Share this article on:

A Comparison of Lidocaine and Bupivacaine Digital Nerve Blocks on Noninvasive Continuous Hemoglobin Monitoring in a Randomized Trial in Volunteers

Miller, Ronald D. MD, MS*; Ward, Theresa A. BSN, RN*; McCulloch, Charles E. PhD; Cohen, Neal H. MD, MPH, MS*

doi: 10.1213/ANE.0000000000000144
Technology, Computing, and Simulation: Research Report

BACKGROUND: Blood hemoglobin can be monitored continuously and noninvasively with a noninvasive spectrophotometric sensor (Masimo SpHb). The perfusion index (PI) of the finger is directly related to the clinical accuracy of SpHb. We evaluated those variables that influence PI without the influences of surgery and anesthesia.

METHODS: Based on our past studies, 12 awake adult volunteers were studied. A SpHb sensor was attached to the same finger of each hand. The temperature of each finger was measured via a skin surface probe. A digital nerve block (DNB) was performed with 1% lidocaine on one finger and 0.25% bupivacaine on the other finger of the opposite hand. SpHb, PI, and finger temperature were monitored continuously 30 minutes before and 3 to 4 hours after placement of the DNB. A random effects spline regression was used to flexibly model the outcomes before and after the DNB and to compare the effects of lidocaine and bupivacaine.

RESULTS: The DNBs increased the PI for both lidocaine and bupivacaine (P < 0.0001) and finger temperature from both lidocaine (P < 0.0001) and bupivacaine (P = 0.02). The duration of action of bupivacaine was markedly longer than that of lidocaine (P < 0.0001). Between 45 and 75 minutes after insertion of the DNB, the PI with bupivacaine was substantially higher than that of lidocaine. The PI was directly related to changes in finger temperature and SpHb. During this time interval, 11 of the 12 volunteers receiving bupivacaine descriptively had increases in finger temperature ranging from no change to 6.1°C. In contrast, only 6 of the 12 lidocaine volunteers had increases in finger temperature ranging from no change to 4°C. Changes in PI were directly correlated with SpHb values (correlation coefficient = 0.7).

CONCLUSIONS: A DNB increases PI and finger temperature. These increases lasted 2 to 3 hours longer with bupivacaine than lidocaine. The increases in PI were associated with slightly higher SpHb values. We conclude that the DNB induces increases in PI and temperature of the finger. Because of the close relationship between finger temperature, PI, and SpHb, consistently increasing finger temperature and PI could increase the accuracy of SpHb.

From the Departments of *Anesthesia and Perioperative Care, and Epidemiology and Biostatistics, University of California, San Francisco, San Francisco, California.

Accepted for publication January 8, 2014.

Funding: This work was supported by the Department of Anesthesia and Perioperative Care at the University of California, San Francisco. The Masimo Corporation provided the Radical 7 Pulse Co-Oximeter with SpHb™, sensors and software for the study.

Conflict of Interest: See Disclosures at the end of the article.

Reprints will not be available from the authors.

Address correspondence to Ronald D. Miller, MD, MS, Department of Anesthesia and Perioperative Care, University of California, San Francisco, 533 Parnassus Ave., U450, San Francisco, CA 94143-0648. Address e-mail to

Blood hemoglobin (Hb) levels are critical to decision-making regarding blood transfusions and patient blood management strategies.1,2 With recent advances in monitoring, Hb can now be continuously and noninvasively measured by using spectrophotometric finger sensors (Masimo Corporation, Irvine, California, SpHb).3 While the technology reflects a significant advance in clinical management, debate remains over when and under what circumstances SpHb is sufficiently accurate to guide clinical transfusion decision-making.4–8 Despite advances in sensor technology, the accuracy of SpHb has not improved.4–8

One of the important aspects of measurement of SpHb is the perfusion of the finger as defined by the perfusion index (PI).4,8,9 PI is a relative assessment of the pulse strength at the monitoring site. Specifically, if the PI is good (a high reading), accuracy is usually better than when the PI is low. Accordingly, increasing PI by administration of a local anesthetic-induced digital nerve block (DNB) should increase the accuracy of SpHb.4,8,9 Yet a lidocaine DNB only modestly increased the PI and accuracy of SpHb.9 Perhaps the duration of the DNB was too short to optimize the PI during lengthy surgical procedures with anticipated large blood loss. If the PI could be more consistently increased, SpHb may be more accurate for a longer time. Yet many other variables, such as surgery, temperature, and anesthesia also probably influence the accuracy of SpHb.10,11

