Advertisement

Efficacy, safety, and economics of bracing after spine surgery: a systematic review of the literature

  • Author Footnotes
    1 These authors have contributed equally to this work.
    Mary P. Zhu
    Footnotes
    1 These authors have contributed equally to this work.
    Affiliations
    Division of Neurosurgery, Department of Surgery, St. Michael's Hospital, University of Toronto, 30 Bond St, Toronto, Ontario M5B 1W8, Canada
    Search for articles by this author
  • Author Footnotes
    1 These authors have contributed equally to this work.
    Lindsay A. Tetreault
    Footnotes
    1 These authors have contributed equally to this work.
    Affiliations
    Division of Neurosurgery, Department of Surgery, Toronto Western Hospital, University of Toronto, 399 Bathurst St, Toronto, Ontario M5T 2S8, Canada

    Graduate Entry Medicine, School of Medicine, University College Cork, Brookfield, College Rd, Cork, T12 K8AF, Ireland
    Search for articles by this author
  • Fatimah Sorefan-Mangou
    Affiliations
    Graduate Entry Medicine, School of Medicine, University College Cork, Brookfield, College Rd, Cork, T12 K8AF, Ireland
    Search for articles by this author
  • Philip Garwood
    Affiliations
    Graduate Entry Medicine, School of Medicine, University College Cork, Brookfield, College Rd, Cork, T12 K8AF, Ireland
    Search for articles by this author
  • Jefferson R. Wilson
    Correspondence
    Corresponding author. Division of Neurosurgery, Department of Surgery, St. Michael's Hospital, University of Toronto, 30 Bond St, Toronto, Ontario M5B 1W8, Canada. Tel.: + 6476684497.
    Affiliations
    Division of Neurosurgery, Department of Surgery, St. Michael's Hospital, University of Toronto, 30 Bond St, Toronto, Ontario M5B 1W8, Canada

    Li Ka Shing Knowledge Institute, St. Michael's Hospital, University of Toronto, 209 Victoria St, Toronto, Ontario M5B 1T8, Canada
    Search for articles by this author
  • Author Footnotes
    1 These authors have contributed equally to this work.
Published:January 17, 2018DOI:https://doi.org/10.1016/j.spinee.2018.01.011

      Abstract

      Background Context

      Bracing is often used after spinal surgery to immobilize the spine, improve fusion, and relieve pain. However, controversy exists regarding the efficacy, necessity, and safety of various bracing techniques in the postsurgical setting.

      Purpose

      In this systematic review, we aimed to compare the effectiveness, safety, and cost-effectiveness of postoperative bracing versus no postoperative bracing after spinal surgery in patients with several common operative spinal pathologies.

      Study Design/Setting

      A systematic review was carried out to compare postoperative bracing and no postoperative bracing.

      Methods

      A systematic search was conducted of MEDLINE, Embase, and the Cochrane Collaboration Library from 1970 to May 2017, supplemented by manual searching of the reference list of relevant studies and previously published reviews. Studies were included if they compared disability, quality of life, functional impairment, radiographic outcomes, cost-effectiveness, or complications between patients treated with postoperative bracing and patients not receiving any postoperative bracing. Each article was critically appraised independently by two reviewers, and the overall body of evidence was rated using guidelines outlined by the Grading of Recommendation Assessment, Development and Evaluation (GRADE) Working Group.

      Results

      Of the 858 retrieved citations, 5 studies met the inclusion criteria and were included in this review, consisting of 4 randomized controlled trials and 1 prospective cohort study. Low to moderate evidence suggests that there are no significant differences in most measures of disability, pain, quality of life, functional impairment, radiographic outcomes, and safety between groups. Isolated studies reported statistically significant and inconsistent differences between groups with respect to Neck Disability Index at 6 weeks postoperatively or Short Form-36 Physical Component Score at 1.5, 3, 6, and 12 months postoperatively.

      Conclusions

      Based on limited evidence, postoperative bracing does not result in improved outcomes after spinal surgery. Future high-quality randomized trials will be required to confirm these findings.

      Keywords

      Introduction

      Bracing is routinely used after surgery for a number of spinal pathologies, including degenerative disease of the lumbar and cervical spine, thoracolumbar fractures, and scoliosis [
      • Connolly P.J.
      • Grob D.
      Bracing of patients after fusion for degenerative problems of the lumbar spine—yes or no?.
      ]. The intended goal of this practice is to immobilize the spine, relieve pain, improve fusion rates, and remind patients to avoid certain activities that may compromise their recovery [
      • Connolly P.J.
      • Grob D.
      Bracing of patients after fusion for degenerative problems of the lumbar spine—yes or no?.
      ,
      • Soliman H.A.
      • Barchi S.
      • Parent S.
      • Maurais G.
      • Jodoin A.
      • Mac-Thiong J.M.
      The early impact of postoperative bracing on pain and quality of life following posterior instrumented fusion for lumbar degenerative conditions: a randomized trial.
      ]. However, a number of important complications can arise from bracing, including dysphagia, nerve palsies, pressure ulcers, and skin rashes [
      • Connolly P.J.
      • Grob D.
      Bracing of patients after fusion for degenerative problems of the lumbar spine—yes or no?.
      ]. Furthermore, braces can be uncomfortable for some patients and costly [
      • Soliman H.A.
      • Barchi S.
      • Parent S.
      • Maurais G.
      • Jodoin A.
      • Mac-Thiong J.M.
      The early impact of postoperative bracing on pain and quality of life following posterior instrumented fusion for lumbar degenerative conditions: a randomized trial.
      ].
      Evidence & Methods

      Context

      The authors performed a systematic review to assess the possible benefits of bracing following spine surgery.

      Contribution

      Using improved methods over previous similar studies, they arrived at similar conclusions, ie, that bracing does not appear to make an impact.

      Implications

      The conclusions of systematic reviews are only as good as the studies included. Diversity in patient pathologies, other confounding medical conditions and surgical techniques in the available data limit validity. Accordingly, based on this and other studies, we can conclude that bracing generally does not appear to make an impact except on the individual/symptomatic level while acknowledging that there may be subgroups as yet insufficiently studied who may benefit.
      Given the paucity of high-quality comparative studies, it is unclear whether postoperative bracing can effectively limit and restrict spinal movements, reduce rates of pseudoarthrosis, and optimize patient recovery. Certain advances in spinal surgery have allowed rigid internal stabilization of the spine and, arguably, have decreased the requirement for external immobilization. Although these techniques may be sufficient to achieve successful fusion, there may still be a role for postoperative bracing in higher risk patients, including those who smoke, suffer from osteoporosis, or require an extensive multilevel surgery.
      As a result of the limited body of evidence available, spine surgeons often based their decision to use postoperative bracing on their own clinical experience and training [
      • Connolly P.J.
      • Grob D.
      Bracing of patients after fusion for degenerative problems of the lumbar spine—yes or no?.
      ]. This finding is supported by a survey that highlighted substantial disagreement among spinal surgeons with respect to the appropriate type, duration, and indication for use of postoperative bracing after anterior cervical spine surgery [
      • Bible J.E.
      • Biswas D.
      • Whang P.G.
      • Simpson A.K.
      • Rechtine G.R.
      • Grauer J.N.
      Postoperative bracing after spine surgery for degenerative conditions: a questionnaire study.
      ]. Given the heterogeneity in management strategies, there is a need to synthesize results from high-quality studies and establish recommendations surrounding care after spinal surgery.
      This systematic review addresses four key questions (KQs). KQ1: What is the efficacy and effectiveness of postoperative bracing compared with no bracing based on disability, pain, quality of life, and functional outcomes? KQ2: What is the impact of postoperative bracing compared with no bracing on radiographic outcomes? KQ3: What is the safety profile of postoperative bracing compared with no bracing? KQ4: What is the cost-effectiveness of postoperative bracing? Importantly, this systematic review will assess the overall strength of the evidence using methodology developed by the Grading of Recommendation Assessment, Development and Evaluation (GRADE) Working Group.

