volumes [139]. International guidelines allow for unrestricted intake of clear fluids up to 2 h before elective surgery in patients not considered to have impaired gastric emptying [140,142].
In spinal surgical patients without delayed gastric emptying standard societal fasting implementations can be made. Patients should be allowed to eat up until 6 h and take clear fluids including CHO drinks, up until 2 h before initiation of anaesthesia. Preoperative treatment with oral CHOs may not be suitable in patients with documented delayed gastric emptying, gastrointestinal motility disorders and in patients undergoing emergency surgery.

Pre-emptive analgesia
A number of studies found that pre-emptive administration of gabapentin reduced the opioid consumption and pain scores in the postoperative period in spinal patients [143][144][145][146]. The most effective dose in lowering the postoperative pain scores was found to be 600 mg [144]. Impact of multimodal anti-inflammatory regimes combined with gabapentinoids, is significant in lowering the postoperative pain scores [147,148]. Parecoxib and ketorolac were found to be equally effective in improving postoperative pain measures. Both were superior to placebo in patients undergoing posterior lumbar fusion [149]. Preemptive epidural analgesia for thoracolumbar spine surgery has not been deemed effective [150].
Multimodal pre-emptive analgesia utilizing individual gabapentinoids and/or non-steroidal anti-inflammatory agents improves pain scores and functional measures in the immediate post-operative period. There is high quality evidence for preemptive administration of gabapentinoids in patients undergoing surgery of the spine.
There is moderate quality evidence for pre-emptive administration of multimodal anti-inflammatory regimes combined with gabapentinoids. Evidence quality is low for sole administration of individual anti-inflammatory agents.

Intraoperative period 9. Prevention of nausea and vomiting (PONV)
Postoperative nausea and vomiting (PONV) are amongst the most frequent postoperative complications, impacting the quality of recovery and causing patient dissatisfaction [151]. Patients should be risk-stratified according to the baseline risk of PONV. In patients undergoing spinal surgery a particular risk factor for increased PONV is the need for significant intraoperative and postoperative opioid administration. Standard societal guidelines for PONV prophylaxis and management apply to patients undergoing spinal surgery [151]. With nonpharmacological measures, avoidance of fasting and dehydration has been recommended. Multi-modal analgesia with opioid sparing effect has a beneficial influence on the risk of PONV [152,153].
Risk assessment of patients according to the anaesthetic and procedural factors is recommended.
Stepwise non-pharmacological and pharmacological PONV prophylaxis according to the guidelines is recommended. Use of anaesthetic techniques which minimize risk of PONV in high-risk patients should be considered.
There is high quality evidence for risk stratification of patients and appropriate anti-emetic prophylaxis.
There is moderate quality evidence for opioid sparing techniques as well as avoidance of nitrous oxide and volatile anaesthesia; Large prospective cohort studies have demonstrated that the surgical site infection rate is equivalent with both topical chlorhexidine gluconate and povidone in spinal surgical patients [155,156]. Topical chlorhexidine with alcohol compared to povidone alone, decreased the bacterial load significantly in spinal surgery [157,158]. A review evaluating randomized controlled trials in all types of surgeries concluded that alcohol-based agents are superior to aqueous solutions [159].
There is high quality evidence for using alcohol-based preparations. There is moderate quality evidence for decreasing the viable bacterial load utilizing CHG with alcohol solution.

