In the June 21, 2017 edition of The Journal of Bone & Joint Surgery, Kato et al. add to the evidence base regarding this question. They report on results from an analysis comparing the two approaches in 80 pairs of “propensity-matched” patients who had multilevel compression myelopathy. Propensity matching allowed the authors to adjust for multiple baseline factors and MRI characteristics, thus minimizing the risk of selection bias.
After the propensity-matched analysis, there were no two-year between-group differences in mJOA score, Neck Disability Index, or SF-36 Physical Component score. The overall rates of perioperative complications were similar between the two groups, although dysphagia and dysphonia were reported only in the anterior group, while surgical site infection and C5 radiculopathy were reported only in the posterior group.
The authors claim that propensity matching helps to “reflect the ‘real-world’ clinical setting and likely has greater generalizability than a smaller, narrowly randomized controlled trial,” but they ultimately conclude that the surgical approach in such cases “should be carefully chosen by evaluating risk profiles in a shared decision-making process on a case-by-case basis.”