Standard monitoring technologies in nonintubated patients are often inadequate to identify early signs of respiratory compromise.1 Current protocols rely primarily on clinical assessment, pulse oximetry, and, occasionally, capnography. Pulse oximetry, while widely used, is an indirect surrogate for ventilation because it reflects an already deteriorated state of respiratory compromise, rather than providing a real-time quantitative assessment of ventilation. Capnography has never gained widespread acceptance for use in nonintubated patients in hospital settings such as the general hospital floor and the postanesthesia care unit.2,3 Furthermore, even when capnography equipment is tolerated by the patient, it provides a lagging indicator of respiratory performance, rather than a direct measure of changes in respiratory volumes. If respiratory changes can be detected earlier, then interventions could be initiated earlier and the chance of serious complications could be significantly reduced.
Postoperative respiratory compromise generally begins with a decrease in ventilation, not oxygenation, followed by hypercarbia, and eventually hypoxemia.4 Pulse oximeters measure oxygenation, not ventilation, a fact that most clinicians do not appreciate, as indicated by a 2002 study which found that only 35% of nurses and 39% of physicians knew that pulse oximetry does not directly reflect changes in ventilation.5 Pulse oximetry is a valuable technique for assessing oxygenation, but is unfortunately incapable of detecting early changes in ventilation. If a patient is also receiving supplemental oxygen, hypoxemia may be further delayed in the spiral of respiratory depression.6 More recently, capnography was introduced and proposed as a better technique for the detection of early onset respiratory depression, aimed at increasing patient safety.7–10 Unfortunately, monitoring end-tidal CO2 (EtCO2) in nonintubated patients has proven to be less reliable than anticipated and often poorly tolerated by patients, rendering this technology impractical on the general care floor.2,11
Recently, a Food and Drug Administration–cleared impedance-based noninvasive respiratory volume monitor (RVM) became available to continuously measure minute ventilation (MV), tidal volume (TV), and respiratory rate (RR). Studies have shown it to be an effective, reliable way of providing accurate real-time continuous respiratory volume metrics, trends, and waveforms when tested against “gold standard” spirometry and ventilator measurements, with approximately 90% accuracy for MV and TV, and 98% accuracy for RR.12,13 The RVM uses a small amount of electrical current across several vectors through the lung and exploits the conductivity difference between air and tissue to calculate the amount of air moving in and out of the lung in real time. The RVM displays an easily interpretable breath-by-breath waveform as well as 30-second averages for MV, TV, and RR. In this study, we compared capnography measurements of EtCO2 and RR to noninvasive RVM measurements of MV and RR in a controlled environment in cooperative, nonintubated subjects without supplemental oxygen. We hypothesized that RVM measurements of MV will change more rapidly and by a larger degree than changes in EtCO2 in response to changes in ventilation.
IRB and Consent
This study was approved by the Schulman Associates IRB, Cincinnati, OH (SAIRB-11-0020). All subjects responded to an IRB-approved advertisement and gave informed written consent. Subjects were recruited between March 6, 2014, and May 29, 2014.
Continuous respiratory data were recorded from an impedance-based RVM (ExSpiron, Respiratory Motion, Inc, Waltham, MA) simultaneously with capnography data (Capnostream 20, Covidien, Mansfield, MA). Inclusion criteria were men and women aged 18 to 99 years. Exclusion criteria were hospitalization within 30 days before the study and pregnant females.
The RVM collected bioimpedance traces via an electrode padset placed in the recommended positions: sternal notch, xiphoid, and right midaxillary line at the level of the xiphoid. The skin was prepped and the padset applied in a fashion similar to that used in standard electrocardiogram electrode placement. No subject reported discomfort or problems resulting from the RVM electrode padset. At the beginning of the study, each subject breathed normally for 2.5 minutes through a heated pneumotachometer (Heated FVL, Morgan Scientific, Haverhill, MA) with disposable mouthpiece (QRS Diagnostic, Maple Grove, MN) while simultaneously collecting RVM data. The MV measured by the pneumotachometer during this period was then entered into the RVM to calibrate the MV estimated by the RVM during this same period. The RVM then calculated real-time measurements of MV, TV, and RR from the bioimpedance trace for the duration of the study.