      Methods

      Eligibility criteria

      Population

      Our review targeted studies including patients undergoing surgery for any spinal pathology, including cervical and lumbar degenerative disease, trauma, oncology, and adolescent idiopathic scoliosis. Studies were excluded if patients under study were treated non-operatively (Table 1).
      Table 1Inclusion and exclusion criteria for studies reviewed
      CharacteristicInclusionExclusion
      PopulationPatients with any spinal pathology treated surgically and followed postoperatively (eg, scoliosis, spinal trauma, cervical myelopathy)Patients treated non-operatively
      InterventionPostoperative bracing
      ComparisonNo postoperative bracing

      Standard of care
      OutcomeKQ1: Efficacy and effectiveness

      • Disability (eg, NDI)
      • Pain (eg, VAS)
      • Quality of life (eg, SF-36)
      • Functional outcomes (eg, JOA, mJOA, Nurick)


      KQ2: Radiographic outcomes

      • Fusion rate
      • Sagittal alignment
      • ROM


      KQ3: Safety

      • Complications or adverse events (eg, dysphagia, hardware, skin complications, reoperation, revision surgery, infection, hematoma)


      KQ4: Cost-effectiveness

      • Incremental cost-effectiveness ratio (or similar)
      • Cost per unit of outcome
      KQ1: Subjective neurologic status, patient satisfaction, improvement of symptoms
      Study designKQ1, 2, 3: Comparative studies (eg, RCT, prospective cohort, case-control studies) designed to compare a postoperative bracing group with a control group; n≥10 per group

      KQ4: Full economic studies
      Case reports

      Non-clinical studies

      Animal studies
      PublicationStudies published in peer-review journals and in EnglishAbstracts, editorials, letters

      Duplicate publications of the same study that do not report on different outcomes

      Narrative or systematic reviews
      JOA, Japanese Orthopedic Association; mJOA, modified Japanese Orthopedic Association; NDI, Neck Disability Index; RCT, randomized controlled trial; ROM, range of motion; SF-36, Short Form-36; VAS, visual analog scale.

      Intervention and comparison

      This review focused on studies that had an intervention group who received postoperative bracing and a control group who received standard of care and no postoperative bracing (Table 1).

      Outcomes

      For KQ1, we sought studies that considered the clinical efficacy of postoperative bracing by measuring patient disability, pain, quality of life, or functional outcomes. For KQ2, we focused on studies that assessed radiographic outcomes, including fusion rate, sagittal alignment, and range of motion (ROM). For KQ3, we sought studies that compared complication rates and adverse events between the intervention and control groups. For KQ4, we focused on studies that examined various measures of cost-effectiveness including incremental cost-effectiveness ratio and cost per unit of outcome (Table 1).

      Study characteristics

      For KQ1, 2, and 3, we sought comparative studies (ie, randomized controlled trials, cohort studies) designed to evaluate differences between a postoperative bracing group and a control group. To be included, studies needed to have at least 10 patients per group. Case reports, non-clinical studies, and animal studies were excluded. For KQ4, we focused on full economic studies. For all KQs, abstracts, editorials, letters, narrative, and systematic reviews were excluded. Duplicate publications of the same study that did not report on different outcomes were also excluded (Table 1).

      Information sources

      A systematic search of MEDLINE, Embase, and Cochrane Collaboration Library was conducted to identify relevant studies. Manual searching of the reference lists of included studies and previously published reviews was also conducted to ensure all relevant studies were located.

      Search strategy

      The search strategy was first developed in MEDLINE and then appropriately modified for the other databases. We used the following search terms to search all databases: Orthotic Devices AND Spinal Diseases AND Postoperative Complications/Care AND Treatment Outcome or Outcome Assessment. Only studies involving humans, written in English, and published in peer-reviewed journals between 1970 and May 2017 were considered for inclusion, with no other limits applied. A detailed search strategy is provided in the Supplementary material.

      Study selection

      All abstracts and titles were reviewed and sorted by our predefined inclusion criteria. Studies were classified as relevant, possibly relevant, or irrelevant. Full-text investigation of all relevant and possibly relevant studies was done for further clarification.

      Data extraction and synthesis

      The following data were extracted from each included article: study design; patient sample and characteristics, including diagnosis, surgical summary, and type of bracing; outcome assessment tools; follow-up schedule; dropout rate; and results of association, including standard deviation, odds ratio, confidence intervals, and p-values (Table 2).
      Table 2Characteristics of included studies
      Authors and yearPopulation, nSex, %Mean age (range, SD)
      Age in years.
      DiagnosisSurgeryBracingOutcome factors assessedFollow-upDropout, n
      Dropout before end of follow-up.
      Abbott et al., 2013 (RCT, pilot)Brace (n=17)M, 53%53.4 (NR, 13)Cervical spondylosis (n=8), cervical disk herniation (n=4), cervical degenerative disc disease (n=5)ACDF with interbody cagePhiladelphia collar daytime only for 6 weeksKQ1: Borg CR-10 (neck and arm pain), NDI, SF-36 (PCS and MCS), FES, unipedal balance standing test

      KQ2: CROM, fusion rate, sagittal alignment

      KQ3: None

      KQ4: None
      1.5, 3, 6, 12, and 24 months postoperative55% (n=18)
      Control (n=16)M, 69%47.3 (NR, 11)Cervical spondylosis (n=5), cervical disk herniation (n=4), cervical degenerative disc disease (n=7)
      Campbell et al., 2008 (Prospective cohort)
      Postoperative care, including immobilization techniques and activity restrictions, was left to the discretion of the attending surgeon.
      Brace (n=149)M, 43.6%44.3 (NR, 8.8)Single-level radiculopathy or myelopathyDecompression and arthrodesis using allograft and anterior cervical plate or arthroplastyCervical collarKQ1: NDI, neck and arm pain scales, SF-36

      KQ2: Fusion success

      KQ3: Second procedure
      Included revisions, removals, reoperations, supplemental fixations, and external bone growth stimulators.
      , instrumentation failure, graft extrusion

      KQ4: None
      1.5, 3, 6, 12, and 24 months postoperativeNR
      Control (n=108)M, 49.1%43.3 (NR, 9.0)
      Christodoulou et al., 1987 (RCT)Brace (n=25)NANRAdolescent idiopathic scoliosis with thoracic curves≥35°Posterior decompression and fusion with Harrington instrumentation augmented by a Cotrel bar or by sublaminal Luque wiresPlaster body cast for 6 monthsKQI: None

      KQ2: Spinal curve

      KQ3: None

      KQ4: None
      3, 6, 12, and 24 months postoperativeNR
      Control (n=25)NANR
      Hida et al., 2017 (RCT)Brace (n=45)M, 73%72.0 (NR, 8.7)Cervical myelopathy secondary to multisegmental cervical spondylotic stenosisDouble-door cervical laminoplasty without instrumentationPhiladelphia collar for 2 weeksKQ1: VAS, JOA, SF-36 (PCS and MCS)

      KQ2: ROM (total, extension, and flexion), lordotic angle C2-7

      KQ3: Perioperative complications
      Included surgical site infection, epidural hematoma, and C5 palsy.