To better understand the variables influencing the PI, we conducted a study in awake volunteers in a room temperature-maintained environment. A major goal was to create conditions that would make SpHb more likely to be useful. In a double-blind manner, we compared DNBs by using 1% lidocaine in 1 finger and 0.25% bupivacaine in the same finger on the opposite hand. The use of the 2 different local anesthetics was chosen to determine whether a bupivacaine-induced DNB with a longer duration of action would improve the PI in a more consistent manner, both in duration and intensity than a DNB by using lidocaine. Clinical studies usually have many variables. If possible, isolating 1 important variable (e.g., PI) facilitates the ultimate understanding of several variables put together. Our study design in volunteers was used to eliminate some of the effects of surgery and anesthesia on PI, including, but not limited to, the direct effects of anesthetic drugs, room temperature, blood loss, and variable hemodynamics.

Back to Top | Article Outline


Based on our past studies,4,9 we designed this study to be performed in 12 volunteers (24 hands). We received approval from the University of California, San Francisco, Human Research Protection Program. Twelve healthy volunteers (6 women, 6 men) aged 18 to 55 years, nonsmokers, with no use of medications or supplements except birth control, and weighing between 52 and 75 kg were studied. Written informed consent was obtained before the study. All volunteers were asked to withhold caffeine for at least 12 hours.

Volunteers were in the supine position with the head slightly elevated on a stretcher for the duration of the study. SpHb was continuously monitored on either the third finger (3 volunteers) or fourth finger (9 volunteers) of each hand by using the Masimo Radical 7 Pulse Co-Oximeter with SpHb™ and Rainbow Adhesive Sensors, version RevF (both by Masimo Corporation). The sensors were covered with an optical shield to prevent optical interference. A noninvasive, surface temperature probe (Bluestar Enterprises, Inc, Omaha, Nebraska) was placed on the palmer surface of the finger distal to the SpHb sensor. The ambient room temperature was recorded hourly and remained between 69°F and 71°F (20.5°C–21.6°C) with the exception of 2 volunteer studies (i.e., 2 volunteers: 4 hands) during which the room temperature was 73° to 74° (22.7°C–23.3°C). SpHb, PI, and finger temperature data were recorded for 30 minutes before the DNB and every 2 minutes for the first hour and every 5 minutes after DNB for up to 4 hours for each finger. Data recorded during this time period served as baseline data points. The volunteers’ verbal responses regarding the time of onset and duration of anesthesia and finger temperature change were also recorded.

A randomization schedule was used to assign each volunteer to receive in the fingers attached to a sensor either 1% lidocaine in the left finger and 0.25% bupivacaine in the right finger (6 volunteers), or 1% lidocaine in the right finger and 0.25% bupivacaine in the left finger (6 volunteers). Two mL was injected into each finger, 1 mL each at the base of the medial and lateral side of each finger, respectively. The study coordinator (second author) responsible for collecting all study data was blinded to the randomization schedule or medication used for each DNB.

Back to Top | Article Outline

Statistical Analysis

The data were analyzed in the following manner: Initially, the consistency of the preblock PI data was analyzed by comparing the lowest and highest PI value before the DNB was administered. Then, the PI data from lidocaine and bupivacaine were compared and plotted against time before and after the DNB. A random effects spline regression was used to flexibly model the outcomes before and after the DNB and to compare the effects of lidocaine and bupivacaine. The model we fit used piecewise linear splines with a baseline period during which the PI was flat, then allowed for a sharp increase in PI initially between local anesthetic administration and 10 minutes, and then different piecewise linear trends between 10 to 60 minutes, 60 to 120, and 120 to 240 minutes. Data beyond 240 minutes were not included in the analysis. Each term in the model was allowed to be different by type of DNB. Because of considerable volunteer to volunteer variability in the trajectories in the PI, the model incorporated individual specific random effects for the baseline value, the rapid increase in PI, the trend from 10 to 60 minutes and from 60 to 120 minutes. An unstructured variance-covariance was used to allow flexibility in the modeling. The tests of the equality of the 2 models were then performed.12,13 Models were fit by using the mixed model routines (xtmixed) in Stata (StataCorp, College Station, TX, versions 12 and 13). Models were checked by qualitatively comparing a generalized estimating equations (GEE) fit with a working exchangeable correlation structure and robust standard errors. The use of robust standard errors gives correct inference even when the assumptions of normality or the assumption of an exchangeable correlation structure is incorrect.13,14