After calibration, each subject performed 6 breathing trials at 3 different prescribed RRs as depicted in Figure 1. In trials 1 and 6, subjects were instructed to breathe normally (12.6 ± 0.6 min−1). In the middle 4 trials, subjects alternated between fast (25 min−1) and slow (5 min−1) breathing as set by a metronome. Trials 1, 2, 3, and 6 were 2.5 minutes long, while trials 4 and 5 were 1.5 minutes. This set of 6 breathing trials was repeated with 2 capnography sampling methods: nasal cannula and in-line sensor. In the nasal cannula set of trials, capnography data were collected using a nasal cannula with oral scoop sampling port (Covidien Smart CapnoLine Plus Oral/Nasal, Boulder, CO). In in-line sensor trials, capnography data were collected using a sensor (Covidien Filterline Set, Boulder, CO) connected in-line with the patient while the subjects breathed through a single-use snorkel mouthpiece and filter and wore a disposable noseclip (A-M Systems, Sequim, WA) to capture all inhaled and exhaled air. The order of the trials was balanced such that half the subjects performed the breathing trials with the nasal cannula first and half performed the breathing trials with in-line sensor first.
Data and Statistical Analysis
For each transition in prescribed RR (eg, when transitioning from fast to slow breathing), we measured the time to reach a new steady state value of MV, EtCO2, and RR for both the RVM and capnograph. Steady state was defined when the metric reached within the resolution of each metric (ie, 0.1 L/min for MV, 0.1 mm Hg for EtCO2, and 1 min−1 for RR). We calculated the ability of the capnography monitor to reflect changes in MV during these transitions in prescribed RRs. Specifically, we defined instrument sensitivity as the ratio between the observed changes in EtCO2 and the corresponding changes in MV measurements from the steady state values before a transition in prescribed RR to the new steady state value after the transition. This instrument sensitivity was calculated separately for each of the 2 EtCO2 sampling methods (ie, nasal cannula and in-line sensor) and averaged over all transitions in prescribed RR. In each patient, we also calculated the average EtCO2 and the average MV values for each of the 3 different breathing conditions (ie, normal, fast, and slow) and 2 different EtCO2 sampling methods (ie, nasal cannula and in-line sensor). In a spontaneously breathing patient at rest (ie, MV ≤ 10 L/min), a clinically relevant change in MV is considered to be at least 1 L/min. Based on the stated EtCO2 accuracy of the capnograph used in this study of ±2 mm Hg,14 the smallest change that can be measured with significance is 4 mm Hg, yielding a clinically useful instrument sensitivity of at least 4 mm Hg/(L/min).
During periods of steady breathing, Pearson correlations and bias between the RVM and capnography measurements of RR were calculated. Paired t tests were used to compare the time to reach a new steady state of MV and EtCO2 during transitions in prescribed RR from slow to fast and fast to slow. Paired t tested were also used to compare the time to reach a new steady state of RR measured with the RVM and capnography following transitions in prescribed RR. Paired t tests were used to compare measurements of MV and EtCO2 at different prescribed RRs and to compare measurements of MV, EtCO2, and instrument sensitivity between nasal cannula and in-line sensor trials. Nasal cannula and in-line sensor EtCO2 measurements were compared using a Bland-Altman analysis.15 All data are presented as mean ± SEM unless otherwise indicated. All analyses were performed in Matlab 2014b (MathWorks, Natick, MA). Results were considered statistically significant at P < 0.01.
Forty-eight subjects (15 females/33 males; age: 46.1 ± 14.3 years; body mass index: 27.6 ± 6.2 kg/m2, mean ± SD) completed the study. Figure 2 depicts a continuous representative RVM trace (A), MV and EtCO2 (B), and RR as measured by RVM and capnography (C) while a subject transitions from a fast (ie, hyperventilating) to slow (ie, hypoventilating) prescribed RR. Averaging over all trials and both sampling methods, the RVM-based RR reached a new steady state following a change in prescribed RR 27.5 ± 3.7 seconds faster (P < 0.0001) than capnography-based RR (37.6 ± 1.5 vs 64.2 ± 3.9 seconds). This delay in RR reported by the capnograph was likely due to differences between the algorithm used by the RVM and by the specific capnograph used in this study. More importantly, the differences in the time to reach a new steady value of MV and EtCO2 during transitions in breathing rate were more pronounced. Specifically, the RVM MV reflected the change in ventilation in 37.7 ± 1.4 seconds on average while EtCO2 changes were notably slower, often failing to reach a new asymptote before a 2.5-minute threshold.