      KQ4: None
      0.5, 3, 6, and 12 months postoperative18% (n=16)
      Control (n=45)M, 62%71.6 (NR, 9.6)
      Yee et al., 2008 (RCT)Brace (n=46)M, 43%
      Based on patients who completed follow-up.
      52 (NR, 15.2)
      Based on patients who completed follow-up.
      Spondylosis/stenosis (n=11), degenerative spondylolisthesis (n=13), isthmic spondylolisthesis (n=8), junctional syndrome (n=1), pseudarthrosis (n=2), iatrogenic/post-op instability (n=2)
      Included later-stage revision surgery caused by symptomatic non-union or later-stage hardware removal caused by prominence or bursitis.
      Posterior lumbar spinal arthrodesisCanvas lumbar corset full time for 8 weeksKQ1: DPQ, SF-36

      KQ2: Fusion rate

      KQ3: Revision surgery
      Included later-stage revision surgery caused by symptomatic non-union or later-stage hardware removal caused by prominence or bursitis.
      , complications
      Included intraoperative incidental durotomy, intraoperative pedicle-screw-placement issues, new postoperative radiculopathy, early postoperative pulmonary embolism, wound seroma/hematoma, deep wound infection, or persistent lumbar radiculopathy postoperatively.


      KQ4: None
      12 and 24 months postoperative20% (n=18)
      Control (n=44)M, 54%
      Based on patients who completed follow-up.
      53 (NR, 15.4)
      Based on patients who completed follow-up.
      Spondylosis/stenosis (n=12), degenerative spondylolisthesis (n=12), isthmic spondylolisthesis (n=5), junctional syndrome (n=2), pseudarthrosis (n=2), iatrogenic/post-op instability (n=1), congenital stenosis (n=1)
      Included later-stage revision surgery caused by symptomatic non-union or later-stage hardware removal caused by prominence or bursitis.
      ACDF, anterior cervical discectomy and fusion; CROM, cervical range of motion; DPQ, Dallas Pain Questionnaire; FES, Falls Efficacy Scale; JOA, Japanese Orthopedic Association; MCS, Mental Component Score; NDI, Neck Disability Index; NR, not reported; PCS, Physical Component Score; RCT, randomized controlled trial; ROM, range of motion; SD, standard deviation; SF-36, Short Form-36; VAS, visual analog scale.
      * Age in years.
      Dropout before end of follow-up.
      Postoperative care, including immobilization techniques and activity restrictions, was left to the discretion of the attending surgeon.
      § Included revisions, removals, reoperations, supplemental fixations, and external bone growth stimulators.
      Included surgical site infection, epidural hematoma, and C5 palsy.
      Based on patients who completed follow-up.
      # Included later-stage revision surgery caused by symptomatic non-union or later-stage hardware removal caused by prominence or bursitis.
      ** Included intraoperative incidental durotomy, intraoperative pedicle-screw-placement issues, new postoperative radiculopathy, early postoperative pulmonary embolism, wound seroma/hematoma, deep wound infection, or persistent lumbar radiculopathy postoperatively.

      Risk of bias in individual studies

      The class of evidence for each article was rated (Class I, II, III, IV) independently by two reviewers using criteria outlined by the Journal of Bone and Joint Surgery for therapeutic studies and modified to encompass both methodological quality and risk of bias. Randomized controlled trials were rated based on patient allocation, intention-to-treat analysis, independent or blinded assessment, whether co-interventions were applied equally, rates of follow-up, statistical power, and control for possible confounding. Prospective cohort studies were rated based on independent or blinded assessment, whether co-interventions were applied equally, rates of follow-up, statistical power, and control for possible confounding. Because of the nature of the intervention, studies were rated as having independent or blinded assessment if surgeons were blinded to the randomization group until after surgery, patients were blinded to the randomization group until day of admission of surgery or after surgery, or radiologists reviewing radiographs were blinded to the randomization group.

      Risk of bias across studies

      The overall body of evidence was assessed using a scoring system developed by the GRADE Working Group with recommendations from the Agency for Healthcare Research and Quality. This methodology allows an assessment of the overall strength of the evidence and is particularly valuable for highlighting critical knowledge gaps.
      The initial strength of the overall body of evidence was graded as “high” if half or more of the studies were randomized controlled trials and “low” if the majority of studies were observational studies. The body of evidence was downgraded 1, 2, or 3 levels if there was risk of bias, results were inconsistent or consistency was unknown, the evidence was indirect, the effect estimates were imprecise (eg, wide confidence intervals), or if there was publication bias. If no downgrades were made, the body of evidence was upgraded 1, 2 or 3 levels based on large magnitude of effect, dose-response gradient, or if all plausible biases would decrease the magnitude of an apparent effect.
      The final rating of the body of evidence expresses our confidence in the estimate of effect and the impact of further research on this topic. An overall strength of “high” means we have high confidence that the evidence reflects the true effect. Further research is very unlikely to change our confidence in the estimate of effect. The overall strength of “moderate” means we have moderate confidence that the evidence reflects the true effect. Further research may change our confidence in the estimate of effect and may change the estimate. A grade of “low” means we have low confidence that the evidence reflects the true effect. Further research is likely to change the confidence in the estimate of effect and likely to change the estimate. A grade of “insufficient” means that evidence is either unavailable or does not permit a conclusion.

      Results

      Study selection

      The initial electronic search yielded a total of 853 citations. Five additional citations were identified through reference scanning. After initial review of abstracts and titles, 841 studies did not meet our inclusion criteria. Following full-text investigation, an additional 12 studies were excluded because (1) they were not comparative studies; (2) patients were not treated surgically; (3) there was no postoperative comparison of intervention and control groups; (4) they had a different outcome of interest; (5) they had no control group; (6) they were a duplicate publication with no new results; or (7) they were not in English. A total of five studies were deemed relevant after this review process (Figure).

      Study characteristics

      For KQ1, we identified four studies (three randomized controlled trials, one prospective cohort) discussing the effect of postoperative bracing on disability, pain, quality of life, and functional outcomes [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ,
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ]. Sample sizes ranged from 33 to 257 surgical patients, with mean ages between 43.9 and 72.7 years. All patients were diagnosed with degenerative cervical myelopathy or radiculopathy, or degenerative disease of the lumbar spine. Bracing included Philadelphia collars, cervical collars, and lumbar corsets for differing lengths of time. Various outcome measures were used across the studies, with the Short Form-36 (SF-36) Physical Component Score (PCS) reported the most frequently (n=4) [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ,
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ], followed by the SF-36 Mental Component Score (MCS) (n=3) [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ,
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ], SF-36 subscales (n=3) [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ,
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ], Neck Disability Index (NDI) (n=2) [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ], neck pain (n=2) [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ], and arm pain (n=2) [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ].
      For KQ2, a total of five studies (four randomized controlled trials, one prospective cohort) met our inclusion criteria [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ,
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ,
      • Christodoulou A.G.
      • Prince H.G.
      • Webb J.K.
      • Burwell R.G.
      Adolescent idiopathic thoracic scoliosis. A prospective trial with and without bracing during postoperative care.
      ]. These studies were designed to assess the impact of postoperative bracing compared with no bracing on radiographic outcomes. Sample sizes ranged from 33 to 257 surgical patients, with mean ages between 14.3 and 72.7 years. Patients were diagnosed with degenerative cervical myelopathy or radiculopathy, degenerative disease of the lumbar spine, or adolescent idiopathic scoliosis. Bracing included Philadelphia collars, cervical collars, body casts, and lumbar corsets for various lengths of time. Fusion rate was the most frequently reported outcome measure (n=3) [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ].
      For KQ3, we identified three studies (two randomized controlled trials, one prospective cohort) examining the safety profile of postoperative bracing compared with no bracing [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ,
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ]. Sample sizes ranged from 50 to 257 surgical patients with mean ages between 43.9 and 72.7 years. Patients were diagnosed with degenerative cervical myelopathy or radiculopathy, or degenerative disease of the lumbar spine. Bracing included Philadelphia collars, cervical collars, and lumbar corsets for various lengths of time. Outcome measures included complications (n=3) [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ,
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ] and revision surgery or second procedure (n=2) [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ].
      No studies met our inclusion criteria for KQ4 on the cost-effectiveness of postoperative bracing.

