Secondary Logo

Journal Logo


Performance Evaluation of a New Pulse Oximeter During Mild Hypothermic Cardiopulmonary Bypass

Irita, Kazuo MD, PhD; Kai, Yoichiro MD, PhD; Akiyoshi, Kozaburo MD; Tanaka, Yukako MD; Takahashi, Shosuke MD, PhD

Author Information
doi: 10.1213/00000539-200301000-00004
  • Free


The use of pulse oximeters (POs) and capnometers as routine monitors during anesthesia is believed to contribute to anesthesia safety (1,2). The usefulness of POs for early detection of hypoxemia has been shown in a clinical setting (3,4) and in a study using an anesthesia simulator (5). PO saturation (Spo2) has been proposed as a fifth vital sign (6). A large scale prospective study has shown that monitoring with POs decreased the rate of perioperative myocardial ischemia, although overall morbidity was not affected (4). POs have been introduced in the postanesthesia care unit (3,4,7–9), in the intensive care unit (10), in the emergency room (6), in the general ward (3,11), and even during out-of-hospital transportation (12,13). The widespread use of POs, however, enhances the number of false alarms due to motion (2, 7–11, 14) and hypoperfusion (10,12,14–18).

Newly developed POs are designed to display accurate Spo2 during motion and hypoperfusion. There are many clinical reports indicating that these new POs display accurate Spo2 during motion (7–9,11,14). However, there are only a few reports concerning the accuracy of newly developed POs during hypoperfusion in a clinical setting (18). We compared the performance of such a PO (7,9,14) with that of a conventional device during cardiopulmonary bypass (CPB) under moderate hypothermia.


The Kyushu University IRB approved the study. Eighteen adult patients (male/female, 10:18) undergoing cardiac surgery with mild hypothermic CPB were enrolled, and informed consent was obtained. Their mean age was 59 ± 18 yr, and their New York Heart Association (NYHA) classification was 2.2 ± 0.8. Performed operations were as follows: valve replacement, n = 13; valve replacement plus coronary artery bypass grafting, n = 2; coronary artery bypass grafting, n = 2; and ventricular septal defect closure, n = 1. Anesthesia was maintained with fentanyl (40–60 μg/kg) and diazepam (0.3–0.4 mg/kg).

CPB was maintained with a roller pump and nonpulsatile flow, which produced a pulse pressure of approximately 12 mm Hg. Pump flow rate was maintained at approximately 2.1–2.4 L · min−1 · m−2. Pao2 during CPB was maintained at 250–300 mm Hg. Acid-base status was maintained according to an alpha-stat procedure. The minimum bladder temperature during CPB was 31.1°C ± 1.0°C. The duration of CPB was 203 ± 78 min, and the duration of aorta cross-clamping (AoX) was 135 ± 66 min. During the study period, vasoactive drugs were used as follows: dopamine (3–10 μg · kg−1 · min−1) in 16 patients, dobutamine (3–8 μg · kg−1 · min−1) in 17 patients, nitrate (0.3–2.0 μg · kg−1 · min−1) in 16 patients, diltiazem (0.3–1.0 μg · kg−1 · min−1) in 11 patients, prostaglandin E1 (5–20 ng · kg−1 · min−1) in 10 patients, and olprinone (0.05–0.4 μg · kg−1 · min−1) in 2 patients. None of the patients developed major postoperative complications.