Antimicrobial prophylaxis
Risk factors for surgical site infections after spinal surgery include extended duration of procedure (longer than 2 h), excessive blood loss (greater than one liter), staged procedure, multilevel fusion, foreign body placement, combined, anterior and posterior fusion, and poor peri-operative glycemic control [160]. Surgical site infection is less likely in procedures at the cervical spine level or with an anterior surgical approach [161]. Current guidelines recommend intravenous cephazolin as the first choice agent for antimicrobial prophylaxis for most surgical procedures [162]. In patients with MRSA, intravenous vancomycin is recommended 1 h prior to skin incision. Clindamycin is an acceptable alternative in patients with a cephalosporin or vancomycin allergy. In the setting of risk for SSI due to gram-negative pathogens, an additional agent may be warranted (such as an aminoglycoside, aztreonam, or a fluoroquinolone). In order to ensure adequate antimicrobial serum and tissue concentrations, repeat intraoperative dosing is warranted for procedures that exceed two half-lives of the drug and for procedures in which there is excessive blood loss [162]. In a metaanalysis incorporating 6 prospective randomizedcontrolled trials, antibiotic prophylaxis was found to decrease the rate of infection [163]. Whether postoperative infections are reduced by continuing use of prophylactic antibiotics remains controversial [164]. In a metaanalysis consisting of 14 mostly class 3 evidence studies, vancomycin powder was found to decrease the likelihood of surgical site infection [165]. Vancomycin powder was found to decrease the rate of deep space infections requiring re-operation [166]. Vancomycin powder should be restricted to procedures and patients most at risk of MRSA-related surgical site infection [167][168][169].
Routine prophylaxis with cefazolin within 1 h prior to skin incision is recommended. Patients with MRSA should be treated prophylactically with vancomycin initiated 1 h prior to skin incision.
There is high quality evidence for intra-operative antibiotic prophylaxis. There is low quality evidence for use of intravenous vancomycin in patients at risk of MRSA.

Local anaesthetic infiltration
The benefits of intra-operative wound infiltration for postoperative analgesia in spinal surgery are controversial. A number of studies have demonstrated conflicting results in this area [170][171][172][173]. A meta-analysis of nine trials exploring the effect of wound infiltration in spinal surgery concluded that only a few trials observed a mild to modest pain score reduction. Of the trials which did show pain reduction, the analgesic benefit was noted in the immediate post-operative period [174].
Local anaesthetic wound infiltration in major spinal surgery has some immediate benefit on postoperative pain scores. There is moderate quality evidence for intra-operative administration of long-acting local anaesthetic administration.

Standard Anaesthetic protocol
Prior systematic reviews and meta-analysis have concluded that recovery parameters are improved with the use of total intravenous anaesthesia (TIVA) [175,176]. There is some evidence that patients receiving TIVA had improved cognitive outcomes in post-anaesthesia recovery unit in all types of surgical patients [177]. Patients anesthetized with propofol-based TIVA reported less pain during coughing and consumed less daily and total PCA fentanyl after lumbar spine surgery [178]. This finding was not consistent across all studies [179]. Remifentanil, ultra-short acting phenyl-piperidine derivative is used in spinal surgery as part of total intravenous anaesthesia or inhalational anaesthesia protocols. Indications for use in spinal surgery include: improved endotracheal tube tolerance, improved surgical conditions and facilitations of peripheral neuromuscular monitoring. Severe postoperative pain after the intraoperative use of remifentanil has repeatedly been linked to the development of acute tolerance and/or opioid induced hyperalgesia [180]. In patients undergoing spinal fusion remifentanil dosage up to 0.16 mg/kg/min did not cause an increased post-operative opioid consumption [181]. In contrast, in patients having correction of scoliosis where higher doses of remifentanil of 0.28 mcg/kg/min were used for longer duration, the requirements for post-operative analgesia were 30 % higher in the remifentanil group [182]. Neurologic monitoring in spinal surgery is performed using the intraoperative somatosensory potentials (SSEP's) and/or the Motor Evoked Potentials (MEP's). During the SSEP monitoring anaesthetic drugs produce a dose dependent increase in latency and a decrease in amplitude. The overall quality of SSEP is superior when propofol total intravenous anaesthesia is used. International Society of Intraoperative Neurophysiology recommends use of propofol and opioid [183]. MEP's display extreme sensitivity to the inhibitory effects of volatile agents even at concentrations as low as 0.25 MAC. Due to a lower level of interference with monitoring MEP's, propofol total intravenous anaesthesia is recommended for patients requiring spinal cord neurophysiological monitoring during surgery.
There is moderate quality evidence for use of total intravenous anaesthesia in patients undergoing surgery of the spine. There is low quality evidence for continuous intra-operative remifentanil infusion use in spine surgery.

