Background/Objectives: Pediatric neurosurgical procedures often involve significant blood loss and rapid hemodynamic shifts, necessitating accurate hemoglobin (Hb) monitoring. While continuous non-invasive Hb (SpHb) monitoring offers real-time trending, its accuracy in high-risk pediatric populations remains debated. We aimed to evaluate the diagnostic accuracy and clinical utility of SpHb compared to invasive arterial blood gas (ABG) analysis in pediatric patients undergoing cranial and spinal surgeries. Methods: This prospective, observational study enrolled 60 pediatric patients (aged 0–16 years) scheduled for high-risk neurosurgery. SpHb was measured continuously and compared with intermittent ABG-Hb values. Statistical analysis included Bland–Altman agreement, Pearson’s correlation, and Error Grid Analysis. Subgroup analyses assessed the impact of the Perfusion Index (PI), hypotension, and metabolic acidosis on device performance. Results: Data from 57 patients (median age: 12 months, interquartile range: 6–42 months; 70.2% aged <24 months) were analyzed. SpHb demonstrated a moderate correlation with ABG-Hb (r = 0.567, p < 0.001) but exhibited systematic overestimation with a mean bias of +1.60 ± 1.54 g/dL. Crucially, SpHb showed 0% sensitivity for detecting critical anemia (Hb < 8.0 g/dL). Device performance was significantly compromised by physiological extremes: severe metabolic acidosis significantly increased bias to +2.27 g/dL (p = 0.038), and intraoperative hypotension significantly widened the limits of agreement (SD of bias: 1.79 g/dL vs. 1.45 g/dL in normotension). Furthermore, hemodynamic analysis revealed a loss of autoregulation during hypotension, where the pressure-perfusion coupling strengthened (r = 0.44) compared to the normotensive state (r = 0.15). Conclusions: SpHb monitoring provides fair Hb trending but is limited by systematic overestimation and poor sensitivity for critical anemia. Accuracy worsens during severe acidosis and hemodynamic instability. Therefore, SpHb should function as a complementary “early warning” trend monitor rather than a sole transfusion trigger, with invasive validation remaining essential for intraoperative decision-making.
Arun et al. (Wed,) studied this question.