Diagnostic Accuracy of Panoramic Radiographic Signs for Predicting Inferior Alveolar Nerve Injury Following Mandibular Third Molar Surgery: A Systematic Review and Meta-Analysis

Objectives: To estimate how accurately individual panoramic radiographic signs predict postoperative inferior alveolar nerve (IAN) injury after mandibular third molar surgery, treating postoperative neurosensory injury, intraoperative nerve exposure and CBCT-defined canal contact as three separate target conditions rather than one. Methods: MEDLINE, Embase, Scopus, the Cochrane Library and Google Scholar were searched to 24 June 2025. Risk of bias was assessed with QUADAS-2 and, for randomised trials, RoB 2. Sensitivity and specificity were pooled by sign using DerSimonian–Laird random-effects models and re-examined with a bivariate random-effects model. Results: Thirty-seven studies were included; ten contributed two × two data, nine against postoperative neurosensory injury and one against intraoperative exposure. Diversion of the canal discriminated best (specificity 0.95, 95% CI 0.90–0.98; diagnostic odds ratio 8.18, 95% CI 3.31–20.20) but was insensitive (0.27). Only darkening of the root and diversion of the canal withstood removal of any single study. Negative predictive values of 94.2–96.7% added just 0.2–2.5 percentage points beyond pre-test probability, whereas diversion of the canal raised estimated risk from about 5% to 25%. Heterogeneity was substantial (I2 91–98% for specificity), no contributing study was at low risk of bias in all four QUADAS-2 domains, and the reference standard was at high risk in seven of ten. Conclusions: Panoramic signs serve risk stratification and are informative when present rather than when absent. The pooled estimates are provisional and should not be transferred to individual practice without allowance for heterogeneity and bias.