Accuracy of noninvasive and invasive point-of-care total blood hemoglobin measurement in an outpatient setting

Objective: The purpose of this study was to determine the accuracy of noninvasive hemoglobin measurement using pulse CO-oximetry (Pronto? SpHb?, Masimo Corp.) and a commonly used, invasive, point-of-care, automated spectrophotometer (HemoCue 201+?, HemoCue, Inc.), in comparison with hemoglobin measurements obtained from a laboratory hematology analyzer (measuring total blood hemoglobin) in the outpatient setting. Methods: Adult patients presenting to an outpatient research clinic were tested for total blood hemoglobin measurement by 3 methods: noninvasive pulse CO-oximetry (SpHb?), finger-stick blood sample on a point-of-care device, and venous sample on a laboratory hematology analyzer (reference device). Bias and standard deviation (SD) of SpHb? and HemoCue 201+? compared with the values obtained with the laboratory hematology analyzer were calculated and Bland-Altman graphs were generated. Results: Samples from 152 subjects were assessed (average age, 46 years; 69% female). The bias ± SD compared with the reference method was -0.5 ± 1.0 g/dL for SpHb? and 0.3 ±1.0 g/dL for HemoCue 201+?. The Bland-Altman plots assessing agreement of the test methods to the reference method had limits of agreement of -2.5 to 1.5 g/dL for SpHb? and -1.7 to 2.3 g/dL for HemoCue 201+?. A noninvasive measurement could not be obtained in 4 subjects after 2 attempts (2.5% failure rate), whereas the HemoCue 201+? measurements were obtained for all subjects. Conclusion: Noninvasive SpHb? testing had bias and SD similar to those of HemoCue 201+?. Because SpHb? measurement is noninvasive, it may offer additional patient and provider benefits.