      Risk of bias

      We critically appraised the five studies included in our review. The inter-rater reliability was 80%; disagreement on the fifth study, by Yee et al., surrounding the “intention to treat” analysis was resolved through discussion. It was noted that although 90 patients were randomized, only 72 were included in their analysis, and therefore the study did not use intention-to-treat analysis [
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ]. Of the studies included, four were considered Class II and one was rated Class III. The four Class II studies were randomized controlled trials, and were downgraded from Class I because they did not include intent-to-treat analysis, independent or blind assessment, adequate sample size, random sequence generation or statement of concealed allocation, had unreported follow-up rates or follow-up rates<80%, or did not control for possible confounders. The Class III study was a prospective cohort study that was downgraded from Class II because co-interventions were not applied equally and follow-up was not reported.

      Results of individual studies

      KQ1: What is the efficacy and effectiveness of postoperative bracing compared with no bracing based on disability, pain, quality of life, and functional outcomes?

      Degenerative cervical myelopathy or radiculopathy

      Three studies compared disability, pain, quality of life, or functional outcomes between the postoperative bracing and non-bracing groups in patients with degenerative cervical myelopathy or radiculopathy [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ].
      According to Abbott et al. [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ], patients who received postoperative bracing had better NDI scores at 6 weeks after surgery compared with patients who did not receive postoperative bracing (mean difference between groups −4.4, 95% confidence interval [CI] −8.6 to −0.2, p=.042). However, Campbell et al. [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ] reported that patients in the non-braced group had better NDI scores at 6 weeks after surgery compared with patients in the braced group (p=.008).
      Abbott et al. [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ] also found that patients in the control group had better SF-36 PCS scores at 6 weeks (mean difference between groups 5.8, 95% CI 0.8–10.7, p=.025), 3 months (mean difference between groups 6.8, 95% CI 0.4–13.1, p=.038), 6 months (mean difference between groups 7.4, 95% CI 1.4–13.4, p=.017), and 12 months (mean difference between groups 7.5, 95% CI 0.3–14.6, p=.041) after surgery compared with patients in the postoperative bracing group. In contrast, Campbell et al. and Hida et al. found no significant differences in SF-36 PCS scores between the two groups at all time points assessed (Campbell et al. 6 months p=.481, 12 months p=.260, 24 months p=.279; Hida et al. p=.537) [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ].
      Two studies evaluated differences in various subscales of the SF-36 between a postoperative bracing group and a control group [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ]. A single study by Abbott et al. reported significantly better SF-36 Bodily Pain (BP) scores in the postoperative bracing group at 6 (mean difference between groups 21.4, 95% CI 4.4–38.5, p=.016) and 12 (mean difference between groups 17.5, 95% CI 1.7–33.2, p=.031) months, as well as better SF-36 Social Functioning (SF) scores at 12 months (mean difference between groups 16.5, 95% CI 0.1–32.9, p=.049) than in the non-bracing group [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ]. The other six subscales (Physical Functioning [PF], General Health [GH], Role Limitations Physical [RP], Vitality [VT], Role Limitations Emotional [RE], and Mental Health [MH]) were not significantly different between treatment groups at all time points assessed (1.5, 3, 6, 12, and 24 months) [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ]. A second study by Hida et al. [
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ] identified no differences between the collar-fixation group and the control group with respect to SF-36 BP subscale (p=.848).
      Other measures of disability, pain, quality of life, and functional impairment, including SF-36 MCS, visual analog scale, Japanese Orthopedic Association recovery rate, Falls Efficacy Scale, unipedal balance standing test, and neck and arm pain were not significantly different between the postoperative bracing and the non-bracing groups (Table 3) [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ].
      Table 3Results of statistical analysis
      Authors and yearStatistical analysisDifferences at baselinePostoperative differences
      Abbott et al., 2013ANCOVA with adjustment for covariates, repeated measure analysis of covarianceThere were no significant differences between the cervical collar group and the control group with respect to gender, age, diagnosis, level of operation, cervical range of motion, unipedal balance and baseline NDI, SF-36 (subscales, PCS and MCS), FES, and Borg CR-10 (neck and arm pain) scores (p=NR)ANCOVA:

      NDI scores at 1.5 months (p=.042)
      No significant differences at all other time points assessed (1.5, 3, 6, 12, and 24 months postoperative).


      Mean difference between groups: −4.4 (95% CI: −8.6 to −0.2); Cohen's effect size: −0.77

      SF-36 BP scores

      6 months (p=.016)
      No significant differences at all other time points assessed (1.5, 3, 6, 12, and 24 months postoperative).


      Mean difference between groups: 21.4 (95% CI: 4.4 to 38.5); Cohen's effect size: 0.73

      12 months (p=.031)
      No significant differences at all other time points assessed (1.5, 3, 6, 12, and 24 months postoperative).


      Mean difference between groups: 17.5 (95% CI: 1.7 to 33.2); Cohen's effect size: 0.57

      SF-36 SF scores at 12 months (p=.049)
      No significant differences at all other time points assessed (1.5, 3, 6, 12, and 24 months postoperative).


      Mean difference between groups: 16.5 (95% CI: 0.1 to 32.9); Cohen's effect size: 0.45

      SF-36 PCS

      1.5 months (p=.025)
      No significant differences at all other time points assessed (1.5, 3, 6, 12, and 24 months postoperative).


      Mean difference between groups: 5.8 (95% CI: 0.8 to 10.7); Cohen's effect size: 0.84

      3 months (p=.038)
      No significant differences at all other time points assessed (1.5, 3, 6, 12, and 24 months postoperative).


      Mean difference between groups: 6.8 (95% CI: 0.4 to 13.1); Cohen's effect size: 0.63

      6 months (p=.017)
      No significant differences at all other time points assessed (1.5, 3, 6, 12, and 24 months postoperative).


      Mean difference between groups: 7.4 (95% CI: 1.4 to 13.4); Cohen's effect size: 0.80

      12 months (p=.041)
      No significant differences at all other time points assessed (1.5, 3, 6, 12, and 24 months postoperative).


