Secondary Logo

Journal Logo

What Is the Accuracy of the High-Fidelity METI Human Patient Simulator Physiological Models During Oxygen Administration and Apnea Maneuvers?

Lejus, Corinne PhD, MD*; Magne, Cécile*; Brisard, Laurent MD*; Blondel, Pascal; Asehnoune, Karim PhD, MD*; Péan, Didier MD*

doi: 10.1213/ANE.0b013e3182991c2d
Technology, Computing, and Simulation: Research Report
Free

BACKGROUND: A widely used physiological simulator is generally accepted to give valid predictions of oxygenation status during disturbances in breathing associated with anesthesia. We compared predicted measures with physiological measurements available in the literature, or derived from other models.

METHODS: Five studies were selected from the literature which explored arterial oxygenation, with or without preoxygenation, in clinical situations or through mathematical modeling as well as the evolution of the fraction of expired oxygen (FEO2) during preoxygenation maneuvers. Scenarios from these studies were simulated on the METI-Human Patient Simulator™ simulator, and the data were compared with the results in the literature.

RESULTS: Crash-induction anesthesia without preoxygenation induces an O2 pulse saturation (SpO2) decrease that is not observed on the METI simulator. In humans, after 8 minutes of apnea, SpO2 decreased below 90% while the worst value was 95% during the simulation. The apnea time to reach 85% was less with obese patients than with healthy simulated patients and was shortened in the absence of preoxygenation. However, the data in the literature include METI simulator confidence interval 95% values only for healthy humans receiving preoxygenation. The decrease in PaO2 during 35-second apnea started at end-expiration was slower on the METI simulator than the values reported in the literature. FEO2 evolution during preoxygenation maneuvers on the METI simulator with various inspired oxygen fractions (100%, 92%, 84%, and 68%) was very close to those reported in humans when perfect mask seal is provided. In practice, this seal is impossible to obtain on the METI simulator.

CONCLUSIONS: SpO2 decreased much later during apnea on the METI simulator than in a clinical situation, whether preoxygenation was performed or not. The debriefing after simulation of critical situations or the use of the METI simulator to test a new equipment must consider these results.

Published ahead of print June 6, 2013.

From the *C.H. U. de Nantes, Service d’Anesthésie et de Réanimation Chirurgicale, Hôtel-Dieu Hôpital Mère Enfant, Place Alexis Ricordeau, Nantes; and Draeger Simulation Center, Paris, France.

Accepted for publication April 3, 2013.

Published ahead of print June 6, 2013.

Funding: This work was supported by the Service d’Anesthésie et de Réanimation Chirurgicale, Hôtel Dieu Hôpital Mère Enfant, Place Alexis Ricordeau, Nantes, France. This sponsor had no role in the design, methods, subject recruitment, data collection, analysis, and preparation of the paper.

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

Reprints will not be available from the authors.

Address correspondence to Corinne Lejus, PhD, MD, Service d’Anesthésie et de Réanimation Chirurgicale, Hôtel-Dieu Hôpital Mère Enfant, Place Alexis Ricordeau, Nantes, F-44000 France. Address e-mail to corinne.lejus@chu-nantes.fr.

Full-scale mannequin-based simulators were designed as educational tools. The METI Human Patient Simulator™ (HPS) (version 6, Medical Education Technologies, Sarasota, FL) is a sophisticated mannequin with integrated physiological and pharmacological models. The METI simulator is used worldwide for teaching with great clinical accuracy.1,2 High-fidelity simulators could also be considered experimental tools for exploring pathophysiological conditions and providing clinically relevant information in cases such as carbon monoxide intoxication.3 They could also be used to evaluate clinical scenarios or to test medical devices, especially in critical situations that cannot be studied in humans for ethical reasons.4,5 This implies that the METI simulator can be trusted to reproduce a real clinical situation. However, the functionalities of the gas exchange hybrid model of the METI simulator have been only partially described6 and the accuracy of this model has never been confirmed. The purpose of this study was to quantitatively investigate the accuracy of the METI simulator oxygenation model when common events such as preoxygenation or apnea are performed.

