Kawagishi, Toshiya MD; Kanaya, Noriaki MD, PhD; Nakayama, Masayasu MD, PhD; Kurosawa, Saori MD; Namiki, Akiyoshi MD, PhD
Pulse oximetry is considered to be a standard of care in both the operating room and postanesthetic care unit, and it is widely used in all critical care settings (1–6). In some patients, such as patients with an arm injury, burn, or breast tumor, measurements of arterial saturation (SpO2) and arterial blood pressure (BP) are often made on the same arm. Important information may not be obtained if the pulse oximetry signal is lost during inflation of a cuff for BP measurement, particularly in patients with he-modynamic instability. Moreover, frequent false alarms cause by cuff-induced motion or hypoperfusion may confound a true hypoxic situation. A frequent incidence of false alarms in pulse oximetry may lead to complacency of care providers who have become desensitized to alarms. This may result in delays in response to significant events. False alarms may also prompt unnecessary medical interventions.
Newly developed pulse oximeters are designed to display accurate SpO2 during motion and hypoperfusion. Masimo SET oximetry showed excellent performance during motion in a number of clinical studies (6–9). However, there are only a few reports on the accuracy of newly developed pulse oximeters during hypoperfusion in a clinical setting (9). The purpose of the present study was to test the performance of different pulse oximeters during BP cuff-induced hypoperfusion. We compared the failure times of two motion-resistant pulse oximeters (Masimo SET and Nellcor N-395) with the failure time of a conventional device during BP cuff-induced hypoperfusion in volunteers.
The Institutional Ethics Committee of Sapporo Medical University approved this study, and all subjects gave written informed consent for participation in the study. Ten healthy male volunteers (25–40 yr old) who had not been taking any regular medication and were free of chronic pulmonary disease or other significant medical problems participated in this study.
To simulate a clinical situation, all measurements were performed in the same operating room using the same monitor at the same room temperature. Standard monitoring equipment (HP CMS M1166A; Hewlett Packard, Palo Alto, CA) was used. Lead II of an electrocardiogram (ECG) was continuously monitored, and heart rate (HR) was determined as an average of every 4 s from the ECG monitor. Volunteers were encouraged to rest comfortably in the supine position and were covered to simulate an operation. BP was measured oscillometrically (Hewlett Packard). A pulse oximeter sensor was attached to the index finger, and a BP cuff was attached to the ipsilateral arm of each subject. To prevent light artifact from influencing the saturation measurements, the volunteers’ arms were covered.
The following pulse oximeters were tested: Masimo-SET Radical (Masimo) (Masimo Corp, Irvine, CA), Nellcor N-395 (N-395) (Nellcor, Inc, Pleasanton, CA), Nellcor N-20PA (N-20PA), and Nellcor D-25 (D-25). Time to peak of cuff pressure (TP), time to loss of signal (TLS), time to recovery of signal (TR), and failure interval (FI) were measured to evaluate the failure time of each pulse oximeter. These time intervals were defined as follows:
* TP: time from pressing the Start button to begin cuff inflation until the peak pressure was reached.
* TLS: time from pressing the Start button to begin cuff inflation until the saturation value is no longer the same as the baseline or the pulse rate value did not agree with the pulse measurement by the ECG.
* FI: time interval that no SpO2 value was displayed.
* TR (TLS + FI): time from pressing the Start button to begin cuff inflation until the saturation value returned to baseline and the pulse measurement from the oximeter agreed with the measurement from the ECG.
The first BP measurement was excluded because often BP measurement devices inflate to a higher cuff pressure on the initial measurement and then inflate to a different cuff pressure on subsequent measurements based on the initial BP value. The beginning was marked when the button was pressed on the BP machine to initiate a BP measurement. Time interval measurements were performed with a stopwatch by an observer who was not one of the investigators. Before each measurement, subjects were told to lie quietly for at least 3 min to obtain a stable hemodynamic period. All pulse oximeters were tested on each volunteer. The order of pulse oximeter examination was decided with a random card in each subject.
All measurements were performed three times for each pulse oximeter and were averaged. Differences among the pulse oximeters were analyzed using analysis of variance. P < 0.05 was taken as significant. All data are expressed as means ± sd.