The Table compares RR, MV, and EtCO2 measurements during normal, slow, and fast breathing trials averaged across all patients and both capnography sampling methods. During steady breathing, RR measured by the RVM and capnograph were strongly correlated (R = 0.98 ± 0.02) and consistent for all breathing rates (12.5 vs 12.9 min−1, 6.9 vs 6.7 min−1, and 24.6 vs 25.2 min−1 for normal, slow, and fast breathing trials, respectively). The capnograph-based RR was 0.21 ± 1.24 (SD) min−1 higher than the RVM-based RR during steady state breathing for all breathing trials. During the study, subjects modulated their MV from 23.6 ± 1.4 L/min while hyperventilating to 2.1 ± 0.2 L/min during hypoventilation. The corresponding EtCO2 measurements ranged from 23.3 ± 1.0 mm Hg (hyperventilation) and 38.3 ± 0.7 mm Hg (hypoventilation) (Figure 3C). No EtCO2 values above 44 mm Hg were recorded throughout all trials.
Figure 3 summarizes the instrument sensitivity (A), MV (B), and EtCO2 (C) for normal, slow, and fast breathing trials. As expected, a negative relationship was found between changes in MV and EtCO2 for both the nasal cannula and in-line sampling (Figure 3A). However, large changes in MV resulted in relatively small changes in EtCO2 (instrument sensitivity = −0.71 ± 0.11 and −0.55 ± 0.11 mm Hg per 1 L/min for nasal and in-line sampling, respectively, P = 0.13).
During normal breathing, the average EtCO2 measurement from the in-line sensor was higher than the average EtCO2 measurement from the nasal cannula (36.3 ± 0.6 vs 33.3 ± 0.6 mm Hg, P = 0.0005). Despite having higher EtCO2, MV measurements during in-line sampling normal breathing trials were also higher than nasal sampling trials (8.9 ± 0.5 vs 7.3 ± 0.4 L/min, P = 0.009) during normal breathing. Similar disparities between in-line and nasal cannula measurements were observed during hypoventilation (MV: 2.4 ± 0.2 vs 1.8 ± 0.2 L/min, P = 0.006; EtCO2: 40.0 ± 0.6 vs 36.5 ± 0.6 mm Hg, P = 0.0005) and hyperventilation (MV: 25.0 ± 1.4 vs 22.2 ± 1.3 L/min, P = 0.0002; EtCO2: 25.8 ± 0.9 vs 20.9 ± 1.0 mm Hg, P < 0.0001) (Figure 3, B and C). Importantly, regardless of sampling methodology, EtCO2 displayed low instrument sensitivity in response to large swings in MV, especially during hypoventilation. Specifically, MV decreased from 7.0 during normal breathing to 2.0 L/min during slow breathing (P < 0.0001) and EtCO2 increased from 33.7 to 36.8 mm Hg (P < 0.0001) during the nasal cannula sampling trials. Similarly, MV decreased from 8.7 down to 2.9 L/min (P < 0.0001) and EtCO2 increased from 36.7 to 40.3 mm Hg (P < 0.0001) from normal to slow breathing during the in-line sensor trials.
A strong correlation was found between in-line and nasal cannula sampling measurements of EtCO2 (R = 0.81) (Figure 4A). A Bland-Altman analysis was performed to compare the difference between the 2 sampling techniques across the range of mean EtCO2 values recorded (Figure 4B). In-line sampling capnography measurements were found to be consistently higher than nasal cannula measurements (bias = 4.0 mm Hg, solid line; 99% limits of agreement = −3.6 to 11.6 mm Hg, dashed lines).
Comparing RVM to capnography data across a wide range of RRs, our data demonstrate that the capnograph displayed low instrument sensitivity in response to changes in ventilation (−0.71 ± 0.11 and −0.55 ± 0.11 mm Hg 1 L/min for nasal and in-line sampling, respectively). We saw small changes in EtCO2 associated with large changes in MV. For example, MV was nearly 4 times lower during hypoventilation (2.1 L/min) compared to normal breathing (8.0 L/min) while EtCO2 was only 10% higher during hypoventilation (38.3 mm Hg) compared to normal breathing (34.8 mm Hg). The measured capnograph instrument sensitivity was significantly <4 mm Hg/(L/min), the instrument sensitivity needed to reflect a clinically relevant change in MV of 1 L/min based on the EtCO2 accuracy of the capnograph used in this study14 (Figure 3A).