Back to Top | Article Outline

METHODS

The patient was simulated by using the METI-HPS located in the Dräger® Simulation Center (Anthony, France). This full-scale high-fidelity simulator used a hybrid (mathematical and mechanical) self-regulating lung model,6 with a real physical system to model pulmonary gas exchange and lung mechanics of a simulated patient. Briefly, uptake and excretion of oxygen (O2), carbon dioxide (CO2), nitrous oxide, and volatile anesthetic gases were physically created based on the measured concentrations in the bellows of the simulated lung and in a software model representing uptake, distribution, storage, consumption, and/or production in the body. Lung perfusion was also accounted for in this model by modeling the cardiovascular subsystem of the patient being simulated. Gas inspired values were directly measured in the bellows representing the lungs, and the excretion calculated to produce realistic alveolar fractions. The O2 pulse saturation (SpO2) measurement was simulated with the data from the model and was continuously recorded by METI-HPS software in a physiologic data log file for each procedure. The models were integrated in this software with the possibility to create new patients, when modifying the pulmonary or hemodynamic variables (O2 consumption, pulmonary shunt, weight, functional residual capacity).

The design of the study was based on the method used by Hardman et al.7 to validate the Nottingham Physiology Simulator by reproducing previous clinical studies.7 Eight studies were examined.8–15 Five studies were selected,9–12 and 3 were excluded because it was not possible to reproduce these studies on the simulator.8,13,15 The investigators (1 senior anesthesiologist and 1 resident in anesthesia) who were in charge of planning and performing the simulations were blinded to the results of the clinical studies.

Back to Top | Article Outline

Scenario 1

In the clinical study by Drummond and Park,9 no preoxygenation was performed. The simulator software was initialized, and Scenario 1 was started after the oxygen alveolar pressure (PAO2) was stabilized in ambient air. Successively, IV boluses of thiopental (4 mg/kg) and succinylcholine (100 mg) were given over a period of 15 seconds. The mannequin was insufflated once by facemask with an FIO2 of 1, and the trachea was intubated with an oral tube of internal diameter 8 mm (Portex™, Smiths Medical, Ashford, Kent, United Kingdom). The black mark was positioned at the level of the vocal cords, and the cuff was inflated. The tube was cut at 22 cm, and controlled ventilation was applied with an FIO2 of 1 and a tidal volume of 8 mL/kg at a rate of 12 per minute (Dräger, Primus IE™). SpO2 was recorded at baseline (T0) after stabilization in air, immediately after drug injection and disappearance of respiratory movements (TIND), 1 minute after the previous measure (T1), after tracheal intubation and after a 30-second period of controlled ventilation. The Drummond and Park study9 included 20 patients (12 females and 8 males). The mean percentage of expected weight was respectively 104% (84%–150%). Before each test, weight was fitted to the clinical data of the corresponding patient.

Back to Top | Article Outline

Scenario 2

In the study by Naguib et al.,10 after preoxygenation in 100% O2 to an end-tidal O2 fraction (FetO2) above 90%, induction was performed using fentanyl (2 µg/kg) and propofol (2 mg/kg) followed by succinylcholine (1 mg/kg) or saline (control). When apnea occurred, the facemask was removed and SpO2 was monitored for the next 8 minutes. Ventilation was manually assisted only when SpO2 decreased to 90%. In the Naguib et al. study,10 20 patients were included in each control group and succinylcholine (1 mg/kg) group. In the present study, we planned to achieve an equivalent number of tests and to adjust the weight of the simulator to the clinical data of the corresponding patient.

Back to Top | Article Outline

Scenario 3

Farmery and Roe11 constructed a mathematical model to describe the rate of desaturation during apnea in various pathophysiological states. They calculated the apnea time required to reduce SpO2 to 85% for a “standard” adult (body weight 70 kg) and an obese adult (body weight 127 kg) both with and without preoxygenation. In the present study, we planned to determine this apnea time in a simulated standard adult and an obese adult on the METI simulator. Cardiac output, alveolar volume, and O2 consumption of the METI simulator were matched with the clinical set as follows: cardiac output was set at 4.9 L/min, alveolar volume at 2.5 L, and O2 consumption at 0.25 L/min for the standard adult. For the obese adult, cardiac output was set at 7 L/min, alveolar volume at 1 L, and O2 consumption at 0.378 L/min. Preoxygenation with an FIO2 of 1 was performed using an 18 L/min O2 gas flow until an FetO2 of 87% was obtained. Regarding the experiment without preoxygenation, FetO2 concentration was set at 13% and end-tidal CO2 fraction was set at 35 mm Hg.

Back to Top | Article Outline

Scenario 4

Sasse et al.12 quantified the change in oxygen arterial partial pressure (PaO2) occurring during a 35-second breath-holding period initiated at functional residual capacity. A clinical dataset was obtained from 8 patients (median age 42 years, range 18–63 years) with normal spirometry. Simulation was conducted with the “Standard Man” model of a 70-kg body weight patient. The initial physiological variables were adjusted to the mean value measured in the clinical series (PaO2 110 mm Hg, PaCO2 37 mm Hg, pH 7.43).