We routinely used the Nihon (N) Kohden AY-900P (Nihon Kohden, Tokyo, Japan), a conventional PO. We prospectively compared a newly developed PO, the Masimo (M) SET Radical (Masimo Corp., Irvine, CA), with N. PO sensors were clip-on and reusable in N and were tape-on and disposable in M. PO sensors and a skin temperature probe (YS1409J; Nihon Kohden) were applied on randomly chosen fingers on the ipsilateral hand. PO data were collected in real-time with a personal computer data acquisition system and handwritten notes. The sensitivity of M to detect pulse waves was set at the highest level. In this setting, M might display Spo2, even when its sensor slipped off the patients. However, the pulse rate recorded in the computer corresponded to the rotation speed of the roller pump when M displayed Spo2, suggesting that artifacts due to the detachment of PO sensors did not affect our results. PO failure was defined as failure to show no Spo2 data (no Spo2) and/or to display incorrect Spo2 for longer than 3 min continuously (incorrect Spo2). Incorrect Spo2 was defined as <97% during CPB, because arterial saturation obtained with cooximetry was >99% during CPB. The duration of PO failure was calculated as the duration of no Spo2 plus the duration of incorrect Spo2. The dropout rate, proposed by Barker and Shah (14), corresponds to the percentage of time when the PO shows no Spo2 data. The PO signal strength, monitored by M, was calculated as modulation of the infrared photoplethysmogram [(max − min)/mean] and was expressed as a percentage. We also examined PO performance from the standpoint of preoperative diuretic therapy and intraoperative hyperlactatemia. Of 18 patients, 6 had preoperative diuretic therapy. We divided the patients into two groups according to their maximum intraoperative arterial lactate concentration: nine patients whose lactate was less than 80 mg/dL (Group L) and nine patients whose lactate was more than 80 mg/dL (Group H). Blood lactate level was examined with the Radiometer 725 analyzer (Radiometer, Copenhagen, Denmark).

Data are shown as a mean ± sd. Statistical analysis was performed with Student’s unpaired t-test and the χ2 test. A P value <0.05 was considered significant.


Typical baseline signal strength was 1% immediately prebypass. CPB produced measurable pulsations of 0.1% at approximately 1.6 Hz. During CPB, PO failure developed in 14 patients with N (78%), whereas it developed in 4 patients with M (22%;P = 0.0022;Table 1). All four patients with PO failure in M developed PO failure in N. PO failure occurred immediately after the initiation of CPB in four patients and just after AoX in nine patients.

Table 1
Table 1:
Number of Patients Who Developed Pulse Oximeter (PO) Failure Lasting Longer Than Three Minutes During Hypothermic Cardiopulmonary Bypass

The duration of PO failure was 36% ± 31% of the duration of CPB with N and was 6% ± 15% with M (P = 0.0006;Table 2). The duration of PO failure was 46% ± 43% of the duration of AoX with N and was 5% ± 15% with M (P = 0.0005). No Spo2 was provided for 36% ± 39% of the duration of AoX with N and for 4% ± 12% with M (P = 0.002). The observed PO failure was mainly due to no Spo2 rather than to incorrect Spo2. In 4 patients, in whom PO failure developed both with M and N, the duration of no Spo2 during AoX was 18.2% ± 22.5% with M, compared to 66.9% ± 38.3% in N (P = 0.0584). In 14 patients, in whom PO failure developed with N, the duration of no Spo2 during AoX was 54.4% ± 34.5%, which was not statistically different from 18.2% ± 22.5% with M.

Table 2
Table 2:
Duration of Pulse Oximeter (PO) Failure and Dropout Rate (DR) During Cardiopulmonary Bypass (CPB) and Aorta Cross-Clamping (AoX)

Skin temperature, bladder temperature, and mean arterial blood pressure (MABP) when PO failure started to occur and ended were similar between N and M (Table 3). The minimum MABP during which PO displayed an Spo2 of 99%–100% was 42 ± 12 mm Hg with N, whereas it was 36 ± 10 with M (P = 0.0034).

Table 3
Table 3:
Temperatures and Mean Arterial Blood Pressure (MABP) When Pulse Oximeter (PO) Failure Started to Occur and Ended

Among six patients with preoperative diuretic therapy, five developed PO failure in N, but only one in M (P = 0.0008). The duration of no Spo2 during AoX was 6.5% with M in one patient in whom diuretics were given before surgery and in whom PO failure developed during CPB. However, it was 35.6% ± 35.6% with N in 5 patients in whom diuretics were given before surgery and in whom PO failure developed during CPB. The incidence of PO failure was similar between N and M among 12 patients without preoperative diuretic therapy: 9 developed PO failure in N and 3 in M (P = 0.2258).