Surgical access (open and minimally invasive spinal surgery, including robotic surgery)
Minimally Invasive Spinal Surgical (MISS) techniques can be viewed as a critical component of enhanced recovery in spinal surgery protocols (ERSS) [184]. Reduced length of stay together with significant cost saving has been identified in studies utilizing MISS techniques [32]. MISS techniques have been efficacious in decreasing postoperative pain in observational studies [185]. With a focus on minimally invasive transcutaneous lumbar inter-body fusion, Wang et al. demonstrated that ERAS in this group of patients was feasible and afforded improved early functional outcomes [186]. MISS approach studied within the enhanced recovery protocol was found to be effective in oncological spinal patients, where it was found to decrease the pain scores and lower the opioid consumption [187]. Significantly faster mobilization was demonstrated in patients undergoing minimally invasive thoracic inter-lumbar body fusion compared to open procedure [188]. In contrast to single studies and qualitative reviews, a quantitative metaanalysis found there was equipoise in patients undergoing lumbar minimally invasive procedures [189]. A multicenter study found equivalent outcomes for obese patients having spinal MISS or open techniques [190]. Conversely, Senkar et al. found minimally invasive surgical techniques had the highest utility in patients with multiple comorbidities [191].
There is evidence that minimally invasive surgical approaches improve pain scores, decrease opioid consumption and decrease length of stay, when used within the appropriate clinical context. There is moderate quality evidence for the intervention in appropriate clinical context.

Maintenance of Normothermia
Maintenance of normothermia has been shown to decrease the frequency of morbid cardiac events and the rate of blood product transfusion in major surgery [192,193]. In spinal procedures with potential neurological cord compromise, maintenance of normothermia and avoidance of hyperthermia is recommended [194]. There is little scientific literature supporting the neuroprotective effects of hypothermia on the spinal cord in elective or emergency spinal surgery [195]. In pediatric spinal surgery maintenance of normothermia was found to be associated with a lower allogenic red blood cell transfusion rate [196]. In contrast hypothermia may be associated with a lower rate of acute kidney injury in spinal surgery under general anaesthesia [197].
Measures to maintain normothermia should be implemented in spinal surgical patients. There is moderate quality evidence for maintenance of intraoperative normothermia.

Intraoperative fluid and electrolyte therapy
For the minor range of spinal surgeries intraoperative fluid management goals are achievable with routine monitoring. In major surgery, goal-directed therapy has been recommended [198]. Advanced haemodynamic monitoring equipment chosen should be based on a clinical risk-management strategy and patient, anaesthetic, surgical and institutional factors. Prior meta-analysis have demonstrated that pre-emptive hemodynamic monitoring and proactive therapy reduces mortality and morbidity in major surgical procedures [199,200]. In a retrospective observational trial in patients undergoing prone spinal surgery, goal directed fluid management was found to decrease blood loss and transfusion, improve postoperative respiratory performance and allow for faster return of bowel function [201]. In contrast, liberal fluid strategy was associated with an increased rate of pulmonary complications [202].
Goal-directed fluid management may decrease the rate of complications and duration of stay when implemented in the appropriate clinical context. There is low quality evidence for goal-directed intraoperative fluid management using contextually appropriate indicators and measurements of cardiac output in patients undergoing major surgery of the spine.