      Mean difference between groups: 7.5 (95% CI: 0.3 to 14.6); Cohen's effect size: 0.66

      There were no significant differences between the cervical collar group and the control group with respect to

      (1) neck pain, arm pain, FES, SF-36 PF, SF-36 RP, SF-36 GH, SF-36 VT, SF-36 RE, SF-36 MH, SF-36 MCS at all time points assessed (1.5, 3, 6, 12, and 24 months) (p>.05)

      (2) all components of the unipedal balance test (right/ foot, right/left foot soft surface, right/left foot eyes closed) at all time points assessed (1.5, 3, 6, 12, and 24 months) (p>.05)

      (3) all components of CROM (right/left lateral flexion, flexion, extension, right/left rotation) at all time points assessed (1.5, 3, 6, 12, and 24 months) (p>.05)

      (4) fusion rates and sagittal alignment (p=NR)

      Repeated measures analysis of covariance showed that controlling for the combined effects of all prospective measures gave significantly better outcome for the cervical collar group in neck pain (p=.038), SF-36 PCS (p=.010), and SF-36 BP (p=.029).
      Campbell et al., 2008ANOVA, ANCOVA, Fisher exact test, Student t testSF-36 PCS was higher in the control group (31.1±7.2) than in the braced group (33.4±7.8) (p=.019)

      There were no significant differences between the braced group and the control group with respect to age (p=.367), gender (p=.447), worker's compensation (p=.458), litigation (p=1.000), smoking status (p=1.000) and occupational status (p=.695), and baseline NDI (p=.141), neck pain (p=.523), and arm pain (p=.710) scores
      Mean improvement in NDI scores at 1.5 months (p=.008)

      Braced: 21.6±18.4

      Not braced: 28.4±19.0

      There were no significant differences between the braced group and the control groups with respect to NDI scores at 3 months (p=.468), 6 months (p=.169), 12 months (p=.415) and 24 months (p=.693), SF-36 PCS at 6 months (p=.481), 12 months (p=.260) and 24 months (p=.279), average neck pain scores at 24 months (p=.622), and average arm pain scores at 24 months (p=.260).

      There were no significant differences in fusion success at any time period between groups, although higher rates of fusion were reported in the non-braced group (p=.552 at 24 months).

      There were no significant differences in rates of secondary surgeries or procedures between the braced group and the control group: revisions (p=.653), removals (p=.724), reoperations (p=1.000), supplemental fixations (p=.286), any surgery (p=.184).
      Christodoulou et al., 1987NRMean curves (p=NR)

      Braced: 58.0°

      Not braced: 54.0°
      Postoperative mean curves (p=NR)

      Braced: 23.0°

      Not braced: 22.8°

      There was no significant difference between the braced group and the control group with respect to the mean loss of correction at 24 months (7.0° for braced group, 6.3° for control group, p=NR).
      Hida et al., 2017Two-way repeated ANOVA, Fisher exact test, Student t test, chi-square testThere were no significant differences between the collar-fixation group and the control group with respect to age (p=.73), gender (p=.26), height (p=.59), weight (p=.66), operation time (p=.57), intraoperative blood loss (p=.69), number of operated levels (p=.67), VAS (p=.33), JOA (p=.67), lordotic angle (p=.84), ROM (p=.88), SF-36 PCS (p=.68), MCS (p=.80), and BP (p=.57).There were no significant differences between the collar-fixation group and the control group with respect to

      (1) VAS (p=.487), JOA recovery rates (p=.80), SF-36 PCS (p=.537), SF-36 MCS (p=.504), and SF-36 BP subscores (p=.848) at 12 months follow-up

      (2) the decrease in the C2-7 lordotic angle (p=.82) and ROM (p=.61)

      (3) incidence of complications (p=.53)
      Yee et al., 2008Mann-Whitney U test, chi-square test, Fisher exact test, two-way ANOVAThere were no significant differences between the brace group and the control group with respect to age (p=.97), gender (p=.35), CCI (p=.6), number of levels included in the arthrodesis (p=.42), smoking status (p=.89), worker's compensation or litigation (p=.48), revision surgery (p=.88), BMI (p=.74), diagnosis (p=.8), and preoperative SF-36 MCS (p=.9), SF-36 PCS (p=.19), SF-36 domain scores (p>.05), and DPQ category scores (p>.05)There were no significant differences between the brace group and the control group with respect to

      (1) the distribution of surgical complications or subsequent revision rates (p=.8)

      (2) postoperative DPQ category scores (p=.34 for the daily activity category, p=.67 for the work/leisure category, p=.17 for the anxiety-depression category and p=.40 for the social category)

      (3) postoperative SF-36 domain and component scores (p=.38 for PF, p=.28 for BP, p=.23 for GH, p=.41 for RP, p=.25 for VT, p=.79 for SF, p=.86 for RE, p=.30 for MH, p=.30 for PCS and p=.57 for MCS)

      (4) rates of fusion seen radiographically at 12 months (p=.8) or 24 months (p=.9) postoperatively

      (5) Rates of revision surgery caused by symptomatic non-union (p=.43)
      ANCOVA, analysis of covariance; ANOVA, analysis of variance; BMI, body mass index; BP, bodily pain; CCI, Charlson Comorbidity Index; CI, confidence interval; CROM, cervical range of motion; DPQ, Dallas Pain Questionnaire; FES, Falls Efficacy Scale; GH, general health; JOA, Japanese Orthopedic Association; MCS, Mental Component Score; MH, mental health; NDI, Neck Disability Index; NR, not reported; OR, odds ratio; PCS, Physical Component Score; PF, physical functioning; RE, role limitations emotional; ROM, range of motion; RP, role limitations physical; SF, social functioning; SF-36, Short Form-36; VAS, visual analog scale; VT, vitality.
      * No significant differences at all other time points assessed (1.5, 3, 6, 12, and 24 months postoperative).

      Degenerative disease of the lumbar spine

      A single study by Yee et al. examined pain and quality of life outcomes for patients with a degenerative disease of the lumbar spine [
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ]. There were no significant differences in the Dallas Pain Questionnaire (DPQ; daily activity category p=.34, work or leisure category p=.67, anxiety depression category p=.17, social category p=.40), SF-36 PCS (p=.30), SF-36 MCS (p=.57), and SF-36 subscales (PF p=.38, BP p=.28, GH p=.23, RP p=.41, VT p=.25, SF p=.79, RE p=.86, MH p=.30) between the braced and the non-braced groups at all time points assessed [
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ].

      KQ2: What is the impact of postoperative bracing compared with no bracing on radiographic outcomes?

      Degenerative cervical myelopathy or radiculopathy

      Three studies assessed the radiographic outcomes between the postoperative bracing and the non-bracing groups in patients with degenerative cervical myelopathy or radiculopathy [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ].
      Abbott et al. [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ] reported no significant differences in cervical range of motion (p>.05), fusion rate (p=NR [not reported]), and sagittal alignment (p=NR) between the two groups. Campbell et al. [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ] also found no significant differences in fusion success between the bracing and the non-bracing groups (p=NR). Similarly, Hida et al. [
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ] concluded that there were no significant differences in ROM (p=.61) or decrease in lordotic angle C2-7 (p=.82) between groups.

      Degenerative disease of the lumbar spine

      Yee et al. [
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ] found no significant differences in fusion rate at 12 (p=.8) or 24 months (p=.9) postoperatively between patients who wore a lumbar corset and those who did not.

      Adolescent idiopathic scoliosis

      Based on a single study, there was no significant difference in mean loss of spinal curve correction between the braced and the non-braced groups at all time points assessed in patients with adolescent idiopathic scoliosis (p=NR) [
      • Christodoulou A.G.
      • Prince H.G.
      • Webb J.K.
      • Burwell R.G.
      Adolescent idiopathic thoracic scoliosis. A prospective trial with and without bracing during postoperative care.
      ].

      KQ3: What is the safety profile of postoperative bracing compared with no bracing?

      Degenerative cervical myelopathy or radiculopathy

      Campbell et al. [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ] reported rates of instrumentation failure, graft extrusion, and second procedures including revisions, removals, reoperations, supplemental fixations, and external bone growth stimulators, whereas Hida et al. [
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ] considered all perioperative complications including surgical site infection, epidural hematoma, and C5 palsy.
      Both Campbell et al. and Hida et al. identified no significant differences in the incidence of complications between the bracing and the non-bracing groups (Campbell et al. no events of instrumentation failure or graft extrusion in either group, p=NC; Hida et al. p=.53) [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ]. In addition, Campbell et al. [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ] found no significant differences between groups in the rate of revision surgery (p=.653), removals (p=.724), reoperations (p=1.000), supplemental fixations (p=.286), or any second operation (p=.184).

