Continuous noninvasive hemoglobin monitoring estimates timing for detecting anemia better than clinicians: A randomized controlled trial

Abstract

Background Hemoglobin measurement is important for transfusion decision-making. Pulse CO-Oximetry provides real-time continuous hemoglobin (SpHb) monitoring. The triage role of SpHb trends based on hemoglobin measurements was investigated. Methods In this diagnostic randomized controlled trial, 69 patients undergoing spine or cytoreductive surgery were randomly enrolled into SpHb-monitoring and standard-care groups. Diagnostic blood samples were drawn for CO-oximetry Hb (CoOxHb) when the SpHb decreased by 1 g/dl or at the clinician’s discretion in the standard-care group. The positive predictive value (PPV) was defined as the ability to detect a decrease in CoOxHb > 1 g/dl or a CoOxHb < 10 g/dl; the PPVs were compared using Fisher’s exact test. The SpHb and trend accuracies were calculated. The transfusion units and postoperative hemoglobin levels were compared. Results The PPV of a decrease in CoOxHb > 1 g/dl was 93.3% in the SpHb group vs 54.5% without SpHb monitoring (p = 0.002). The PPV of CoOxHb < 10 g/dl was 86.7% vs. 50.0% for these groups (p = 0.015). The CoOxHb was never < 7 g/dl with SpHb monitoring. Sixty SpHb–CoOxHb data pairs and 28 delta pairs (ΔSpHb-ΔCoOxHb) were collected. The bias, precision and limits of agreement were -0.29, 1.03 and -2.30 to 1.72 g/dl, respectively. When ΔSpHb and ΔCoOxHb were > 1 g/dl, the concordance rate for changes in hemoglobin reached 100%. The delta pairs revealed a positive correlation [ΔSpHb = 0.49 * ΔCoOxHb - 0.13; r = 0.69, 95% confidence interval (0.53, 0.82)]. No significant differences were found in the transfusion volume or postoperative anemia state. Conclusions The SpHb trend tracked changes in hemoglobin satisfactorily during surgery and more accurately estimated the appropriate timing for invasive hemoglobin measurements than the clinicians. Trial Registration ChiCTR1800016290, Date of registration