Peri-operative analgesia
Poorly controlled pain in the post-operative period can influence mobility and result in increased rate of complications of deep venous thrombosis, pulmonary embolism and pneumonia [203].
NSAIDS and acetaminophen A recent meta-analysis of eight trials identified that NSAIDs are effective in postoperative analgesia after lumbar spine surgery. The study found that NSAID dose, different surgery types, and analgesic type might influence the efficacy of NSAIDs [204]. A meta-analysis of 17 studies demonstrated that addition of NSAIDs to opioid analgesics alone resulted in lower pain scores and less morphine equivalents consumed [205]. In a meta-analysis of seven spine fusion studies, no statistically significant association between NSAID exposure and nonunion was identified (odds ratio = 2.2, 95% confidence interval 0.8-6.3) [206]. It is likely that adverse effects of NSAID's on bone healing/ fusion in adult spine surgery are dose-dependent [207]. While there is limited evidence for the use of acetaminophen specifically in spinal surgery, it is a well-established analgesic agent for a wide range of related surgeries [203].
N-methyl D-aspartate antagonists Randomized controlled trials demonstrating decreased opioid consumption and lower pain scores following intraoperative and post-operative ketamine [208][209][210]. These findings are in line with a meta-analysis of eight trials [211]. A single study showed no benefit of low dose ketamine in major lumbar surgery [212]. Methadone and magnesium through their NMDA antagonism may also be of benefit; however, data are limited and further studies are indicated [213][214][215].
Alpha-2 receptor agonists Data supporting the use of alpha-2 receptor agonists in major spine surgery are limited; studies have demonstrated conflicting findings [216].
Gabapentinoids In a systematic review and metaanalysis by Yu et al., perioperative administration of gabapentinoids was found to decrease opioid consumption and pain intensity in the immediate post-operative period [145]. Other high quality prospective studies have deemed gabapentinoids effective at reducing the opioid consumption when continued for at least 24 h postoperatively [146]. A prospective, double-blind study, randomized control trial by Khurana et al. showed a stronger benefit for pregabalin over gabapentin versus placebo for pain and functional status in the postoperative period and at 3 months [217].
Intravenous lignocaine In a number of controlled trials in both adult and pediatric major and minor spine surgery, perioperative lignocaine infusion was demonstrated to improve pain scores and decrease opioid consumption [218][219][220][221]. Conversely, in a randomized controlled trial of 70 patients undergoing posterior spine surgery, there was no analgesic benefit of a systemic lignocaine infusion as compared to placebo [222].
Regional analgesia Intrathecal morphine administration in a wide dosage range as a single injection has been found to be effective as a postoperative analgesic in spinal surgery, though doses greater than six mcg,kg − 1 are associated with postoperative respiratory depression [223][224][225][226][227][228][229][230]. In a meta-analysis of eight randomized controlled trials, intrathecal morphine was an effective analgesic [231].
Multimodal regimens Prior review articles have highlighted multimodal analgesia as a significant contributor to enhanced recovery in spinal surgery [10,203,232]. Multimodal analgesia bundles have been incorporated into most care pathways of enhanced recovery in spinal surgery [8,12,233]. A numbed of retrospective studies have demonstrated decreased pain measurement outcomes including post-operative opioid consumption [234][235][236][237]. In contrast to other studies, a single randomized controlled trial of optimally dosed multimodal regime did not show any benefit over the placebo components when evaluated in terms of quality of recovery scores or analgesic components [238]. Minimally invasive opioid free enhanced recovery protocols in spinal surgery have been shown to have a favorable profile on perioperative opioid consumption [239].
Simple analgesics such as acetaminophen and NSAIDs are safe and efficacious, particularly in combination. There is high quality evidence for perioperative administration of NSAID's. Ketamine in both intraoperative and post-operative infusions, reduces pain scores, opioid requirements in the immediate and late post-operative phases. There is moderate quality for intraoperative ketamine administration. There is very low quality evidence for administration of other NMDA antagonists and alpha-2-agonists. There is high quality evidence for perioperative gabapentinoid administration. Consideration should be given to perioperative intravenous lignocaine infusion administration. There is moderate quality evidence for perioperative intravenous lignocaine administration. There is moderate quality evidence for use of intrathecal morphine in spinal surgery, although its utility may be limited by logistical factors. Clinically appropriate multimodal opioid-sparing regimens should be considered in all patients undergoing spine surgery. There is moderate quality evidence for instituting perioperative multimodal analgesia.