      Degenerative disease of the lumbar spine

      Based on a single study, there were no significant differences in rates of revision surgery, a second procedure, or complications between the braced and the non-braced groups at all time points assessed (p=.8) [
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ]. Revision surgery included later-stage revision surgery caused by symptomatic non-union or later-stage hardware removal caused by prominence or bursitis, and complications included intraoperative incidental durotomy, intraoperative pedicle-screw-placement issues, new postoperative radiculopathy, early postoperative pulmonary embolism, wound seroma or hematoma, deep wound infection, or persistent lumbar radiculopathy postoperatively.

      KQ4: What is the cost-effectiveness of postoperative bracing?

      No studies were identified that evaluated the cost-effectiveness of postoperative bracing.

      Summary of evidence

      The overall quality of evidence ranged from “insufficient” to “moderate.” Evidence was downgraded because of risk of bias, unknown or inconsistent consistency of results, or imprecise effect estimates (eg, wide confidence intervals) (Table 4).
      Table 4Summary and strength of evidence
      Patient population and outcome of interestStudiesStrength of evidenceOverall effect and conclusions
      KQ1: What is the efficacy and effectiveness of postoperative bracing compared with no bracing based on disability, pain, quality of life, and functional outcomes?
      Degenerative cervical myelopathy or radiculopathy
       NDI

       SF-36 PCS

       SF-36 MCS

       SF-36 BP subscale

       SF-36 SF subscale

       SF-36 other subscales

       VAS

       JOA recovery rate

       FES

       Unipedal balance standing test

       Neck pain

       Arm pain
      n=2
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.


      n=3
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.


      n=2
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.


      n=2
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.


      n=1
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.


      n=1
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.


      n=1
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.


      n=1
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.


      n=1
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.


      n=1
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.


      n=2
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.


      n=2
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      Insufficient

      Insufficient

      Low

      Insufficient

      Insufficient

      Insufficient

      Low

      Low

      Insufficient

      Insufficient

      Moderate

      Moderate
      Results were inconsistent across studies: (1) a prospective cohort study reported that patients in the non-braced group had better NDI scores at 6 weeks after surgery compared to patients in the braced group (Campbell et al., 2009), whereas (2) a pilot randomized controlled trial indicated that patients in the braced group had better NDI scores at 6 weeks after surgery compared to patients in the non-braced group (Abbott et al., 2013).

      Results were inconsistent across studies: (1) a pilot randomized controlled trial indicated that patients in the non-braced group had better SF-36 PCS scores at 6 weeks, 3, 6, and 12 months after surgery compared with patients in the braced group (Abbott et al., 2013), whereas (2) two studies (a prospective cohort study and a randomized controlled trial) reported no significant differences in SF-36 PCS scores between the braced and non-braced groups at all time points assessed (Campbell et al., 2009, Hida et al., 2017).

      Results were inconsistent across studies: (1) a pilot randomized controlled trial indicated that patients in the braced group had better SF-36 BP subscale scores at 6 and 12 months after surgery compared with patients in the non-braced group (Abbott et al., 2013), whereas (2) a randomized controlled trial reported no significant differences in SF-36 BP subscale scores between the braced and non-braced groups at all time points assessed (Hida et al., 2017).

      Patients in the braced group had better SF-36 SF subscale scores at 12 months after surgery compared to patients in the non-braced group (Abbott, Halvorsen et al. 2013).

      There were no significant differences in SF-36 MCS and other subscales, VAS, JOA recovery rate, FES, unipedal balance standing test, neck pain, and arm pain between the braced and non-braced groups at all time points assessed.
      Degenerative disease of the lumbar spine
       DPQn=1
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      LowThere were no significant differences in DPQ, SF-36 PCS, MCS, and subscales between the braced and the non-braced groups at all time points assessed.
       SF-36 PCSn=1
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      Low
       SF-36 MCSn=1
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      Low
       SF-36 subscalesn=1
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      Low
      KQ2: What is the impact of postoperative bracing compared with no bracing on radiographic outcomes?
      Degenerative cervical myelopathy or radiculopathy
       ROMn=2
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      LowThere were no significant differences in ROM, fusion rate, sagittal alignment, and decrease in lordotic angle C2-7 between the braced and the non-braced groups at all time points assessed.
       Fusion raten=2
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      Moderate
       Sagittal alignmentn=1
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      Low
       Lordotic angle C2-7n=1
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      Low
      Degenerative disease of the lumbar spine
       Fusion raten=1
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      LowThere was no significant difference in fusion rate between the braced and the non-braced groups at all time points assessed.
      Adolescent idiopathic scoliosis with thoracic curves≥35°
       Spinal curven=1
      • Christodoulou A.G.
      • Prince H.G.
      • Webb J.K.
      • Burwell R.G.
      Adolescent idiopathic thoracic scoliosis. A prospective trial with and without bracing during postoperative care.
      LowThere was no significant difference in mean loss of spinal curve correction between the braced and the non-braced groups at all time points assessed.
      KQ3: What is the safety profile of postoperative bracing compared with no bracing?
      Degenerative cervical myelopathy or radiculopathy
       Revision surgery or second proceduren=1
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      LowThere were no significant differences in rate of revision surgery, second procedure, or complications between the braced and the non-braced groups at all time points assessed.
       Complicationsn=2
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      Moderate
      Degenerative disease of the lumbar spine
       Revision surgery or second proceduren=1
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      LowThere were no significant differences in rate of revision surgery, second procedure or complications between the braced and the non-braced groups at all time points assessed
       Complicationsn=1
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      Low
      KQ4: What is the cost-effectiveness of postoperative bracing?
      Nonen=0NANone
      CROM, cervical range of motion; DPQ, Dallas Pain Questionnaire; FES, Falls Efficacy Scale; JOA, Japanese Orthopedic Association; MCS, Mental Component Score; NDI, Neck Disability Index; NA, not applicable; PCS, Physical Component Score; ROM, range of motion; SF-36, Short Form-36; VAS, visual analog scale.
      Based on low evidence, postoperative bracing in patients with degenerative cervical myelopathy or radiculopathy does not result in improved SF-36 MCS, visual analog scale, Japanese Orthopedic Association recovery rate, sagittal alignment, ROM, decrease in lordotic angle C2-7, and rate of revision surgery or second procedure outcomes. Based on moderate evidence, postoperative bracing does not result in improved neck or arm pain, fusion rate, and incidence of complications in this patient population.
      Low evidence suggests no improvement in DPQ, SF-36 PCS, SF-36 MCS, SF-36 subscales, fusion rate, incidence of complications, and rate of revision surgery or second procedure after postoperative bracing in patients with a degenerative disease of the lumbar spine.
      Based on low evidence, postoperative bracing is not associated with improved loss of spinal curve correction in adolescents with idiopathic scoliosis.

      Discussion

      The use of bracing after surgery for a variety of spinal pathologies remains controversial with limited evidence available to the surgeon to make an informed decision. Historically cited reasons to use bracing include to limit mobility and stabilize the spine, improve fusion rates, prevent graft dislodgement or subsidence, reduce postoperative pain, and optimize outcomes. Bracing, however, can be uncomfortable, lead to social isolation, and be associated with complications such as dysphagia, nerve palsies, and pressure ulcers. This review summarizes the current literature on the efficacy, safety, and cost-effectiveness of bracing after spinal surgery using rigorous methodology and is the first to synthesize results using methodology proposed by the GRADE Working Group. This knowledge is valuable in a clinical setting and can be used by clinicians to determine the most appropriate postoperative management strategies. Clinical judgment, however, is still required to determine whether a patient may benefit from additional external immobilization.

