Response to Letter to the Editor for “Impact of Concomitant Grade I Degenerative Spondylolisthesis on Clinical and Radiographic Outcomes Following Biportal Endoscopic Versus Microscopic Laminectomy for Lumbar Spinal Stenosis: A Post Hoc Analysis of a Multicenter Randomized Clinical Trial”

We thank the correspondents for four points that go to the heart of how a small exploratory subgroup should be read, and have reanalyzed the trial data to answer each. ¹

Our radiographic models were not complete-case analyses: linear mixed-effects models are likelihood-based, retain every observed visit, and give valid estimates under a missing-at-random mechanism without imputation.^2,3 We nonetheless repeated them using multiple imputation by chained equations (m = 30; completeness among the 40 patients with grade I spondylolisthesis was 97–100% preoperatively and 60% at 12 months). Estimates were essentially unchanged: the 12-month between-technique difference was −0.26 percentage points (95% CI, −1.69 to 1.17) for slip percentage and 0.11° (95% CI, −3.35 to 3.57) for the instability ratio, versus −0.38 and 0.14° as observed, and no interaction became significant. Imputation attenuated the within-arm slip increase to nonsignificance while the increase in angular motion persisted — a signal we flagged rather than dismissed.

On the equivalence margin, the value applied was not derived post hoc: ±12.8 points is the equivalence margin of the parent trial, used for its sample size calculation, ⁴ and the visual analog scale and EuroQol-5 Dimension thresholds were reported in our Methods. What was not prespecified was a formal equivalence hypothesis for these subgroup comparisons. Applying that margin, all four 12-month confidence intervals fell entirely within it, with significant two one-sided tests (Oswestry Disability Index, −2.98 [−12.69 to 6.72] against ±12.8, p = 0.024; the other three outcomes, p ≤ 0.013). We present this as supportive rather than confirmatory. ⁵

On the third point we agree. Individual slip trajectories were reassuring: the median change was 0.4 points and no patient progressed beyond 5 points. Segmental motion is the more relevant signal — two of 24 evaluable patients met level-specific instability thresholds at 12 months, one in each arm and both marginally (20.9° and 21.4° against a 20° threshold), though neither underwent fusion or revision, and a small number of patients may genuinely have destabilized. Moreover, conventional flexion–extension radiography is itself insensitive for segmental instability, ⁶ so this count likely understates the true proportion. We did not advance this as a finding because it was not statistically supported; nonetheless, we accept the correspondents’ reading that a population-averaged estimate can conceal individual progression.

On the fourth point we also agree, and it is why we reported the analysis as we did. Of the 97 between-group and interaction tests, only one reached raw p < 0.05 (a baseline characteristic), and none survived Benjamini-Hochberg or Bonferroni correction. ⁷ Because uncorrected comparisons in a subgroup of this size invite over-interpretation, we reported the conservative reading that no difference was detected.

We therefore share the correspondents’ central caution. Our position is that decompression alone is an acceptable option in stable grade I spondylolisthesis and that these data do not distinguish the two techniques, while segmental motion does increase modestly in both arms and more than modestly in occasional patients — warranting vigilance and longer follow-up rather than blanket reassurance. Full analyses are available on request.

ORCID iDs

Hyun-Jin Park https://orcid.org/0000-0002-8075-2813

Min-Seok Kang https://orcid.org/0000-0002-9773-4486

Kwang-Sup Song https://orcid.org/0000-0002-9238-8908

Dae-Woong Ham https://orcid.org/0000-0003-4278-2701

Ho-Joong Kim https://orcid.org/0000-0002-8205-4648

Jin-Sung Kim https://orcid.org/0000-0001-5086-0875

Sang-Min Park https://orcid.org/0000-0001-6171-3256

Ethical Considerations

The original RCTs were approved by the institutional review boards of the participating hospitals (B-2102/666-007, B-2102/667-003) and registered at cris.nih.go.kr (KCT0006057). The present post hoc analysis was approved by the Institutional Review Board (B-2502-957-101).

Data Availability Statement

Data supporting the findings of this study are available from the corresponding author upon request.*