Because the PI values correlate with the accuracy of SpHb, we needed to understand what other variables might influence PI.4,8,9 Accordingly, we analyzed whether finger temperature could influence PI. We also analyzed whether changes in PI during the 30 minutes before the DNB (control) clinical period (i.e., no anesthesia or surgery) would influence SpHb values. For the analysis of changes in PI and SpHb resulting from the DNB, we compared the averages of individual volunteer’s data during the 45- to 75-minute time window after DNB. This time was selected because changes (if any) as a result of the DNB would have taken place but before the PI would be expected to return to the baseline pre-DNB level. We report correlation coefficients for these analyses for ease of interpretation, but we tested for the existence of a relationship by regressing PI on finger temperature or SpHb by using GEE methods to account for the clustering by person.

To compare the PI values between lidocaine and bupivacaine in various temperature ranges, we divided the finger temperature values into quartiles (Table 1). We then compared the PI values by using GEE methods to account for the repeated measures across time and fingers within a person (Table 1).

Back to Top | Article Outline


For most of the volunteers, the PI and finger temperature were observed to remain stable for the 30 minutes before insertion of the DNB. However, in 7 of the 24 hands (12 volunteers-2 hands each), the PI decreased >1.5% before insertion of the DNB. Only 1 hand had the finger temperature change >1.5°C; specifically, 1 finger had a 1.8° increase in temperature.

The DNB from both local anesthetics had overall effects on both PI and finger temperature and were statistically significantly different (P < 0.0001 for both PI and finger temperature). Fig. 1 demonstrates a plot of the model fit for PI. As the figure indicates and as we detail below, the initial increase from a lidocaine DNB was larger than bupivacaine but short in duration. The pharmacodynamic model fit and the data are illustrated in 1 volunteer (i.e., 2 hands) to facilitate understanding of our approach (Fig. 2). All the data were plotted for all volunteers with individual lines of identity calculated and displayed in Fig. 1. Lidocaine increased PI more than bupivacaine initially (P < 0.0001). However, the initial increase in PI after lidocaine DNB was short in duration, and by 100 minutes, the bupivacaine had higher PI (P < 0.0001). This difference between the 2 drugs persisted even out to 200 minutes (Fig. 3, P < 0.0001). The results by using the GEE model fits with robust standard errors were qualitatively similar to the random effect model fits (all P values above were <0.015, except the P value for the initial increase in PI, which was P = 0.08). The PI data during lidocaine versus the bupivacaine blocks were statistically significantly different (P < 0.001), by using both random effects regression models or GEE analyses.

The post-DNB PI values were related to finger temperature (Fig. 4) and SpHb (Fig. 5). Each data point in the figure is the average of all their measurements in the time window from 45 to 75 minutes. As a result, there are 24 values reflected in each dot (2 values per volunteer, i.e., 1 value for each hand). This time was selected because changes (if any) from the DNB would have occurred but before the data response would be returning to baseline. There was a strong direct correlation between changes in finger temperature and PI (correlation coefficient = 0.83; P < 0.0001) (Fig. 4). There was also a significant correlation (correlation coefficient = 0.7) between changes in SpHb and PI (P < 0.0001) (Fig. 5). While there was a tendency for some bupivacaine volunteers to have higher PIs at a given finger temperature, this relationship is not statistically significant.

There was a modest descriptive relationship between changes in PI and surface finger temperature. After the lidocaine DNB, most volunteers (5 of 6) had the PI decrease with a decrease in finger temperature (Fig. 4). Alternatively, with 1 exception, increases in finger temperature after lidocaine tended to be associated with increases in PI.

The change in SpHb values were directly related to changes in the PI (Fig. 5). Analyzing the variables in a 45- to 75-minute period after DNB resulted in a direct relationship between PI and SpHb. More specifically, there was a moderate positive relationship between the 2 variables. Bupivacaine tended to contribute to larger changes in both SpHb and PI, but for both local anesthetics, the positive association was consistent.

All the PI and finger temperature averages from the 25- to 45-minute window are illustrated in Fig. 6. For both lidocaine and bupivacaine, there is a direct relationship between finger temperature and PI. For lidocaine, the PI was consistently <4.0% when finger temperature was <30°C. In contrast, many of the PI values were higher than 4.0% after the bupivacaine DNB even when finger temperature was <30°C. Yet when the finger temperature was higher than 34°C, the PI values were quite variable after both lidocaine and bupivacaine DNB including those that were still <4.0%.

