Total hemoglobin (tHb) is one the most frequently ordered laboratory measurements in both acute and outpatient settings.1 Current laboratory methods are invasive, time consuming, and provide only intermittent hemoglobin measurements. Noninvasive and continuous hemoglobin measurement would offer many advantages in the assessment of both acute and chronic anemic status in a variety of clinical settings. Pulse CO-Oximetry™ (Masimo Corp., Irvine, CA) is a multi-wavelength spectrophotometric method that has recently received Food and Drug Administration 510(k) clearance for noninvasive, immediate, and continuous hemoglobin monitoring (SpHb).a The purpose of this study was to compare simultaneous measurements of hemoglobin using noninvasive pulse CO-Oximetry and invasive laboratory CO-Oximetry (tHb) in normal subjects undergoing hemodilution.
English-speaking male and nonpregnant female volunteer subjects between 18 and 30 years of age were eligible for the study if they were categorized as ASA physical status I (normal, healthy subject without organic, physiologic, or psychiatric disturbances) or II (subject with controlled medical conditions without significant systemic effects). Subjects were excluded if they had hemoglobin <12 g/dL, known drug or alcohol abuse, skin abnormalities affecting the digits, experienced a head injury in the previous year, known neurologic and psychiatric disorders, known chronic usage of psychoactive or anticonvulsive drugs within 90 days or any use within the previous 7 days, hypertension, baseline heart rate <50 bpm, or any medical condition which, in the judgment of the investigator, rendered them inappropriate for participation in the study. The study was approved by an independent IRB, which adheres to the Declaration of Helsinki and United States Code of Federal Regulations. All patients provided written, informed consent to the study procedures as outlined in the protocol.
After screening but before enrollment, subjects underwent a medical history and physical examination with preprocedure vital signs taken. An anesthesiologist was present throughout the procedure. Before hemodilution, continuous electrocardiogram and oscillometric blood pressure monitoring was initiated and recorded throughout the procedure for patient safety purposes. A first-generation Pulse CO-Oximeter (Masimo Radical-7 running board software version 7211; Masimo Corp.) with a spectrophotometric adhesive sensor (Masimo Rainbow R1 25 and R1 25L) using 12 wavelengths of light was used to measure SpHb. Data were continuously recorded and downloaded to a computer running data collection software (Physiolog; Masimo Corp.). Up to 3 sensors were placed on 1 or both hands of each subject (index, middle, or ring finger). Optical interference was prevented through the use of certified optical shields over each sensor. Perfusion index was also measured automatically and simultaneously with SpHb from the Pulse CO-Oximeter. Perfusion index is the pulsatile signal indexed against the nonpulsatile signal and is an indication of localized perfusion.
A peripheral IV catheter was placed in the subject's upper extremity. After venous access was established, an intraarterial catheter was placed in the radial artery to facilitate continuous blood pressure monitoring, for removal of aliquots of blood to induce anemia, and to allow for frequent determination of tHb. Measurement of tHb throughout the procedure was obtained by taking an arterial blood sample from the arterial catheter and analyzing the blood with a laboratory CO-Oximeter (model ABL820; Radiometer, Copenhagen, Denmark). The laboratory CO-Oximeter was calibrated each day according to manufacturer's specifications. The reported bias of the ABL820 for hemoglobin concentration of the blood samples used in our study compared with the hemiglobincyanide (HiCN) method is between 0.04 and 0.37.2
The Pulse CO-Oximeter in our study has the capability to measure either venous or arterial hemoglobin concentration. We chose to use arterial blood samples for comparison because arterial blood is the “gold standard” analyte for blood gas measurement.
One unit of blood, approximately 500 mL, was drawn from the subject through the arterial or venous catheter. The subject then rapidly received a multi-electrolyte crystalloid (Isolyte®; B. Braun Medical, Inc., Irvine, CA) IV to compensate for the decrease in intravascular volume and further reduce hemoglobin concentration. Subjects continued to receive IV fluid until they reached the goal of 30% reduction in hemoglobin or had received a maximum of 30 mL/kg IV fluid. Arterial samples were drawn after each approximately 500 mL of fluid administered. The process was stopped if there was any suggestion of subject distress, including but not limited to a blood pressure decrease of >30% from baseline, an increase in heart rate to >130 bpm, or complaints of feeling faint, gastrointestinal stress, or alteration of mental status. At the conclusion of the procedure, the venous and arterial catheters were removed and the subject was monitored for safety and then discharged. The unit of blood initially removed was not reinfused. Twenty-four hours after the procedure, subjects returned for a physical examination to exclude any complications.
Device and patient data were collected by study coordinators and entered into a database in which statistical analysis was performed (SigmaPlot and SigmaStat; Systat Software, Inc., San Jose, CA). SpHb data were excluded from analysis if the Pulse CO-Oximeter device indicated a low signal quality index, an automatic determination that is made by the device using a combination of factors including waveform amplitude and signal-to-noise ratios. Data are expressed as mean ± 1 SD and n (%). For regression statistics, significance was determined at P < 0.05.
SpHb accuracy was analyzed by pairing the SpHb measurement obtained 20 seconds after the start of the invasive blood draw with the resulting tHb test result. Bias, precision, and the average root-mean-square error were calculated. The association of perfusion index to SpHb accuracy was analyzed by linear regression, correlating perfusion index to the absolute value of the difference between SpHb and tHb.