Our data also clearly demonstrate that while RRs are measured equally well by both RVM and capnography at steady state, respiratory changes are observed by the RVM more rapidly than by capnography. Specifically, the MV reached a new steady state value following a change in breathing rate in 37.7 ± 1.4 seconds while EtCO2 changes often failed to reach a new asymptote before a 2.5-minute threshold. During hypoventilation, CO2 takes time to build up in the expired air and therefore changes in EtCO2 are always delayed from changes in MV. The capnography-based RR measurements were also delayed following changes in breathing rates compared to the RVM-based RR measurements. This delay in RR reported by the capnograph is likely too short to have clinical relevance and is dependent on the difference between the algorithm used by the RVM and by the specific capnograph used in this study.
As anticipated, we noted a significant difference across capnography sampling techniques, even when used with cooperative subjects. The more commonly used nasal cannula sampling performed consistently less well when attempting to identify respiratory depression, yielding lower EtCO2 readings than in-line sampling, likely due to mixing of exhaled air with ambient air. Higher respiratory volumes seen with an in-line sensor were likely due to greater subject discomfort from the nose clip and the focus on maintaining the mouthpiece in a good position.
The introduction of noninvasive capnography has been proposed as a superior alternative to SpO2 with regard to monitoring ventilation.7–10 However, capnography has never achieved wide clinical adoption in the general hospital floor because of inaccuracies due to inadvertent cannula dislodgement, patient noncompliance of nasal cannulas, and complexity in interpreting CO2 waveforms.2,3 In addition, multiple studies have demonstrated that EtCO2 does not consistently reflect PaCO2,16–21 particularly in situations of hemodynamic instability.22 Moreover, elevated PaCO2 can occur in the face of either cardiac or respiratory failure and, as such, use of CO2 data for respiratory assessment can be confusing when there are potential alterations in cardiovascular status. Heines et al20 demonstrated that EtCO2 measurements are not reliable to estimate PaCO2 in postoperative ventilated cardiac surgery patients due to a high ventilation to perfusion ratio or hypovolemia.
Improving safety for nonintubated patients requires early detection. This goal might be best achieved by monitoring decreases in MV as opposed to waiting for an alarm related to a decrease in RR, an increase in EtCO2 provided by capnography data, or a decrease in oxygen saturation provided by pulse oxymetry.1,4,23 By alerting clinicians earlier, before respiratory failure manifests as a cardiopulmonary arrest, fewer rapid response team activations may be required, morbidity and mortality reduced, and cost of care decreased.24 Incorporation of MV data into systems that measure other vital variables, such as heart rate, arterial blood pressure, pulse oximetry, and/or temperature drives improvement in early warning scores developed to provide information far enough in advance of patient decompensation to permit intervention.23,24
The RVM has been shown to accurately measure respiratory volume metrics12,13 and has been successfully used in a variety of clinical scenarios. For example, RVM has been shown to accurately reflect opioid sensitivity in varied postanesthesia care unit populations, leading to protocols for patient-specific titration of opioids or changes to analgesic management and reducing the risk of consequent respiratory depression.25–27 When used to monitor respiratory mechanics in postextubation periods of predicted difficult airways or borderline extubation intensive care unit patients, the RVM could potentially prevent the morbidity associated with an emergent reintubation by converting the reintubation to a more controlled urgent reintubation.28 RVM has also been used successfully in patients with traumatic thoracic injuries to aid with the assessment of respiratory function.29 The work here provides basic data as to the relationship between RVM measurements and EtCO2 data in nonintubated patients. This understanding is critical for clinical use of this new technology.
There are several limitations to this study. First, the study was conducted on healthy volunteers. A study in nonintubated, sedated patients is planned to perform similar analyses in patients with the altered physiologies seen in clinical practice. Second, we could not adequately determine a false alarm rate in this study. This will be important to address in the hospital setting, given that alarm fatigue is a major problem facing clinicians currently and one of the reasons why pulse oximeters and capnographs are unpopular.30,31 Specifically, we could not quantify the RVM’s effectiveness at minimizing false alarms by filtering out motion artifacts and averaging over 30-second segments, which provides a balance between detecting rapid changes in ventilation while integrating over transient data spikes. Finally, in this study, the RVM was calibrated by a heated pneumotachometer for each patient to maximize its accuracy. This calibration adds an additional step to the measurement process which might be undesirable or difficult in a postoperative environment, where some accuracy may have to be sacrificed in favor of streamlining the workflow.