The difference between the lidocaine and bupivacaine DNB was inversely related to finger temperature (Table 1). Specifically, the estimated difference in PI between lidocaine and bupivacaine was relatively large in the range <32.8°C (PI of lidocaine about 1 less than bupivacaine). However, this did not reach statistical significance (0.2 < P< 0.1) (Table 1).

Back to Top | Article Outline


Clinical decisions regarding blood transfusions are partly based on Hb values as determined by routine laboratory Co-Oximetry (tHb). We previously found that noninvasive Hb monitoring (SpHb) is not as accurate as necessary for such decisions.4 The measure of accuracy is usually “SpHb-tHb.” During our initial study,4 we also noticed that overall variability and absolute differences in SpHb-tHb values decreased with increasing PI. The interpretation was that SpHb is likely to be more accurate with higher PI values. To test this interpretation, we then successfully increased the PI by administering a DNB with lidocaine.9 The mean PI increased by 0.55%, and accuracy improved. Still, the use of lidocaine produced an inconsistent increase in PI probably because the duration of the DNB was too short. This shortcoming is now illustrated in the current manuscript. Nevertheless, this was the first study that demonstrated a benefit to improving the physiology of the finger.9 Both of these studies4,9 indicate that higher PI values are frequently associated with greater accuracy. We then decided that understanding the physiology of the finger without the influence of anesthesia and surgery and evaluating bupivacaine would provide baseline guidelines for designing our future clinical studies. We found that bupivacaine consistently increased PI for at least 3 hours. Last, there was a direct relationship between changes in PI, SpHb, and finger temperature (Figs 4 and 5).

Several studies have suggested that the SpHb may not accurately reflect a laboratory-derived blood hemoglobin level (tHb) in a number of situations. Specifically, SpHb may not be sufficiently accurate to facilitate transfusion decisions.15–19 For example, Gayat et al.15 conclude that SpHb is “too unreliable to guide transfusion therapy.” Applegate et al.16 and Frasca et al.17 both state that with large blood loss, tHb, not SpHb should be used as a guide for transfusion decisions. Lamhaut et al18 found that 46% of the SpHb values had a SpHb-tHb difference >1.0 g/dL that is similar to the findings from our recently published study.4 Causey et al.5 suggest that further study is needed to determine the circumstances under which SpHb is an accurate measure of tHb. We certainly agree, and this is the basis of our current study. If the PI is sustained at an adequate level, acceptable accuracy with the SpHb is possible. While our study does not determine “accuracy” per se, we have evaluated variables that can influence PI.

Several reasons have been proposed to explain the lack of correlation between tHb and SpHb and the circumstances that impact the relationship between the noninvasive and invasive measurement methodologies.3 Rather than the PI, Berkow et al.6 used the “signal-quality indicator” as an indicator of whether the SpHb value is reliable. They concluded that the SpHb has “acceptable” clinical accuracy even though 56 of their 186 data pairs had unacceptably low signal-quality indicator values and were not included in their conclusions.

An alternative indicator of reliability is the PI used in the present study. SpHb is often inaccurate (SpHb-tHb >1.5–2.0 g/dL) when the PI is <2.0% (4, 8, 9). As a result, Masimo and several investigators have recommended the PI as a guide to ensuring accuracy of SpHb as a measure of tHb. Nguyen et al.8 and we9 found that SpHb is much more reliable when the PI is higher than 2.0% to 3.0%. Based on these findings, a PI of higher than 2.0% and preferably 3.0% is required to ensure reliability of the SpHb as a measure of tHb.8,9 Our study was designed to determine whether the PI could be improved by administration of a DNB. We assume that the accuracy of SpHb would be adequate if PI values were persistently above 2.0% to 3.0%.

Previously we found that the PI transiently increased and the accuracy of SpHb improved after a lidocaine DNB9 and, in this clinical model, a PI higher than 3.0%, is associated with a frequently accurate SpHb. However, Nguyen et al.8 concluded that SpHb was not accurate unless the PI was >2.0% at a minimum and preferably higher. Clearly, the specific PI at which the accuracy improves is in question. Yet Applegate et al.16 did not report the PI values but came to the conclusion that tHb should be used instead of SpHb when large blood losses occur. In 2012, additional studies in obstetrics20 and liver surgery21 have also questioned whether SpHb is sufficiently accurate for transfusion decisions. Even though discussed, PI data were not reported in the obstetrics study.20 In the liver surgery study,21 the PI was “slightly” related to the accuracy of SpHb although specific data were not provided.