Between April 21 and May 21, 2008, 20 subjects were enrolled at a single, 911-bed academic hospital (Loma Linda University Medical Center, Loma Linda, CA). Eleven subjects (55%) were male and the average age of all subjects was 24 ± 2 years (range, 20–27 years). Mean tHb was 12.7 ± 1.02 g/dL (mean ± SD). There were no adverse events because no subject experienced complications during the procedure or during the study follow-up period.
One hundred sixty-five tHb measurements were collected (average 9 per subject) with 335 paired SpHb measurements from 1 or both hands of each subject (index, middle, or ring finger). The average perfusion index level was 4.1% ± 2.0% and the range was 0.9% to 9.9%. Eight (2.4%) SpHb measurements had device-indicated low signal quality index and were not included in the analysis. The average decrease in tHb during the blood removal and hemodilution procedure was 2.4 ± 0.8 g/dL. Collected blood samples had tHb values that ranged from 7.5 to 13.8 g/dL, with 117 (71%) tHb measurements <12 g/dL and 40 (24%) <10 g/dL.
A scatterplot with linear regression statistics of tHb and SpHb measurements is shown in Figure 1. A Bland-Altman analysis is shown in Figure 2. The average difference (bias) between 335 paired measurements of SpHb and tHb was −0.15 g/dL, 1 SD of the difference (precision) was 0.92 g/dL, and the average root-mean-square difference was 0.94 g/dL. Table 1 shows the differences between SpHb and tHb as a function of tHb. There was no correlation between the difference between SpHb and tHb and the perfusion index (R2 = 0.02)
In this study, we compared simultaneous measurements of hemoglobin using noninvasive pulse CO-Oximetry and invasive laboratory CO-Oximetry (tHb) in normal subjects undergoing hemodilution. The average difference between 335 paired measurements of SpHb and tHb was −0.15 g/dL, 1 SD of the difference was 0.92 g/dL, and the average root-mean-square difference was 0.94 g/dL. The difference between SpHb and tHb was <2.0 g/dL for 97% of the measurements.
When evaluating any new method for measuring hemoglobin, the reference method contains inherent variability. Of all the methods available to measure tHb, CO-Oximeters are considered the most accurate and have the highest degree of interdevice reliability.3 In an intradevice comparison of 5 different CO-Oximeters, there was a range of 0.1 to 1.3 g/dL difference and an average standard deviation of 0.5 g/dL when measuring the same blood sample on 2 separate devices of the same model type.4 In an interdevice comparison of 31 different CO-Oximeters, there was a 0.4 to 0.9 g/dL range of difference depending on what range of tHb was being analyzed.5
The tHb measurement is performed >400 million times per year in the United States, either by itself or as part of a panel of other laboratory tests.1 This makes tHb measurement more frequent than any other laboratory variable, and indicates a large potential benefit if a validated noninvasive method were available.6 Some of the potential benefits of continuous noninvasive hemoglobin monitoring include hastening the detection of postoperative bleeding, preventing the overtransfusion of blood products during surgery, reducing phlebotomy-induced anemia in the intensive care unit, and increasing patient safety and comfort in all care areas where hemoglobin testing is done.
The limitations of this study are that the monitor was not evaluated in critically ill subjects, those with extremely low perfusion, or those with known peripheral vascular disease. No data were collected when tHb was <7 g/dL. This study in healthy subjects in a controlled procedure did not measure the potential clinical advantages of continuous monitoring on blood management and earlier detection of bleeding.
This study of the first commercialized noninvasive pulse CO-Oximetry found that its SpHb measurement is accurate to 1 g/dL (1 SD) compared with laboratory CO-Oximeter tHb measurement in subjects undergoing hemodilution. Future studies should examine SpHb accuracy in other populations and its impact on clinical decision making.
a Food and Drug Administration. 510(k) K080238 Masimo Radical 7/Rad 87/Rad 57t-SpHb. May 12, 2008. Available at: http://www.accessdata.fda.gov/cdrh_docs/pdf8/K080238.pdf. Accessed June 5, 2009.
1. DeFrances CJ, Lucas CA, Buie VC, Golosinskiy A. 2006 National Hospital Discharge Survey. Natl Health Stat Report 2008;(5):1–20
2. Performance characteristics. In: ABL800 Reference Manual. Copenhagen: Denmark Radiometer Medical ApS, 2004:5.28–5.30
3. Shapiro BA, Peruzzi WT, Templin R. Clinical Application of Blood Gases. 5th ed. Chicago: CV Mosby, 1994:334–6
4. Gehring H, Duembgen L, Peterlein M, Hagelberg S, Dibbelt L. Hemoximetry as the “gold standard”? Error assessment based on differences among identical blood gas analyzer devices of five manufacturers. Anesth Analg 2007;105:S24–30
5. RNA Medical. CVC 223 CO-Oximeter Calibration Verification Controls. Devens, MA: RNA Medical, Division of Bionostics
6. Rosen S. Hematology. In: Carlson B ed. The Worldwide Market for In Vitro Diagnostic Tests. 5th ed. New York: Kalorama Information, 2006:235–44