Ten healthy male volunteers (25–40 yr old) participated in this study. There were no significant differences among HR and BP measurements for any of the pulse oximeters studied (Table 1). There were no significant differences in TP intervals for any of the measurement periods. TLS for Masimo, N-395, N-20PA, and D-25 was 24.7 ± 4.0 s, 9.2 ± 1.9 s, 8.1 ± 1.5 s, and 9.7 ± 1.4 s, respectively. TLS was longest with the Masimo pulse oximeter as compared with other pulse oximeters (P < 0.05). TR for Masimo, N-395, N-20PA, and D-25 was 29.1 ± 3.7 s, 28.3 ± 7.3 s, 36.6 ± 4.2 s, and 35.7 ± 7.3 s, respectively. Masimo and N-395 showed significantly shorter TR than did the other two pulse oximeters (P < 0.05). The Masimo pulse oximeter failed to display a numerical SpO2 value (FI) for 4.4 ± 2.8 s, N-395 for 17.1 ± 6.9 s, N-20PA for 28.5 ± 5.0 s, and D-25 for 26.0 ± 6.8 s. The Masimo FI was significantly shorter than the FI measured using the N-395, N-20PA, or D-25 (P < 0.05).
The pulse oximeter is useful for detecting hypoxia in various situations. Although widely accepted as a standard of care and considered a great advance (1–6), there are several potential sources of error that need to be considered for the evaluation and appropriate use of pulse oximeters (10–12). Motion artifact, light artifact, and peripheral low perfusion are major causes of measurement error. During an operation, the BP cuff is often placed on the same extremity as the pulse oximeter and can cause false oximeter alarms when the cuff inflates. The frequency of BP cuff-induced false alarms depends on the BP measurement interval or cuff pressurization time. If false-positive alarms are frequent, the alarms may be disabled, potentially leading to a failure to detect true hypoxemia.
The Masimo SET oximeter showed the longest TLS and shortest TR, resulting in the shortest FI of all the oximeters studied. The TLS-interval measurement is worthy of closer inspection because it indicates that the Masimo oximeter was the only oximeter to maintain a signal well after the peak pressure in the cuff had been reached (TP). All oximeters include time averaging in their algorithms to minimize the influence of transient noise on the data. A longer averaging interval for the Masimo device could explain the longer TLS after peak pressure. This was not the case because the information available from the manufacturers indicated that the averaging intervals for the Masimo, N-395, N-20PA, and D-25 oximeters were similar at 8 s, automatic, 5–7 s, and 10 s, respectively. The longer TLS in the Masimo oximeter may be due to its signal processing system or algorithm. Although we did not measure flow directly, it is possible that the peak cuff pressure does not completely occlude the arterial flow and that an algorithm better adapted to measurements during low perfusion could continue to provide data after the peak of the cuff pressure.
The TR was similar between the Masimo and N-395, but the FIs were different because of the longer TLS of the Masimo device. Although we did not measure false alarm rates explicitly, a longer FI is likely to correlate with more false alarms. Shorter TR and FI intervals would therefore be an advantage during BP cuff-induced hypoperfusion.
The results of many studies on the effects of motion on the performance of a pulse oximeter may support our results. Barker and Shah (8) compared the levels of accuracy and data dropout rates of pulse oximeters (Nellcor N-200, N-3000, and Masimo SET) during standardized motion in healthy volunteers. They found that the error and dropout rate performance of Masimo was superior to that of the other two instruments, even during a period of rapid hypoxia caused by a decrease in inspired oxygen concentration. Similar results were obtained for 20 pulse oximeters, including 5 motion-resistant pulse oximeters (7).
There are several limitations to our study. The experimental method we used is different from true low perfusion conditions such as hypovolemic shock, hypothermia, or cardiopulmonary bypass because inflation of the BP cuff causes venous occlusion rather than low peripheral perfusion. Therefore, the results of this study do not apply directly to peripheral low perfusion in a clinical setting. However, tourniquets and BP cuffs to occlude blood flow have been widely used in laboratory studies of oximeter performance (13,14). Moreover, it has been reported that the Masimo pulse oximeter was useful in patients during mild hypothermic cardiopulmonary bypass (9). Masimo was able to display accurate SpO2 values more frequently and for a longer time than was a conventional pulse oximeter, even during nonpulsatile cardiopulmonary bypass (9). These findings support our results showing that Masimo could display SpO2 for a longer time than could the other pulse oximeters, resulting in the shortest FI during disruption of blood flow by inflation of the BP cuff.
