1075 The Effect of Regional Procedural Preference on Outcomes After Posterior Decompression for Degenerative Cervical Myelopathy: Individual Participant Data Meta-Analysis of a Prospective, Multinational Dataset of 1047 Cases

Alex Beomju Bak,Ali Moghaddamjou,Paul M. Arnold,James S. Harrop,Michael G. Fehlings,Laureen D. Hachem
DOI: https://doi.org/10.1227/neu.0000000000002809_1075
IF: 5.315
2024-04-01
Neurosurgery
Abstract:INTRODUCTION: There is considerable debate regarding the relative effectiveness of laminoplasty vs. laminectomy and fusion for posterior surgical decompression of degenerative cervical myelopathy (DCM). METHODS: Subjects with DCM that underwent posterior decompression was derived from three independent, prospective, multicentre clinical trials (CSM-NA, CSM-I, CSM-Protect). Primary endpoint was change in SF36-PCS (minimum clinically important difference [MCID] 4) at 1yr compared to pre-operative assessment. Secondary endpoints were change in mJOA (MCID 2) score, Neck Disability Index (NDI; MCID 15) score, NDI Pain Intensity score, and SF36-MCS (MCID 4) score. Two comparison cohorts were created: i) laminoplasty (LP) and ii) laminectomy and fusion (LF). One-stage hierarchical mixed-effects meta-analyses with study and treatment exposure as random effects were performed. As subgroup analysis, the influence of procedural volume between LP-predominant sites, LF-predominant sites, and volume-equivalent sites was investigated. RESULTS: From a total of 1047 patients with DCM, 369 patients met eligibility criteria. There was geographic variability in surgical choice with higher rates of laminectomy and fusion in Brazil and Canada and higher laminoplasty rates in Japan and India. USA had similar rates of LF and LP. When compared to LF-predominant sites, LP-predominant sites achieved greater rates of MCID for mJOA at 1yr (85.4% v. 68.7%, p=0.036) with their predominant technique. There were no significant differences in MCID rates of NDI, SF36-PCS, or SF36-MCS. When comparing LF-predominant and LP-predominant sites with USA, a volume equivalent site, there were no significant differences in MCID rates of all outcomes. CONCLUSIONS: Sites that primarily used LP achieved greater rate of MCID in mJOA with laminoplasty than laminectomy with fusion from primarily LF sites. This suggests that experience with either LP or LF can drive outcomes which has interesting healthcare delivery implications.
surgery,clinical neurology
What problem does this paper attempt to address?