This study supports previous work that the noninvasive RVM provides valuable, accurate quantifiable data that can be used to assess real-time respiratory mechanics. RVM may more rapidly detect changes, with greater instrument sensitivity in nonintubated patients when compared to capnography, especially during variations in respiratory patterns, the time most relevant to patient care. These findings support the potential for RVM to become the preferred clinical tool in nonintubated patients.
Name: George W. Williams, II, MD.
Contribution: This author helped design the study, analyze the data, and write the manuscript.
Conflicts: George W. Williams, II, reported no conflicts of interest.
Name: Christy A. George, MD.
Contribution: This author helped write the manuscript.
Conflicts: Christy A. George reported no conflicts of interest.
Name: Brian C. Harvey, PhD.
Contribution: This author helped analyze the data and write the manuscript.
Conflicts: Brian C. Harvey worked for Respiratory Motion, Inc.
Name: Jenny E. Freeman, MD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Conflicts: Jenny E. Freeman worked for and has equity interest in Respiratory Motion, Inc.
This manuscript was handled by: Maxime Cannesson, MD, PhD.
1. Lee LA, Caplan RA, Stephens LS, Posner KL, Terman GW, Voepel-Lewis T, Domino KB. Postoperative opioid-induced respiratory depression: a closed claims analysis. Anesthesiology 2015;122:659–65.
2. Deiorio NM. Continuous end-tidal carbon dioxide monitoring for confirmation of endotracheal tube placement is neither widely available nor consistently applied by emergency physicians. Emerg Med J 2005;22:490–3.
3. Jabre P, Jacob L, Auger H, Jaulin C, Monribot M, Aurore A, Margenet A, Marty J, Combes X. Capnography monitoring in nonintubated patients with respiratory distress. Am J Emerg Med 2009;27:1056–9.
4. Lynn LA, Curry JP. Patterns of unexpected in-hospital deaths: a root cause analysis. Patient Saf Surg 2011;5:3.
5. Attin M, Cardin S, Dee V, Doering L, Dunn D, Ellstrom K, Erickson V, Etchepare M, Gawlinski A, Haley T, Henneman E, Keckeisen M, Malmet M, Olson L. An educational project to improve knowledge related to pulse oximetry. Am J Crit Care 2002;11:529–34.
6. Fu ES, Downs JB, Schweiger JW, Miguel RV, Smith RA. Supplemental oxygen impairs detection of hypoventilation by pulse oximetry. Chest 2004;126:1552–8.
7. Vargo JJ, Zuccaro G Jr, Dumot JA, Conwell DL, Morrow JB, Shay SS. Automated graphic assessment of respiratory activity is superior to pulse oximetry and visual assessment for the detection of early respiratory depression during therapeutic upper endoscopy. Gastrointest Endosc 2002;55:826–31.
8. Lightdale JR, Goldmann DA, Feldman HA, Newburg AR, DiNardo JA, Fox VL. Microstream capnography improves patient monitoring during moderate sedation: a randomized, controlled trial. Pediatrics 2006;117:e1170–8.
9. Miner JR, Heegaard W, Plummer D. End-tidal carbon dioxide monitoring during procedural sedation. Acad Emerg Med 2002;9:275–80.
10. Waugh JB, Epps CA, Khodneva YA. Capnography enhances surveillance of respiratory events during procedural sedation: a meta-analysis. J Clin Anesth 2011;23:189–96.
11. van Loon K, van Rheineck Leyssius AT, van Zaane B, Denteneer M, Kalkman CJ. Capnography during deep sedation with propofol by nonanesthesiologists: a randomized controlled trial. Anesth Analg 2014;119:49–55.
12. Voscopoulos C, Brayanov J, Ladd D, Lalli M, Panasyuk A, Freeman J. Special article: evaluation of a novel noninvasive respiration monitor providing continuous measurement of minute ventilation in ambulatory subjects in a variety of clinical scenarios. Anesth Analg 2013;117:91–100.