The timing intervals used to evaluate performance of the oximeters had the potential to be influenced by variation in the BP and HR of the subjects and the response of the sphygmomanometer to these variations. The data indicate minimal variation in HR and BP for the pooled data from all subjects for each oximeter. Therefore, we believe our results demonstrate actual differences in pulse oximeter performances.
One might argue that there is some subjectivity in the measurement of the time intervals and that the observer might be influenced by expectations about the performance of a particular device because the observer was not blinded to the device being measured. We attempted to minimize this possibility by using a separate observer who was not an investigator and deciding the order of pulse oximeter examination with a random card in each subject. We performed this study to simulate clinical setting. Thus, we defined the end and return of a pulse oximeter signal using the loss and return of a stable saturation measurement. A stable saturation measurement was defined as a SpO2 value with a pulse signal corresponding to an ECG-measured pulse. Further study using a computer to record the data from the pulse oximeters and the sphygmomanometer might offer the ability to measure the time intervals more accurately and eliminate the subjective element of the study.
Although we recorded the return of the oximeter measurements, the accuracy of the oximeter data is not assured because SpO2 measurements were not confirmed with arterial blood co-oximetry. In a study using co-oximetry, Barker and Shah (8) demonstrated a bias and precision range of 0.8 ± 2.1 to 1.4 ± 3.5 with Masimo compared with a range of 4.5 ± 5.5 to 5.3 ± 5.3 with conventional pulse oximeter.
The longest FI measured in this study was approximately 30 seconds. Whereas an absent pulse oximeter signal for approximately 30 seconds may be within acceptable limits for most clinical situations, critically ill patients may have circulatory states that could exacerbate the differences we found in the present study. It is possible that our results in young, healthy volunteers under normoxic conditions significantly underestimate delays in the detection of rapidly developing hypoxemia by pulse oximetry in the clinical setting.
In conclusion, under conditions of BP-cuff–induced hypoperfusion, the Masimo maintained a longer measurement period after the peak of the cuff pressure (TLS) than any of the other oximeters studied. TR was similar for both the Masimo and the N-395 and significantly shorter than the N-20PA and D-25 oximeters. Although the data were obtained by studying only healthy volunteers, and critically ill patients may have other issues compromising oximeter performance, these observations suggest that data will be more available with fewer false-positive alarms when using the Masimo oximeter followed by the N-395.
1. Eichhorn JH. Prevention of intraoperative anesthesia accidents and related severe injury through safety monitoring. Anesthesiology 1989;70:572–7.
2. Aoyagi T, Miyasaka K. Pulse oximetry: its invention, contribution to medicine, and future tasks. Anesth Analg 2002;94:S1–3.
3. Wouters PF, Gehring H, Meyfroidt G, et al. Accuracy of pulse oximeters: the European multi-center trial. Anesth Analg 2002; 94:S13–6.
4. Tremper KK, Barker SJ. Pulse oximetry. Anesthesiology 1989; 70:98–108.
5. Chambrin MC, Ravaux P, Calvelo-Aros D, et al. Multicentric study of monitoring alarms in the adult intensive care unit (ICU): a descriptive analysis. Intensive Care Med 1999;25: 1360–6.
6. 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.
7. Barker SJ. “Motion-resistant” pulse oximetry: a comparison of new and old models. Anesth Analg 2002;95:967–72.
8. Barker SJ, Shah NK. The effects of motion on the performance of pulse oximeters in volunteers (revised publication). Anesthesiology 1997;86:101–8.
9. Irita K, Kai Y, Akiyoshi K, et al. Performance evaluation of a new pulse oximeter during mild hypothermic cardiopulmonary bypass. Anesth Analg 2003;96:11–4.
10. Ralston AC, Webb RK, Runciman WB. Potential errors in pulse oximetry. I. Pulse oximeter evaluation. Anaesthesia 1991;46: 202–6.
11. Webb RK, Ralston AC, Runciman WB. Potential errors in pulse oximetry. II. Effects of changes in saturation and signal quality. Anaesthesia 1991;46:207–12.
12. Jopling MW, Mannheimer PD, Bebout DE. Issues in the laboratory evaluation of pulse oximeter performance. Anesth Analg 2002;94:S62–8.
13. Wilkins CJ, Moores M, Hanning CD. Comparison of pulse oximeters: effects of vasoconstriction and venous engorgement. Br J Anaesth 1989;62:439–44.
14. Lawson D, Norley I, Korbon G, et al. Blood flow limits and pulse oximeter signal detection. Anesthesiology 1987;67:599–603.