Back to Top | Article Outline

Scenario 5

McGowan and Skinner14 studied the changes in FetO2 in 16 healthy male volunteers (mean weight 69.1 kg) breathing 100%, 92%, 84%, and 68% O2 using a tight-fitting facemask. Because the METI simulator has a poor facemask seal (leak pressure around 0 cm H2O), the experiment was performed after orotracheal intubation of the mannequin (internal diameter 8 mm, inflated cuff and tube cut at 22 cm). A fresh gas flow of 10 L/min was administered using a reservoir bag of 4 L, filled with an appropriate mixture before each trial. The end-tidal O2 concentration was recorded every 20 seconds for 5 minutes. Simulation was conducted with the Standard Man model of a 70-kg body weight patient. Before starting the experiments, to verify that orotracheal intubation did not influence the results, physiological variables (tidal volume, respiratory rate, O2 and CO2 alveolar and arterial pressure, SpO2) were compared on the METI simulator during spontaneous ventilation with and without orotracheal intubation. Values were recorded every 5 seconds for 1 minute with 3 trials.

Back to Top | Article Outline

RESULTS

Scenario 1

Since the variability of the SpO2 was very low, the tests were arbitrarily interrupted by the investigators after the simulation of the first 9 patients. Their mean percentage of expected weight was 114% (92%–150%). Simulation of a rapid sequence induction without preoxygenation as described by Drummond and Park9 did not show significant desaturation contrary to what was observed in the clinical dataset, 1 minute after induction (Fig. 1).

Figure 1

Figure 1

Back to Top | Article Outline

Scenario 2

During apnea, induced after preoxygenation, Naguib et al.10 observed a decrease of SpO2 <90% during the 8-minute period of the study in 45% and 85% of the patients receiving, respectively, saline (n = 20) or succinylcholine (n = 20). On the METI simulator, no oxygenation alteration was observed during the entire procedure. After saline injection, SpO2 was maintained at 100% for 375 seconds and then decreased to 99% until the end of the study. After succinylcholine administration, a decrease of SpO2 to 99% occurred at 360 seconds and then SpO2 remained stable until the end of the study. In the clinical dataset, the lowest SpO2 was correlated with the time of spontaneous diaphragmatic movement reappearance which was observed after succinylcholine injection with a mean time of 4.8 ± 2.5 minutes. With the METI simulator, even if the recovery of spontaneous ventilation was inhibited by the maintenance of a simulated profound muscular block, a decrease of the SpO2 to 99%, 98%, 96%, and 95% was noted, respectively, only at 315, 405, 465, and 480 seconds. The initial planned number of tests was therefore not completed, and the experiment was stopped after these 2 observations.

Back to Top | Article Outline

Scenario 3

The apnea time required to reduce SpO2 to 85% (Table 1) when simulated with the METI simulator was shorter in obese adults than in standard patients and was shorter without preoxygenation than with preoxygenation. The values calculated with the model of Farmery and Roe11 were included in the 95% confidence interval obtained with the METI simulator only in the case of a standard adult with preoxygenation.

Table 1

Table 1

Back to Top | Article Outline

Scenario 4

Changes of PaO2 during a 35-second breath-holding period initiated at end-expiration were shown by Sasse et al.12 to be a curviline described by the following equation, PaO2 (t) = 58.2 + 55/(1 + e 0.18[t − 19.5]). The decrease in PaO2 observed through the METI simulator was considerably smaller (Fig. 2).

Figure 2

Figure 2

Back to Top | Article Outline

Scenario 5

PaO2 was slightly higher and PaCO2 was slightly lower when the mannequin was intubated than when it breathed spontaneously in air. However, the difference was not clinically relevant (Table 2). End-tidal O2 concentration values during preoxygenation with various inspired oxygen concentrations observed with the METI simulator are shown in Figure 3. They were well correlated with the clinical data produced by McGowan and Skinner14 (Fig. 4).

Figure 3

Figure 3

Figure 4

Figure 4

Table 2

Table 2

Back to Top | Article Outline

DISCUSSION

Five scenarios were simulated on the METI simulator, and the data were compared with the results in the literature. During preoxygenation maneuvers, the METI simulator provided valid predictions of expired oxygen fractions. Nevertheless, this supposes a perfect seal of the upper airway, which cannot be obtained by the simple application of a facemask. During apnea, the model did not yield the expected results. SpO2 decreased much later during apnea on the METI simulator than in a clinical situation.

