EEG-Derived Wavelet Index for Assessment of Procedural Sedation in Children Undergoing Magnetic Resonance Examination: A Diagnostic Study.

BACKGROUND

Procedural sedation is often essential for children undergoing imaging examinations. Whether the wavelet index (WLI) could outperform Ramsay Sedation Score (RS) in assessing sedation depth and identifying children at risk of MRI sedation failure was unknown.

METHODS

This prospective, single-blind study was performed at a tertiary care hospital. WLI and RS were measured in 74 pediatric patients who underwent MRI scanning after procedural sedation with chloral hydrate and dexmedetomidine. Once the patient reached RS scores 5 or 6, the MRI examination was initiated. Successful MRI sedation was defined as completion of the planned MRI examination at the first attempt with no clinically significant motion artifact and no need for additional rescue sedation or conversion to general anesthesia.

RESULTS

Sixty-five children successfully completed the MRI examination, and nine failed due to insufficient sedation. The AUC for WLI was 0.902 (95% CI, 0.814-0.990) versus 0.734 (95% CI, 0.603-0.876) for RS, with a significant paired difference between the two ROC curves (ΔAUC, 0.162; 95% CI, 0.004-0.329; p = 0.038). The WLI-based model showed better overall predictive performance than the RS-based model, with a lower Brier score (0.0704 vs. 0.0908). On decision curve analysis, the WLI-based model provided greater net benefit throughout the clinically relevant threshold range (0.05-0.40).

CONCLUSIONS

WLI may provide an adjunctive, non-stimulating measure for risk stratification in pediatric MRI sedation, although the clinical utility of WLI-guided management should be confirmed in prospective studies.

TRIAL REGISTRATION: Clinical Trial Registry of China: ChiCTR2500103661.