As noted, our earlier studies of the relationship between SpHb and measures of perfusion concluded that: (1) the physiology of the finger is an important influence on the accuracy of the SpHb;9 and (2) the PI reflects the overall perfusion of the finger and is generally related to the accuracy of SpHb.9 We previously also noted that when the PI is >4.0%,4,9 the SpHb value is more accurate than it is when the PI is <4.0%, supporting the findings of Nguyen et al.8 This study was designed to better understand how or if the PI can be persistently increased above the threshold we previously reported to be associated with better SpHb accuracy. We decided to achieve this goal in normal volunteers not subjected to anesthesia or surgery with associated blood loss. We compared the response to a lidocaine DNB with a longer-acting bupivacaine DNB to determine if the longer duration of the bupivacaine DNB would optimize the PI and provide a longer period for obtaining reliable measures of SpHb. Based on this comparison, we found that lidocaine induced a small and transient increase in PI. These modest increases in PI probably account for the modest increase in accuracy with our clinical study of SpHb.9 In contrast, the longer-acting bupivacaine produced a larger and sustained increase in PI. After the study was finished, we noticed that a bupivacaine DNB appeared to sustain a higher PI than lidocaine especially with a cooler finger temperature (22.3 to 28.7°C (Table 1) (Fig. 6). However this trend was not statistically significant (Table 1).

The goal of this study was to improve the physiology of the finger for a sustained period of time. We arbitrarily placed emphasis on the PI as an indicator of the status of the physiology of the finger. Specifically, we tried to maintain a PI of >3.0%, a value known to be associated with more consistently accurate SpHb values. We demonstrated that improving perfusion to the finger by providing a DNB directly impacts the magnitude of the PI signal. It is tempting to recommend a bupivacaine DNB to clinicians who choose to use SpHb for transfusion decisions. However, there should be some caution with such a recommendation. Of prime importance is that our data were obtained in volunteers without the presence of the many variables associated with anesthesia and surgery. Yet we have demonstrated the relationship between the physiology of the finger and a sustained PI that is known to be associated with SpHb accuracy. Consequently, we expect accuracy of the SpHb to be better when the PI is >3%. A bupivacaine DNB increased the PI of about 2% over the preblock value. Further studies evaluating other methods for improving blood flow to the finger will be helpful in determining the optimal method to improve the PI and subsequent SpHb accuracy. For example, would local warming of the hand alone be effective in sustaining a higher PI? Would a similar approach have value in the anesthetized surgical patient or the patient receiving vasoactive drugs? Addressing the impact of these approaches to optimizing perfusion to the finger are critical to understanding the clinical situations in which the SpHb is a reliable guide to assess blood loss and transfusion decisions.

Back to Top | Article Outline


Name: Ronald D. Miller, MD, MS.

Contribution: This author helped design and conduct the study, analyze the data, and write the manuscript.

Attestation: Ronald D. Miller has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.

Conflicts of Interest: Ronald D. Miller received honoraria in the past from Masimo. Dr. Miller is a member of the Masimo Scientific Advisory Board, the manufacturer of the SpHb, and has received travel reimbursement in the past. The Masimo Corporation provided the Radical 7 Pulse Co-Oximeter with SpHb™, sensors and software for the study.

Name: Theresa A. Ward, BSN, RN.

Contribution: This author helped design and conduct the study, analyze the data, and write the manuscript.

Attestation: Theresa A. Ward has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Conflicts of Interest: The author has no conflicts of interest to declare.

Name: Charles E. McCulloch, PhD.

Contribution: This author helped analyze the data and write the manuscript.

Attestation: Charles E. McCulloch has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Conflicts of Interest: The author has no conflicts of interest to declare.

Name: Neal H. Cohen, MD, MPH, MS.

Contribution: This author helped analyze the data and write the manuscript.

Attestation: Neal H. Cohen has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.

Conflicts of Interest: The author has no conflicts of interest to declare.

This manuscript was handled by: Maxime Cannesson, MD, PhD.

Back to Top | Article Outline


The authors thank the Masimo Corporation and the faculty, residents, and staff of the University of California, San Francisco, Moffitt/Long Operating Rooms, who facilitated the conduct of our patient studies.