Although simulations have become popular and effective in medicine to improve teaching and continuing medical education, few data are available about the validity of models.16,17 High-fidelity simulation training is increasingly used in anesthesia and emergency medicine,18–20 especially for difficult airway management training,21 and should provide a realistic model. Recently, a comparison of 4 high-fidelity simulators has shown that although the HPS (METI®, Sarasota, FL) is the most realistic, its pharyngeal airspace differed significantly from that of humans.17 The originality of our work is underlined by the absence of data about the validity of its oxygenation model while the first objective of difficult airway management is to provide adequate oxygenation in critical situations when one cannot ventilate and intubate.

Some biases should be discussed. The major difference between the METI simulator and the Nottingham Physiology Simulator22 is that the METI simulator mechanical pulmonary model has 2 bellows representing the lungs, without pulmonary alveoli. In addition, programming of the METI simulator does not make it possible to consider all of the physiologic variables described in human studies (e.g., body mass index). On the contrary, the availability of patient data was limited in the selected studies. However, this limitation does not alter the reliability of our conclusion with results matching for the 5 scenarios. As in the Hardman et al. investigations,7 the blinded design reinforces the value of our message. Finally, a tight seal between the facemask and the face is crucial for the optimization of preoxygenation even if many patients feel uncomfortable.14 An inadequate facemask seal is a common problem observed with the METI simulator, possibly related to the rigidity of materials used by the manufacturers. We hope that this issue will be corrected in new versions of the METI simulator. Finally, the study would be more informative if we also studied the evolution of arterial CO2 during apnea. However, although this parameter is interesting from a physiological point of view, in a critical situation of difficult airway management, it is less useful in a “cannot-intubate, cannot-ventilate” situation.

Preoxygenation increases oxygen reserves and delays the occurrence of desaturation during apnea.22 It is highly recommended in all patients insofar as ventilation and intubation difficulties cannot always be predicted. An end-expiratory oxygen >90% is clinically satisfactory: this can usually be reached after 3 minutes of normal breathing with an FIO2 of 1.23 The strong realistic changes of the fraction expired O2 obtained on the simulator could be exploited for the learning of the preoxygenation maneuver that must be optimized in all patients.24

The reasons for the slow O2 desaturation during apnea of the standard healthy model of the METI simulator are not clear. Oxygenation is affected by both circulatory and ventilator factors. We can argue the METI simulator does not appear to have more than 1 body compartment for its circulation. Thus, the impact of recirculation of desaturated blood will be less than in a real situation wherein 75% of the cardiac output goes to only 25% of the body mass and mixed venous saturation will decrease much faster than the METI model predicts.

This characteristic should be known by the instructors who control the METI simulator. A more realistic decrease of SpO2 can be obtained by modifying variables such as the size of the shunt, residual functional capacity, and O2 consumption. These adjustments could be the subject of further studies while waiting for the ideal simulated patient. The scenarios of emergencies in anesthesia often interfere with oxygenation variables. Patient simulators are currently hybrid models controlled by a computer. Low-fidelity systems require long and tedious programming. High-fidelity simulators are much more expensive but have the major advantage of the physiological models without some of the programming. The need for significant adjustments for important basic vital variables such as oxygenation negates much of their interest.

On one hand, “cannot-ventilate/cannot-intubate” situations are very stressful in anesthesia, even if they are simulated. A delayed decrease in SpO2 during a simulation could be considered an advantage since it provides more time to apply the proper strategy before circulatory arrest occurs. It is usually advisable to play simulation scenarios in real time. On the other hand, the educational impact of simulated patient death that may result from a real-time decrease in SpO2 is the subject of controversy. Debriefing with an expert is always of great importance after an airway-focused simulation session.18 At that moment, students must be informed that the occurrence of O2 desaturation during apnea is going to be longer than in real life. Finally, although the METI simulator may be used to evaluate the efficacy of ventilation and oxygenation devices,25 our results caution against the unqualified use of the METI simulator for such purposes.

Our results support those of several recent studies which suggest that the lack of realism of the simulator airway could compromise the validity of comparative trials of airway devices.16 Our results emphasize that model validation is an important step before any simulation process. In conclusion, our study underscored the need to focus future research on reproducing the most realistic human patient simulation possible.

