Electromagnetic navigation and ventricular catheter placement: a propensity-score-matched analysis of early proximal revision and parenchymal misplacement.

Abstract

Ventricular catheter placement is fundamental to hydrocephalus management, but freehand techniques carry risk of malplacement and early revision. Single-centre EM versus freehand comparisons are often confounded. We assessed whether EM guidance is associated with lower early revision and misplacement using propensity-score matching. We reviewed 1,847 ventricular catheter placements at Erasmus MC (2017-2023). The primary endpoint was 10-day proximal revision in a variable-ratio PSM cohort targeting up to four controls per EM placement (exact match on etiology and age category; propensity score on age, sex, and entry point). Secondary endpoints included 30-day revision, Hayhurst grade, and a ventricle-adjusted ordinal model. Sensitivity analyses added for this revision and an E-value were computed. A parallel prospective cohort (N = 16) served as a Post-Market Clinical Follow-up pilot. After PSM (N = 401), 10-day revision occurred in 5.5% of EM-guided versus 17% of freehand placements (OR 0.31, 95% CI 0.12-0.82, p = .018); in pediatric patients, revision was 3.8% versus 25% (OR 0.12, 0.02-0.80, p = .028). Thirty-day revision was 14% versus 28% (OR 0.45, 0.21-0.94, p = .034), and Grade 3 misplacement was 3.1% versus 16% (OR 0.16, p = .004). Ventricle-size-adjusted ordinal modelling was consistent (OR 0.56, p = .056). Sensitivity analyses were directionally consistent (OR 0.21-0.32); the E-value was 3.01 for the point estimate and 1.45 for the confidence limit. The prospective pilot showed 75% Grade 1 and no device-related serious adverse events. EM-guided placement was associated with lower odds of early revision and parenchymal misplacement in a propensity-matched analysis. A multicentre randomised trial is warranted before routine universal adoption.