RESUMO
BACKGROUND: In patients presenting with severe hemorrhage, the authors conducted an equivalence trial that compared noninvasive occlusion spectroscopy and the capillary blood method to determine hemoglobin level. METHODS: This prospective observational study included patients admitted to their intensive care unit for gastrointestinal bleeding. A ring-shaped sensor, connected to a NBM-200MP (OrSense, Nes Ziona, Israel), was fitted onto the patient's thumb to intermittently measure hemoglobin (SotHb). During the first 24 h after admission, venous hemoglobin level, considered as the reference method, was determined at the laboratory every 8 h and was compared to SotHb and the capillary blood method. The primary endpoint was the proportion of inaccurate measurements, defined as greater than 15% difference compared with reference values or their unavailability for any technical reason. RESULTS: The study was scheduled to include 68 patients but was stopped prematurely after an interim analysis of 34 patients. The proportion of inaccuracies revealed that SotHb could not be considered equivalent to the capillary blood method (47% [95% CI, 43-51] and 24% [95% CI, 20-28]). Considering venous hemoglobin level as a reference method, the mean biases for SotHb (n = 133) and the capillary blood method (n = 135) were, respectively, -0.4 ± 2.0 and 0.8 ± 1.2 g/dl (P < 0.05). SotHb was associated with an increased incidence of failed transfusion. The inaccuracy of SotHb tended to be increased in patients receiving vasopressor agents. CONCLUSIONS: Noninvasive determination of hemoglobin level based on occlusion spectroscopy lacks accuracy in patients presenting with severe gastrointestinal bleeding and cannot be considered equivalent to the capillary-based method. This inaccuracy seems to be moderately influenced by the infusion of vasopressor agents.