      Degenerative cervical myelopathy or radiculopathy

      Based on this review, postoperative bracing of the cervical spine has no impact on (1) most measures of pain, disability, functional impairment, and quality of life; (2) radiographic outcomes such as fusion rate and ROM; and (3) rates of complications and reoperations. In the study by Abbott et al. [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ], patients who received postoperative bracing exhibited superior improvements in neck disability (NDI at 6 weeks) and various metrics of quality of life (SF-36 PCS at 6 weeks to 12 months, SF-36 SF at 12 months, and SF-36 BP at 6–12 months) than those who did not. These findings can be partly explained by psychological factors, including a sense of security provided by the brace, increased coping mechanisms, improved functional self-efficacy, and less fear avoidance [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ]. In contrast, Campbell et al. [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ] identified that patients with postoperative bracing had worse NDI scores, likely caused by the discomfort and disability associated with wearing a brace.
      Previous studies in healthy subjects have demonstrated that cervical bracing reduces velocity of eye movements and causes deterioration in the anterior to posterior body sway induced by vibration of the calf muscles [
      • Burl M.M.
      • Williams J.G.
      • Nayak U.S.
      Effects of cervical collars on standing balance.
      ,
      • Karlberg M.
      • Magnusson M.
      • Johansson R.
      Effects of restrained cervical mobility on voluntary eye movements and postural control.
      ]. Given these findings, it is hypothesized that restricting cervical motion through external immobilization may significantly impair static postural control and disturb balance during dynamic movement. In the study by Abbott et al. [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ], however, there were no differences between a bracing group and a control group with respect to the unipedal balance standing test.
      Biomechanical studies have indicated that cervical collars help to restrict motion during routine activities and stabilize the spine [
      • DiPaola C.P.
      • Sawers A.
      • Conrad B.P.
      • Horodyski M.
      • DiPaola M.J.
      • Del Rossi G.
      • et al.
      Comparing cervical spine motion with different halo devices in a cadaveric cervical instability model.
      ,
      • Koller H.
      • Zenner J.
      • Hitzl W.
      • Ferraris L.
      • Resch H.
      • Tauber M.
      • et al.
      In vivo analysis of atlantoaxial motion in individuals immobilized with the halo thoracic vest or Philadelphia collar.
      ,
      • Schneider A.M.
      • Hipp J.A.
      • Nguyen L.
      • Reitman C.A.
      Reduction in head and intervertebral motion provided by 7 contemporary cervical orthoses in 45 individuals.
      ]. However, early mobilization exercises can prevent spine contracture and improve ROM after surgery [
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ]. In studies by Hida et al. and Abbott et al., there were no significant differences in cervical range of motion between the postoperative bracing group and the control group [
      • Abbott A.
      • Halvorsen M.
      • Dedering A.
      Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
      ,
      • Hida T.
      • Sakai Y.
      • Ito K.
      • Ito S.
      • Imagama S.
      • Ishiguro N.
      • et al.
      Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
      ]. Although ROM often decreases after surgery, this is more likely due to fusion and fixation techniques, damage to the cervical flexors and extensors and injury to the facet joints. Postoperative bracing may also help to decrease the risk of graft or cage migration, maintain spinal alignment, and improve fusion rates. Advancements in surgical procedures, however, have allowed internal stabilization of the spine and may have decreased the requirement for external immobilization. For example, the use of anterior plates has shown to increase fusion and decrease subsidence rates by limiting motion between the graft and vertebral bodies [
      • Yue W.M.
      • Brodner W.
      • Highland T.R.
      Long-term results after anterior cervical discectomy and fusion with allograft and plating: a 5- to 11-year radiologic and clinical follow-up study.
      ,
      • Wang J.C.
      • McDonough P.W.
      • Endow K.
      • Kanim L.E.
      • Delamarter R.B.
      The effect of cervical plating on single-level anterior cervical discectomy and fusion.
      ]. This finding was confirmed by Campbell et al. who reported no significant difference in rates of fusion between bracing and non-bracing groups [
      • Campbell M.J.
      • Carreon L.Y.
      • Traynelis V.
      • Anderson P.A.
      Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
      ]; these results question the need for postoperative bracing, especially in patients undergoing internal stabilization. There may still be a role for postoperative bracing in patients at a higher risk of pseudoarthrosis and disease progression, including those who smoke, have had a previous spine operation, or are treated without rigid internal fixation [
      • Hilibrand A.S.
      • Fye M.A.
      • Emery S.E.
      • Palumbo M.A.
      • Bohlman H.H.
      Impact of smoking on the outcome of anterior cervical arthrodesis with interbody or strut-grafting.
      ,
      • Yan C.
      • Avadhani N.G.
      • Iqbal J.
      The effects of smoke carcinogens on bone.
      ,
      • Kusin D.J.
      • Ahn U.M.
      • Ahn N.U.
      The effect of smoking on spinal cord healing following surgical treatment of cervical myelopathy.
      ,
      • Tetreault L.A.
      • Kopjar B.
      • Vaccaro A.
      • Yoon S.T.
      • Arnold P.M.
      • Massicotte E.M.
      • et al.
      A clinical prediction model to determine outcomes in patients with cervical spondylotic myelopathy undergoing surgical treatment: data from the prospective, multi-center AOSpine North America study.
      ,
      • Lau D.
      • Chou D.
      • Ziewacz J.E.
      • Mummaneni P.V.
      The effects of smoking on perioperative outcomes and pseudarthrosis following anterior cervical corpectomy: clinical article.
      ,
      • Lee J.C.
      • Lee S.H.
      • Peters C.
      • Riew K.D.
      Adjacent segment pathology requiring reoperation after anterior cervical arthrodesis: the influence of smoking, sex, and number of operated levels.
      ]. Furthermore, surgeons are more likely to use postoperative bracing after a multilevel anterior cervical discectomy and fusion (76%) compared with a single-level anterior cervical discectomy and fusion (55%) [
      • Bible J.E.
      • Biswas D.
      • Whang P.G.
      • Simpson A.K.
      • Rechtine G.R.
      • Grauer J.N.
      Postoperative bracing after spine surgery for degenerative conditions: a questionnaire study.
      ].

      Degenerative disease of the lumbar spine

      Based on this review, the use of a lumbar corset after surgery for degenerative lumbar disease has no impact on pain, disability, functional impairment, quality of life, radiographic outcomes, incidence of complications, and rate of reoperations. Postoperative bracing is often used in this population to relieve pain, limit mobility, improve fusion rates, and optimize outcomes; however, the study by Yee et al. [
      • Yee A.J.
      • Yoo J.U.
      • Marsolais E.B.
      • Carlson G.
      • Poe-Kochert C.
      • Bohlman H.H.
      • et al.
      Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
      ] indicated no advantage or disadvantage to the use of a lumbar corset. There may still be a role for bracing in patients at a higher risk of non-union or pseudoarthrosis, such as those who smoke or require a multilevel fusion.