Back to Top | Article Outline


1. Spahn DR, Madjdpour C. Physiologic transfusion triggers: do we have to use (our) brain? Anesthesiology. 2006;104:905–6
2. Goodnough LT, Shander A. Patient blood management. Anesthesiology. 2012;116:1367–76
3. . Masimo Corporation. Masimo Rainbow SET® Radical 7 Pulse CO-Oximeter and Accessories: 510(k) Summary. Available at: Accessed June 15, 2010
4. Miller RD, Ward TA, Shiboski SC, Cohen NH. A comparison of three methods of hemoglobin monitoring in patients undergoing spine surgery. Anesth Analg. 2011;112:858–63
5. Causey MW, Miller S, Foster A, Beekley A, Zenger D, Martin M. Validation of noninvasive hemoglobin measurements using the Masimo Radical-7 SpHb Station. Am J Surg. 2011;201:592–8
6. Berkow L, Rotolo S, Mirski E. Continuous noninvasive hemoglobin monitoring during complex spine surgery. Anesth Analg. 2011;113:1396–402
7. Morey TE, Gravenstein N, Rice MJ. Assessing point-of-care hemoglobin measurement: be careful we don’t bias with bias. Anesth Analg. 2011;113:1289–91
8. Nguyen BV, Vincent JL, Nowak E, Coat M, Paleiron N, Gouny P, Ould-Ahmed M, Guillouet M, Arvieux CC, Gueret G. The accuracy of noninvasive hemoglobin measurement by multiwavelength pulse oximetry after cardiac surgery. Anesth Analg. 2011;113:1052–7
9. Miller RD, Ward TA, McCulloch CE, Cohen NH. Does a digital regional nerve block improve the accuracy of noninvasive hemoglobin monitoring? J Anesth. 2012;26:845–50
10. Bornmyr S, Svensson H, Söderström T, Sundkvist G, Wollmer P. Finger skin blood flow in response to indirect cooling in normal subjects and in patients before and after sympathectomy. Clin Physiol. 1998;18:103–7
11. Lange KH, Jansen T, Asghar S, Kristensen PL, Skjønnemand M, Nørgaard P. Skin temperature measured by infrared thermography after specific ultrasound-guided blocking of the musculocutaneous, radial, ulnar, and median nerves in the upper extremity. Br J Anaesth. 2011;106:887–95
12. Altman DG, Bland JM. Measurement in medicine: the analysis of method comparison studies. Statistician. 1983;32:307–17
13. Liang KY, Zeger SL. Longitudinal data analysis using generalized linear models. Biometrika. 1986;73:13–22
14. Long JS, Ervin LH. Using Heteroscedasticity Consistent Standard Errors in the Linear Regression Model. Amer Statistician. 2000;54:217–24
15. Gayat E, Bodin A, Sportiello C, Boisson M, Dreyfus JF, Mathieu E, Fischler M. Performance evaluation of a noninvasive hemoglobin monitoring device. Ann Emerg Med. 2011;57:330–3
16. Applegate RL 2nd, Barr SJ, Collier CE, Rook JL, Mangus DB, Allard MW. Evaluation of pulse cooximetry in patients undergoing abdominal or pelvic surgery. Anesthesiology. 2012;116:65–72
17. Frasca D, Dahyot-Fizelier C, Catherine K, Levrat Q, Debaene B, Mimoz O. Accuracy of a continuous noninvasive hemoglobin monitor in intensive care unit patients. Crit Care Med. 2011;39:2277–82
18. Lamhaut L, Apriotesei R, Lejay M, Viven B, Carli P. Comparison between a new noninvasive continuous technology of spectrophotometry-based and RBC count for haemoglobin monitoring. Eur J Anaesthesiol. 2010;27:3AP7–1
19. Ehrenfeld JM, Henneman JP, Sandberg WS. Impact of continuous and noninvasive hemoglobin monitoring on intraoperative blood transfusions. Presented at the American Society of Anesthesiologist Annual Meeting 2010;LB05. 2010 Boston, MA Massachusetts General Hospital
20. Butwick A, Hilton G, Carvalho B. Non-invasive haemoglobin measurement in patients undergoing elective Caesarean section. Br J Anaesth. 2012;108:271–7
21. Vos JJ, Kalmar AF, Struys MM, Porte RJ, Wietasch JK, Scheeren TW, Hendriks HG. Accuracy of non-invasive measurement of haemoglobin concentration by pulse co-oximetry during steady-state and dynamic conditions in liver surgery. Br J Anaesth. 2012;109:522–8
© 2014 International Anesthesia Research Society