Robot-assisted versus conventional cranial stereotactic neurosurgery: a systematic review and meta-analysis of targeting accuracy, procedural efficiency, and safety

Comparative evidence on robotic cranial stereotaxy spans different procedures, platforms and outcome definitions. We evaluated targeting accuracy, procedural efficiency, safety and clinical outcomes by surgical application. Five databases were searched through September 20, 2026. Twenty-one comparative studies were included. Accuracy was summarized using Hedges g and time using mean differences (conventional minus robotic). DerSimonian-Laird (DL)/Wald analyses were compared with alternative variance estimators and Hartung-Knapp (HK) intervals. Surgical applications and clinical outcomes were assessed separately. The exploratory three-study accuracy synthesis favored robotic assistance under DL/Wald (g=1.03, 95% CI 0.36-1.69; I2=78.5%), but its prediction interval crossed zero (- 0.189 to 2.242). REML/HK yielded g=1.03 (95% CI -0.41-2.47; P=0.092); modified HK gave a similar result. Inclusion of electrode-level SEEG also produced a confidence interval crossing zero. Two biopsy studies favored shorter procedure duration under DL/Wald (18.48 min, 95% CI 11.46-25.49), although HK intervals crossed zero. Total biopsy workflow time and intracerebral hemorrhage drainage duration were inconclusive. Risk-of-bias judgments were serious in 17 studies, critical in one, moderate in one, and some concerns in two. No consistent association between improved targeting and better clinical outcomes was established. Selected biopsy cohorts suggest technical accuracy and timing advantages with robotic assistance. Few studies, bias, heterogeneity and sensitivity to interval estimation prevent conclusions of generalized clinical superiority across cranial stereotactic applications.