Accuracy of minimally invasive hemoglobin monitoring in the perioperative period: transcutaneous hemoglobin monitoring and point-of-care testing devices: prospective observational study.

Abstract

To evaluate the perioperative performance of two minimally/non-invasive hemoglobin monitoring modalities-continuous pulse hemoglobin (SpHb) (Revision O: rev. O) and intermittent point-of-care testing (POCT) using HemoCue® Hb801-against an automated hematology analyzer as the reference standard. Fifty patients undergoing general anesthesia with invasive arterial monitoring for surgery with anticipated blood loss were enrolled prospectively. A total of 281 arterial blood samples were analyzed by an automated hematology analyzer (reference standard), a blood gas analyzer, and HemoCue® Hb801; all comparisons, including HemoCue®, used arterial rather than capillary blood. SpHb (rev. O) was monitored continuously. Correlation and Bland-Altman analyses were performed, and polar plot analysis of changes between successive sampling points assessed SpHb trend-tracking. HemoCue® Hb801 agreed closely with the reference (r = 0.979; bias 0.119 g/dL; limits of agreement [LOA] -0.687 to +0.925 g/dL). SpHb correlated with the reference (r = 0.777) but showed wide LOA (- 2.114 to +2.811 g/dL), proportional bias, and greater random error at perfusion index <1.4 (87/281 measurements, 31.0%). In polar plot analysis of 23 successive-change pairs from 9 patients, angular bias was - 5.8° (95% CI -12.2° to +0.5°) and four-quadrant concordance 100%, but radial limits of agreement (- 36.3° to +24.6°) exceeded the ±30° criterion. HemoCue® Hb801 provided accurate Hb measurement from a minimal arterial sample. SpHb (rev. O) had limited absolute accuracy and could not quantify the magnitude of change within accepted limits, but reliably indicated its direction. A hybrid approach-continuous SpHb monitoring triggering confirmatory HemoCue® measurement-may support safer perioperative hemoglobin management.