Postoperative period 17. Thromboprophylaxis
Mechanical thromboprophylaxis is a proven measure to decrease the risk of deep venous thrombosis (DVT) in the absence of chemoprophylaxis [240]. A meta-analysis conducted in 2018 found that the incidence of DVT and pulmonary embolism (PE) in spinal surgical population was relatively low regardless of prophylaxis type. The authors commented that there was a higher mean incidence of DVT and PE in the mechanoprophylaxis group (DVT: 1%, PE: 0.81%) compared to the chemoprophylaxis group (DVT: 0.85%, PE: 0.58%) [241]. In this study, when PE occurred it was fatal in 6 % of patients. Perception of true incidence of post-operative epidural haematoma in spinal surgical patients is varied [242].
Patients undergoing spinal surgery should have mechanical thromboprophylaxis by well-fitting compression stockings and/or intermittent pneumatic compression until discharge. There is moderate quality evidence for postoperative mechanical thromboprophylaxis in patients undergoing spinal surgery.
There is low quality evidence for postoperative chemical thromboprophylaxis in patients undergoing spinal surgery.

Urinary drainage
Urinary catheter use beyond 48 h following surgery has been associated with an increase in hospital-acquired urinary tract infections and 30-day mortality [243]. In a nested cohort study in a neurological intensive care unit, an increased rate of urinary infection was noted in patients, where catheter remained in place for longer than 7 days [244]. Risk factors for postoperative urinary retention in spinal surgery include older age, benign prostatic hypertrophy, chronic constipation, longer duration of surgery and posterior spinal fusion [245][246][247].
If urinary drainage is indicated, the duration of catheterization should be individualized based on known risk factors for urinary retention. There is moderate quality evidence for urinary catheter removal within 48 h after surgery.

Postoperative nutrition and fluid management
Many of the studies report on early mobilization in conjunction with dietary libertization [185,248]. When performed together, the two can reduce length of stay and costs without increasing early or late complications in adolescents undergoing posterior spinal fusion [249]. Results of the RELIEF trial suggest that we should be more cautious with postoperative restrictive fluid strategies in patients having major abdominal surgery [250]. In patients having major spine surgery, goal orientated postoperative fluid management may be more appropriate than a restrictive approach, although specific evidence is currently lacking. Intraoperative haemodynamic framework may be continued into the post-operative period in the high-risk patient group. In line with other ERAS guidelines patients should be encouraged to transition as early as tolerated to oral intake.

Postoperative glycemic control
A retrospective cohort study incorporating population undergoing spine surgery found that perioperative hyperglycemia increases the risk of adverse postoperative events in the non-diabetic patient group [251]. Tighter glycemic control may mitigate the risk of surgical site infection in patients with diabetes [252]. There remains insufficient evidence that strict glycemic control is advantageous over conventional management for prevention of surgical site infection [253]. Although it is clear that perioperative hyperglycemia is deleterious, the optimal management paradigm in the postoperative period remains uncertain [252].
It is prudent to maintain more conventional blood glucose target in the postoperative period in patients undergoing spinal surgery. There is low quality evidence for conventional postoperative blood glucose control.

Early mobilization
Early mobilization is thought to be a key component of ERSS [12,254]. There is no clear definition of mobilizing, which may include simple exercise in bed, walking in the room or walking further distances [47]. The overall outcome of these pathways has been that of significant decreased length of stay; as well as improved patient satisfaction measures in selected studies [185,233,248,[255][256][257]. A study focusing on behavioral outcomes of early mobilization and rehabilitation education, identified decreased postoperative patient anxiety and enhanced self-care ability [258]. Reduced complication rates, improved patient-reported outcomes and decreased length of stay were noted in a narrative review in patients undergoing early mobilization [259].
Patients should be encouraged to mobilize actively on the day of surgery as guided by clinical condition and surgical concerns. In the absence of a clear definition of early mobilization, institutions should be encouraged to set their own benchmarks. There is moderate quality evidence for the intervention due to the imprecision in defining mobilization as well as retrospective nature of studies.

Quality of care measures 22. Audit
Systematic audit is the preferred practice pattern in order to review compliance rates with the ERAS implemented interventions [260]. There is evidence in retrospective studies that greater compliance with ERAS processes and protocols improves desired perioperative outcomes [261][262][263][264]. Full compliance with ERAS protocols has been identified to be an issue in prior studies. Compliance with ERAS pathways has been deemed to be a 5-year survival measure [265]. Overall compliance with ERAS protocols has been shown to be associated with better patient reported outcome measures [266]. There is a paucity of audit data in multimodal ERSS protocols.