      Adolescent idiopathic scoliosis

      Historically, molded plaster braces were used to correct the curve after posterior fusions without instrumentation and maintain this correction until solid bony fusion. Techniques proposed in the study by Christodoulou et al. [
      • Christodoulou A.G.
      • Prince H.G.
      • Webb J.K.
      • Burwell R.G.
      Adolescent idiopathic thoracic scoliosis. A prospective trial with and without bracing during postoperative care.
      ], including the use of Harrington distraction rods, however, have decreased the need for postoperative external bracing because of more rigid internal fixation. In this review, a single study examined postoperative bracing in an adolescent idiopathic scoliosis population and found no differences in radiographic outcomes between patients who received bracing versus those who did not. Further investigation, however, is needed to determine the effectiveness of bracing in this population based on other outcome measures.
      Our finding that bracing may not confer additional benefits after spine surgery will have relevant applications in a clinical setting. First, complications such as skin reactions, dysphagia, pressure ulcers, and nerve palsies, as well as costs associated with bracing, can be eliminated. Second, if postoperative bracing is not required, there may be less of an impact on activities of daily living, decreased social isolation, body anxiety, self-perception and body image issues, and an improved ability to return to work or school after surgery.

      Strengths and limitations

      To our knowledge, no other reviews have evaluated the merits of bracing in the postsurgical setting for patients with various spinal pathologies using the GRADE approach. This methodology allows rigorous evaluation of the overall strength of the evidence and helps to identify critical knowledge gaps in the literature (eg, a lack of high-quality comparative studies and limited data on the cost-effectiveness of postoperative bracing). Furthermore, the majority of current studies published on this topic have moderately high to high risk of bias and imprecise estimates of effect (or estimates with unknown precision). Consistency of results is also largely unknown as results are based on single studies.
      Our review has also its limitations. First, our search was restricted to studies published in English and, as a result, some articles with relevant titles or abstracts were excluded. Second, although results were separated based on patient population, the type and length of bracing, as well as surgical technique, varied substantially among studies, thus preventing pooling of data and meta-analysis.

      Conclusions

      Based on the results of this review, postoperative bracing does not result in improved outcomes after spine surgery in patients with various spinal pathologies. Although some outcomes were significantly different between the bracing and the non-bracing groups, firm conclusions cannot be made because of small sample sizes, risk of bias, and low quality of evidence. Finally, given the paucity of studies available, no conclusions can be made regarding the cost-effectiveness of bracing after surgery.

      Supplementary material

      The following is the supplementary data to this article:

      References

        • Connolly P.J.
        • Grob D.
        Bracing of patients after fusion for degenerative problems of the lumbar spine—yes or no?.
        Spine. 1998; 23: 1426-1428
        • Soliman H.A.
        • Barchi S.
        • Parent S.
        • Maurais G.
        • Jodoin A.
        • Mac-Thiong J.M.
        The early impact of postoperative bracing on pain and quality of life following posterior instrumented fusion for lumbar degenerative conditions: a randomized trial.
        Spine. 2018; 43: 155-160
        • Bible J.E.
        • Biswas D.
        • Whang P.G.
        • Simpson A.K.
        • Rechtine G.R.
        • Grauer J.N.
        Postoperative bracing after spine surgery for degenerative conditions: a questionnaire study.
        Spine J. 2009; 9: 309-316
        • Abbott A.
        • Halvorsen M.
        • Dedering A.
        Is there a need for cervical collar usage post anterior cervical decompression and fusion using interbody cages? A randomized controlled pilot trial.
        Physiother Theory Pract. 2013; 29: 290-300
        • Campbell M.J.
        • Carreon L.Y.
        • Traynelis V.
        • Anderson P.A.
        Use of cervical collar after single-level anterior cervical fusion with plate: is it necessary?.
        Spine. 2009; 34: 43-48
        • Hida T.
        • Sakai Y.
        • Ito K.
        • Ito S.
        • Imagama S.
        • Ishiguro N.
        • et al.
        Collar fixation is not mandatory after cervical laminoplasty: a randomized controlled trial.
        Spine. 2017; 42: E253-E259
        • Yee A.J.
        • Yoo J.U.
        • Marsolais E.B.
        • Carlson G.
        • Poe-Kochert C.
        • Bohlman H.H.
        • et al.
        Use of a postoperative lumbar corset after lumbar spinal arthrodesis for degenerative conditions of the spine. A prospective randomized trial.
        J Bone Joint Surg Am. 2008; 90: 2062-2068
        • Christodoulou A.G.
        • Prince H.G.
        • Webb J.K.
        • Burwell R.G.
        Adolescent idiopathic thoracic scoliosis. A prospective trial with and without bracing during postoperative care.
        J Bone Joint Surg Br. 1987; 69: 13-16
        • Burl M.M.
        • Williams J.G.
        • Nayak U.S.
        Effects of cervical collars on standing balance.
        Arch Phys Med Rehabil. 1992; 73: 1181-1185
        • Karlberg M.
        • Magnusson M.
        • Johansson R.
        Effects of restrained cervical mobility on voluntary eye movements and postural control.
        Acta Otolaryngol. 1991; 111: 664-670
        • DiPaola C.P.
        • Sawers A.
        • Conrad B.P.
        • Horodyski M.
        • DiPaola M.J.
        • Del Rossi G.
        • et al.
        Comparing cervical spine motion with different halo devices in a cadaveric cervical instability model.
        Spine. 2009; 34: 149-155
        • Koller H.
        • Zenner J.
        • Hitzl W.
        • Ferraris L.
        • Resch H.
        • Tauber M.
        • et al.
        In vivo analysis of atlantoaxial motion in individuals immobilized with the halo thoracic vest or Philadelphia collar.
        Spine. 2009; 34: 670-679
        • Schneider A.M.
        • Hipp J.A.
        • Nguyen L.
        • Reitman C.A.
        Reduction in head and intervertebral motion provided by 7 contemporary cervical orthoses in 45 individuals.
        Spine. 2007; 32: E1-E6
        • Yue W.M.
        • Brodner W.
        • Highland T.R.
        Long-term results after anterior cervical discectomy and fusion with allograft and plating: a 5- to 11-year radiologic and clinical follow-up study.
        Spine. 2005; 30: 2138-2144
        • Wang J.C.
        • McDonough P.W.
        • Endow K.
        • Kanim L.E.
        • Delamarter R.B.
        The effect of cervical plating on single-level anterior cervical discectomy and fusion.
        J Spinal Disord. 1999; 12: 467-471
        • Hilibrand A.S.
        • Fye M.A.
        • Emery S.E.
        • Palumbo M.A.
        • Bohlman H.H.
        Impact of smoking on the outcome of anterior cervical arthrodesis with interbody or strut-grafting.
        J Bone Joint Surg Am. 2001; 83-A: 668-673
        • Yan C.
        • Avadhani N.G.
        • Iqbal J.
        The effects of smoke carcinogens on bone.
        Curr Osteoporos Rep. 2011; 9: 202-209
        • Kusin D.J.
        • Ahn U.M.
        • Ahn N.U.
        The effect of smoking on spinal cord healing following surgical treatment of cervical myelopathy.
        Spine. 2015; 40: 1391-1396
        • Tetreault L.A.
        • Kopjar B.
        • Vaccaro A.
        • Yoon S.T.
        • Arnold P.M.
        • Massicotte E.M.
        • et al.
        A clinical prediction model to determine outcomes in patients with cervical spondylotic myelopathy undergoing surgical treatment: data from the prospective, multi-center AOSpine North America study.
        J Bone Joint Surg Am. 2013; 95: 1659-1666
        • Lau D.
        • Chou D.
        • Ziewacz J.E.
        • Mummaneni P.V.
        The effects of smoking on perioperative outcomes and pseudarthrosis following anterior cervical corpectomy: clinical article.
        J Neurosurg Spine. 2014; 21: 547-558
        • Lee J.C.
        • Lee S.H.
        • Peters C.
        • Riew K.D.
        Adjacent segment pathology requiring reoperation after anterior cervical arthrodesis: the influence of smoking, sex, and number of operated levels.
        Spine. 2015; 40: E571-E577