BACKGROUND: Lateral lumbar interbody fusion (LLIF) carries a characteristic risk of lumbar plexus injury. Intraoperative neuromonitoring (IONM) is widely used to mitigate this risk, yet its aggregate diagnostic accuracy has not been formally quantified. METHODS: We conducted a systematic review and meta-analysis following PRISMA 2020 guidelines. PubMed, Embase, and Cochrane Library were searched through January 24, 2026. Eligible studies were prospective or retrospective observational cohorts of adults undergoing elective LLIF or XLIF with IONM correlated to postoperative neurologic outcomes and isolable 2 × 2 contingency data. Risk of bias was assessed using QUADAS-2. Analyses employed fixed-effect logistic regression, random-effects generalised linear mixed models, bivariate random-effects logistic regression, and Deeks' funnel plot asymmetry test. RESULTS: Seven studies (n = 1,994) were identified; five (n = 1,082) contributed to the primary analysis. Fixed-effect pooled sensitivity was 79.0% (95% CI: 73.0-83.9%) and specificity 95.9% (95% CI: 94.3-97.0%). The bivariate model estimated sensitivity 99.3% (95% CI: 13.7-100%) and specificity 98.8% (95% CI: 80.2-99.9%), with near-zero sensitivity-false positive rate correlation (ρ = +0.05), indicating no threshold effect. Between-study heterogeneity was high (I2 = 78.8%). Individual study sensitivity ranged from 28% to 100%, driven primarily by monitoring modality. Deeks' test was significant (t = +2.75; p = 0.035, one-tailed) but substantially underpowered at k = 5. CONCLUSIONS: IONM in LLIF demonstrates consistently high specificity but highly variable sensitivity, determined principally by monitoring modality. Multimodal protocols incorporating transcranial motor evoked potentials appear superior to unimodal electromyography. Prospective multicentre trials with standardised protocols and deficit-stratified outcomes are needed.
Velur et al. (Sat,) studied this question.