In Group L, PO failure developed in three patients with N and in two with M (P = 0.6471). However, in Group H, PO failure developed in eight patients with N and two with M (P = 0.0078). The duration of no Spo2 during AoX was 5.3% ± 4.0% with M in two patients who were classified as Group H and developed PO failure, compared to 35.4% ± 33.7% with N in eight patients who were classified as Group H and developed PO failure. There was no statistically significant difference.


This study shows that the incidence of PO failure during mild hypothermic CPB was significantly less in M than in N. Furthermore, the mean duration of PO failure was significantly shorter in M than in N, both during CPB and AoX. Although we used nonpulsatile flow during CPB, the presence of low-amplitude arterial pulsations (0.1%) permitted M to display Spo2 even in patients with mild hypothermic CPB. Among patients in whom PO failure developed, the duration of no Spo2 was shorter with M than with N both during CPB and AoX. However, there was no statistically significant difference, probably because of a great deviation of data and a small sample size. The duration of incorrect Spo2 tended to be shorter in M than in N, but this difference was not significant. However, the mean duration of no Spo2 was longer in N than in M. Therefore, the longer duration of PO failure in N was due to a longer duration of no Spo2, indicating that the ability to detect a pulse wave was superior in M compared with N. It has been shown that M and another type of new PO reduced the incidence of false alarms in a general surgical ward (11) and in a postanesthesia care unit (8).

Skin temperature less than 28°C–30°C and 26.5°C has been identified as a predictor of PO failure with conventional POs in children (17,18) and in adult (15) patients, respectively. However, we could not find any difference of bladder temperature and MABP between the two POs when PO failure started and ended. Villanueva et al. (18) have also shown that pulse pressure and core temperature, as well as percentage flow by laser Doppler, hemoglobin concentration, and age, were not predictors of PO failure. The only significant difference concerning MABP was that the minimum MABP when PO displayed Spo2 of 99%–100% was less with M than with N. We could not find any relationship between PO failure and IV administration of vasoactive drugs.

We observed that preoperative diuretic therapy and intraoperative hyperlactatemia influenced PO performance. In patients with preoperative diuretic therapy, PO failure occurred more frequently in N than in M. Because preoperative diuretic therapy is believed to decrease intravascular volume, and because intravascular volume status may easily affect peripheral perfusion, this observation supported the hypothesis that PO performance with M was superior to that with N during intravascular volume deficiency. Hyperlactatemia associated with hypothermic CPB is indicative of systemic hypoperfusion due to maldistribution of blood flow caused by both artificial circulation and hypothermia (19). Our observation that in Group H, PO failure developed more easily with N than with M also supported the finding that M could display accurate Spo2, even in patients with possible systemic hypoperfusion, more than N.

The effects of hypoperfusion on PO performance have been studied by partial occlusion of the arterial supply of the forearm with manual inflation of the blood pressure cuff (14,16,18). As mentioned previously (18), this method does not mimic physiological vasoconstriction, and an increase in venous pressure might affect PO performance. Furthermore, some of these studies have been performed in healthy volunteers (14,16). In contrast, Goldstein et al. (12) reported that PO failure was less frequent in M than in a conventional PO during interhospital transport by helicopter of five poorly perfused infants with persistent pulmonary newborn hypertension. Two reports describe a performance evaluation of conventional POs after hypothermic cardiac surgery (15,17). This is the first report that describes performance of POs, including a newly developed device, during hypothermic CPB.

A newly developed PO displayed accurate Spo2 significantly more frequently and longer than a conventional device during mild hypothermic CPB. Although we used nonpulsatile flow during CPB, a new-technology PO was able to display Spo2 in patients with possible hypovolemia, hyperlactatemia, or both. New-technology POs seem to be more useful for monitoring Spo2 during hypoperfusion. However, continuous display of accurate Spo2 with the new POs does not always guarantee adequate perfusion. It should be noted that the plethysmogram displayed on the new PO device is inappropriate for confirming the adequacy of circulatory function.