13. Voscopoulos CJ, MacNabb CM, Brayanov J, Qin L, Freeman J, Mullen GJ, Ladd D, George E. The evaluation of a non-invasive respiratory volume monitor in surgical patients undergoing elective surgery with general anesthesia. J Clin Monit Comput 2015;29:223–30.
14. Covidien. Capnostream 20 Portable Bedside Capnograph/Pulse Oximeter: Operator’s Manual. 2014.
15. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307–10.
16. Raemer DB, Francis D, Philip JH, Gabel RA. Variation in PCO2 between arterial blood and peak expired gas during anesthesia. Anesth Analg 1983;62:1065–9.
17. Russell GB, Graybeal JM. Reliability of the arterial to end-tidal carbon dioxide gradient in mechanically ventilated patients with multisystem trauma. J Trauma 1994;36:317–22.
18. Wahba RW, Tessler MJ. Misleading end-tidal CO2
tensions. Can J Anaesth 1996;43:862–6.
19. Poirier MP, Gonzalez Del-Rey JA, McAneney CM, DiGiulio GA. Utility of monitoring capnography, pulse oximetry, and vital signs in the detection of airway mishaps: a hyperoxemic animal model. Am J Emerg Med 1998;16:350–2.
20. Heines SJ, Strauch U, Roekaerts PM, Winkens B, Bergmans DC. Accuracy of end-tidal CO2
capnometers in post-cardiac surgery patients during controlled mechanical ventilation. J Emerg Med 2013;45:130–5.
21. Parrillo JE, Dellinger RP.Critical Care Medicine—Principles of Diagnosis and Management in the Adult. 20144th ed. Philadelphia, PA: Mosby.
22. Belenkiy S, Ivey KM, Batchinsky AI, Langer T, Necsoiu C, Baker W, Salinas J, Cancio LC. Noninvasive carbon dioxide monitoring in a porcine model of acute lung injury due to smoke inhalation and burns. Shock 2013;39:495–500.
23. Politano AD, Riccio LM, Lake DE, Rusin CG, Guin LE, Josef CS, Clark MT, Sawyer RG, Moorman JR, Calland JF; Predictive Monitoring in Patients with Trauma (PreMPT) Group. Predicting the need for urgent intubation in a surgical/trauma intensive care unit. Surgery 2013;154:1110–6.
24. Welch J, Kanter B, Skora B, McCombie S, Henry I, McCombie D, Kennedy R, Soller B. Multi-parameter vital sign database to assist in alarm optimization for general care units. J Clin Monit Comput 2015; [epub ahead of print].
25. Voscopoulos CJ, MacNabb CM, Freeman J, Galvagno SM Jr, Ladd D, George E. Continuous noninvasive respiratory volume monitoring for the identification of patients at risk for opioid-induced respiratory depression and obstructive breathing patterns. J Trauma Acute Care Surg 2014;77:S208–15.
26. Fleming E, Voscopoulos C, George E. Non-invasive respiratory volume monitoring identifies opioid-induced respiratory depression in an orthopedic surgery patient with diagnosed obstructive sleep apnea: a case report. J Med Case Rep 2015;9:94.
27. Voscopoulos C, Theos K, Tillmann Hein HA, George E. A risk stratification algorithm using non-invasive respiratory volume monitoring to improve safety when using post-operative opioids in the PACU. J Clin Monit Comput 2016; [epub ahead of print].
28. Williams GW, Sharma A, Chen PR. The application of a novel non-invasive respiratory monitor (ExSpironTM) in the extubation of a patient with persistent upper airway swelling. J Med Cases 2014;5:388–91.
29. Galvagno SM, Brayanov J, Corneille MG, Voscopoulos CJ, Sordo S, Ladd D, Freeman J. Non-invasive respiratory volume monitoring in patients with traumatic thoracic injuries. Trauma 2014;17:219–23.
30. Drew BJ, Harris P, Zègre-Hemsey JK, Mammone T, Schindler D, Salas-Boni R, Bai Y, Tinoco A, Ding Q, Hu X. Insights into the problem of alarm fatigue with physiologic monitor devices: a comprehensive observational study of consecutive intensive care unit patients. PLoS One 2014;9:e110274.
© 2017 International Anesthesia Research Society
31. Petterson MT, Begnoche VL, Graybeal JM. The effect of motion on pulse oximetry and its clinical significance. Anesth Analg 2007;105:S78–84.