Is cervical collar useful after laminoplasty? A randomized controlled trial

Abstract

Introduction: Cervical myelopathy is commonly treated with laminoplasty, the open-door technique being one of the most widely adopted. Rigid cervical collar immobilization has been thought to be useful after laminoplasty in protecting the hinge opening, preventing hinge fracture and spring-back phenomena. However, not only was there limited evidence on its efficacy, but it may also be detrimental to the patient’s static postural control and dynamic balance with prolonged neck movement restriction. Therefore, we propose that with the advancement of fixation devices such as the miniplates, the need for cervical collar immobilization could possibly be precluded. This could in turn lead to earlier neck mobilization and less neck stiffness, hence better patients’ outcomes. Methodology: Study Design and participants - The study was a prospective, parallel single-blinded randomized controlled trial conducted during the study period from April 2015 to February 2018. Patient recruitment was performed at two tertiary referral centers by the attending spine specialists. Inclusion criteria was any male or female patient aged over 18 years, clinical and radiological signs compatible with cervical myelopathy undergoing open-door laminoplasty for one or more spinal compression levels. Patients were required to be literate and able to comprehend the study to be enrolled. Exclusion criteria included all patients with previous cervical spine surgery, congenital deformities, spine infection or inflammation, tumor, previous spinal fusion surgery, non-Chinese ethnicity, undergoing workman’s compensation and unable or refuse to follow the standardized rehabilitation protocol. All patients were given information sheets regarding the study protocol and both verbal and written informed consents were obtained at the time of recruitment. Ethics approval was obtained from the local institutional review board (UW 15-102) and the study was registered under ClinicalTrials.gov identifier NCT03400644. There were no changes to the methods after trial commencement. Sample size calculation - Due to lack of previous available data comparing patients with collar immobilization after open-door laminoplasty, we performed a pilot study and periodically assessed the difference in axial neck pain (primary outcome) between the two groups measured by the visual analogue scale (VAS). Based on the 2 weeks postoperative data for the first 20 subjects recruited, we found that a sample size of 16 patients in each group could achieve a power of >80% with a significance level of 0.05 to detect a minimal VAS difference of 1.5 as reported as the minimal clinically important difference (MCID). Hence, based on this, we included a total of 35 patients assuming a 10% attrition rate. Randomization and masking - A total of 35 participants (15 female, 20 male) were consecutively recruited and randomized by an independent investigator using a computer program at the time of surgery to either the collar group (n=16) (to receive a Philadelphia collar for 3 weeks postoperatively) or the non-collar group (n=19) (to be allowed free mobilization without any collar immobilization). It was not possible for patients and operating surgeons to be blinded to the group allocation. However, all assessors (range of motion, outcome cores, radiographs) were blinded to the group allocation. All patients randomized to the neck collar group had the collar removed prior to range of motion and outcome assessments to maintain blinding of the assessors. Operative techniques were standardized using titanium miniplates for fixation and performed by four surgeons from the same institute. For patients allocated to the collar group, they were allowed to remove the collar temporarily whilst resting in bed. All patients were prescribed the same doses of analgesics and standardized postoperative rehabilitation protocol. Study parameters and outcome measures - All primary and secondary outcome measures were recorded immediately preoperatively (day before surgery), at 1, 2, 3 and 6 weeks postoperatively, and at 3, 6 and 12 months postoperatively. This was standardized and followed without any changes during the trial. A 1-year follow-up was deemed adequate as collar use was only an early postoperative intervention with unlikely long-term influences. The primary outcome measure was reported axial neck pain by VAS. Secondary outcome measures were divided into clinical (via standardized range of movement measurements , mJOA scores), radiological (serial lateral radiographs for anteroposterior canal diameters and any radiological complications), as well as objective scores, namely short-form 36-item questionnaire (SF-36) and neck disability index (NDI). All clinical, radiological and outcome assessments were performed by investigators blinded to the patient details. All patients were out of collar before clinical and radiographic assessments were performed to avoid bias. There was no dropouts or complications. Results and Analysis: There was no baseline differences preoperatively. Postoperatively, the only statistically significant difference lied in the reduction of postoperative axial neck pain (VAS). There was statistically significantly lower mean VAS for the collar group at postoperative 1 (3.5±2.0 versus 5.4±2.5, p=0.038), 2 (1.5±1.4 versus 3.5±2.4, p=0.028) and 3 (1.3±1.0 versus 2.8±1.9, p=0.031) weeks. Significant reductions in postoperative VAS compared to preoperative VAS were observed only at postoperative 1 (0.8±2.4 versus 3.8±3.5, p=0.016) and 2 (-0.9±3.4 versus 1.8±3.2, p=0.004) weeks for the collar group. There were no differences at postoperative 3 weeks to 12 months follow-up assessments. However, this did not translate to differences in patient-perceived outcomes via NDI and SD-36, probably owing to the predominant impact of myelopathy resolution. Both groups showed similar gradual recovery in mJOA score. There was no statistically significant differences in the cervical alignment, canal diameter and range of motion. Discussion and Conclusion: The authors conclude from the randomized controlled trial that postoperative cervical collar immobilization would not prevent complications, affect neurological recovery, or reduce cervical range of motion. It helps reduce axial neck pain only in the initial postoperative two weeks, without impact on the overall quality of life of patients. Hence, the authors recommend a temporary two-week use of rigid neck collars postoperatively only for better pain relief. Strict compliance is deemed unnecessary as there is no additional benefit.