1. Eichhorn JH. Prevention of intraoperative anesthesia accidents and related severe injury through safety monitoring. Anesthesiology 1989; 70: 572–7.
2. Webb RK, van der Walt JH, Runciman WB, et al. Which monitor? An analysis of 2000 incident reports. Anaesth Intensive Care 1993; 21: 529–42.
3. Moller JT, Jensen PF, Johannessen NW, Espersen K. Hypoxaemia is reduced by pulse oximetry monitoring in the operating theatre and in the recovery room. Br J Anaesth 1992; 68: 146–50.
4. Moller JT, Johannessen NW, Espersen K, et al. Randomized evaluation of pulse oximetry in 20,802 patients. II. Perioperative events and postoperative complications. Anesthesiology 1993; 78: 445–53.
5. Lampotang S, Gravenstein JS, Euliano TY, et al. Influence of pulse oximetry and capnography on time to diagnosis of critical incidents in anesthesia: a pilot study using a full-scale patient simulator. J Clin Monit Comput 1998; 14: 313–21.
6. Mower WR, Myers G, Nicklin EL, et al. Pulse oximetry as a fifth vital sign in emergency geriatric assessment. Acad Emerg Med 1998; 5: 858–65.
7. Dumas C, Wahr JA, Tremper KK. Clinical evaluation of a prototype motion artifact resistant pulse oximeter in the recovery room. Anesth Analg 1996; 83: 269–72.
8. Rheineck-Leyssius AT, Kalkman CJ. Advanced pulse oximeter signal processing technology compared to simple averaging. II. Effect on frequency of alarms in the postanesthesia care unit. J Clin Anesth 1999; 11: 196–200.
9. Malviya S, Reynolds PI, Voepel-Lewis T, et al. False alarms and sensitivity of conventional pulse oximetry versus the Masimo SET™ technology in the pediatric postanesthesia care unit. Anesth Analg 2000; 90: 1336–40.
10. Jensen LA, Onyskiw JE, Prasad NGN. Meta-analysis of arterial oxygen saturation monitoring by pulse oximetry in adults. Heart Lung 1998; 27: 387–408.
11. Christensen M, Lie C, Rosenberg J. Continuous pulse oximetry in the general surgical ward: Nellcor N-200 versus Nellcor N-3000. Anaesthesia 1999; 54: 253–7.
12. Goldstein MR, Liberman RL, Taschuk RD, et al. Pulse oximetry in transport of poorly perfused babies. Pediatrics 1998; 102: 818.
13. Gilmore B, Hardwick W, Noland J, Patton D. Determination of systolic blood pressure via pulse oximeter in transported pediatric patients. Pediatr Emerg Care 1999; 15: 183–6.
14. Barker SJ, Shah NK. The effects of motion on the performance of pulse oximeters in volunteers (revised publication). Anesthesiology 1997; 86: 101–8.
15. Pälve H, Vuori A. Pulse oximetry during low cardiac output and hypothermia states immediately after open heart surgery. Crit Care Med 1989; 17: 66–9.
16. Weber W, Elfadel IM, Barker SJ. Low perfusion-resistant pulse oximetry. J Clin Monit 1995; 11: 284.
17. Picardo S, Di Chiara L, Averardi M, et al. Unreliability of pulse oximetry in hypothermic children after cardiovascular surgery with deep hypothermic circulation. Minerva Anestesiol 1998; 64: 427–30.
18. Villanueva R, Bell C, Kain ZN, Colingo KA. Effects of peripheral perfusion on accuracy of pulse oximetry in children. J Clin Anesth 1999; 11: 317–22.
19. Shangraw RE. Metabolic and splanchnic visceral effects of cardiopulmonary bypass. In: Gravlee GP, Davis RF, Utley JR, eds. Cardiopulmonary bypass: principles and practice. Baltimore: Williams & Wilkins, 1993: 509–41.
© 2003 International Anesthesia Research Society