Implications of this study and future directions
We have identified, delineated and presented the evidence base for first comprehensive multimodal program for Enhanced Recovery in Spinal Surgery (ERSS). A continuous issue when discussing enhanced recovery protocols is that of contention as to which components have the highest clinical utility, accompanied by somewhat arbitrary decisions on incorporating different elements into the program. We identified a high level of evidence for administration of pre-emptive analgesia, peri-operative blood conservation (antifibrinolytic use), surgical site preparation and antibiotic prophylaxis. Although evidence base for cessation of smoking in surgery of the spine is low, there is translational high level evidence from other surgical specialties. In contrast with prior ERSS reviews we identified moderate evidence base for utilization of minimally invasive surgery and use of multimodal analgesia [12]. Although early mobilization and dietary libertization are considered critical in enhanced recovery, we identified a moderate level of evidence for institution of these interventions. Some clinical units may choose to use certain aspects of this proposed perioperative program as suited best to their unique location and practice pattern.
Evidence base is low in certain research areas. Most of the studies assessed were conducted outside the context of enhanced recovery program. This may have a negative bias effect, where the effect of an individual component may be higher than estimated when used within the ERSS pathway. Their combination with other components in a particular pathway is thought to have a synergistic effect. In addition to including studies focusing on individual ERSS elements, we evaluated studies focusing on bundles of care. Through the additive incremental value of each component, this may have a positive bias towards patient care outcomes. Full compliance with ERAS protocols has been identified to be an issue in prior studies. Compliance with ERAS pathways has been deemed to be a 5-year survival measure [265]. Overall compliance with ERAS protocols has been shown to be associated with better patient reported outcome measures [266].
Methodologically heterogenous studies including systematic reviews and meta-analysis, randomized controlled trials, non-randomized controlled studies, and observational studies were eligible for this review. Hence, methodological heterogeneity of included studies rendered any quantitative effect estimates unreliable. In line with our planned protocol, we did not conduct a metaanalysis of studies in this patient population. Inherent clinical heterogeneity was present in this complex systematic review of a multi-pathway intervention. Clinical heterogeneity arose from variability in the participants, types and timing of outcome measures. Participants varied in their nature due to the type of baseline disease (e.g. ERSS for surgery on anatomically abnormal spine versus ERSS in cancer patient population). Participants also varied in inherent comorbidities e.g. young patients undergoing scoliosis surgery versus elderly with comorbidities. These multilevel participant heterogenous characteristics were combined with varied baseline analgesic consumption. Types and timing of outcome measures were compliant with our pre-determined outcome groups. As anticipated, outcome measures were broadly different to be suitable for meta-analytic process. Forest plots were obtained for some pre-determined outcomes however the statistical heterogeneity together with baseline clinical differences made these measures inappropriate for interpretation. In line with our protocol, we performed a pre-planned thematic synthesis.
We have undertaken a number of steps to minimize the underlying meta-biases in this systematic review of complex intervention. We disseminated this protocol through open literature in order to give transparency to our research structure. We assessed the risk of bias in all individual studies. Furthermore, we graded the risk of bias across outcomes [267]. As we have identified 22 components of this pathway, some selection bias due to not identifying all eligible studies was possible. Publication bias across studies, where only data published through positive findings are disseminated poses a risk in any systematic review. Detection bias may have arisen due to problems with classification of exposure or outcomes.

Acknowledgements
The authors gratefully acknowledge Mr. S. Heaton for his help in data collection.
Authors' contributions AL: study development, data collection, data editing; first draft manuscript; final version of manuscript; AS: study development, data collection, manuscript editing, final version of manuscript; HL study design, manuscript editing; JR: study design, manuscript editing; CW: study design, manuscript editing; All authors approved the final manuscript

No external funding
Availability of data and materials The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request Declarations Ethics approval and consent to participate Not applicable as this was a systematic review of available literature; Consent for publication Not applicable as this was a systematic review of available literature;

Competing interests
The authors declare that they have no competing interests.