Back to Top | Article Outline

DISCLOSURES

Name: Corinne Lejus, PhD, MD.

Contribution: This author constructed the design of the study, analyzed the data, and prepared the manuscript.

Attestation: Dr. Corinne Lejus attests to having approved the final manuscript, the integrity of the original data, and data analysis. She is the archival author.

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

Name: Cécile Magne.

Contribution: This author was in charge of conducting the study and data collection.

Attestation: Cécile Magne attests to having approved the final manuscript.

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

Name: Laurent Brisard, MD.

Contribution: This author contributed to data analysis and manuscript preparation.

Attestation: Dr. Laurent Brisard attests to having approved the final manuscript.

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

Name: Pascal Blondel.

Contribution: This author helped conduct the study and was in charge of operating the METI simulator.

Attestation: Blondel Pascal attests to having approved the final manuscript.

Conflicts of Interest: Pascal Blondel is employed by the Dräger Society, that was in the past distributor in France of the METI simulators. But that company no longer provides the distribution of simulators.

Name: Karim Asehnoune, PhD, MD.

Contribution: This author assisted in preparation of the manuscript.

Attestation: Dr. Karim Asehnoune attests to having approved the final manuscript.

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

Name: Didier Péan, MD.

Contribution: This author was in charge of conducting the study, the collection of data, and helped to prepare the manuscript.

Attestation: Dr. Didier Péan attests to having approved the final manuscript and reviewing the original study data. He also attests to the integrity of the original data and the analysis reported in this manuscript.

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

This manuscript was handled by: Dwayne R. Westenskow, PhD.

Back to Top | Article Outline

REFERENCES

1. Pratt SD. Focused review: simulation in obstetric anesthesia. Anesth Analg. 2012;114:186–90
2. Morgan PJ, Cleave-Hogg D, DeSousa S, Tarshis J. Identification of gaps in the achievement of undergraduate anesthesia educational objectives using high-fidelity patient simulation. Anesth Analg. 2003;97:1690–4
3. Cesari WA, Caruso DM, Zyka EL, Schroff ST, Evans CH Jr, Hyatt JP. Study of physiological responses to acute carbon monoxide exposure with a human patient simulator. Adv Physiol Educ. 2006;30:242–7
4. Wetsch WA, Carlitscheck M, Spelten O, Teschendorf P, Hellmich M, Genzwürker HV, Hinkelbein J. Success rates and endotracheal tube insertion times of experienced emergency physicians using five video laryngoscopes: a randomised trial in a simulated trapped car accident victim. Eur J Anaesthesiol. 2011;28:849–58
5. Nicholas TA 4th, Pang H, Bernhagen MA, Boedeker BH. Performance comparison of laryngoscopy and suction techniques in a hemorrhagic airway manikin simulator: direct laryngoscopy with Yankauer vs. Storz CMAC with attached suction tip. Stud Health Technol Inform. 2012;173:307–9
6. van Meurs WL, Good ML, Lampotang S. Functional anatomy of full-scale patient simulators. J Clin Monit. 1997;13:317–24
7. Hardman JG, Wills JS, Aitkenhead AR. Investigating hypoxemia during apnea: validation of a set of physiological models. Anesth Analg. 2000;90:614–8
8. Gold MI, Duarte I, Muravchick S. Arterial oxygenation in conscious patients after 5 minutes and after 30 seconds of oxygen breathing. Anesth Analg. 1981;60:313–5
9. Drummond GB, Park GR. Arterial oxygen saturation before intubation of the trachea. An assessment of oxygenation techniques. Br J Anaesth. 1984;56:987–93
10. Naguib M, Samarkandi AH, Abdullah K, Riad W, Alharby SW. Succinylcholine dosage and apnea-induced hemoglobin desaturation in patients. Anesthesiology. 2005;102:35–40
11. Farmery AD, Roe PG. A model to describe the rate of oxyhaemoglobin desaturation during apnoea. Br J Anaesth. 1996;76:284–91
12. Sasse SA, Berry RB, Nguyen TK, Light RW, Mahutte CK. Arterial blood gas changes during breath-holding from functional residual capacity. Chest. 1996;110:958–64
13. Berthoud MC, Peacock JE, Reilly CS. Effectiveness of preoxygenation in morbidly obese patients. Br J Anaesth. 1991;67:464–6
14. McGowan P, Skinner A. Preoxygenation–the importance of a good face mask seal. Br J Anaesth. 1995;75:777–8
15. Findley LJ, Ries AL, Tisi GM, Wagner PD. Hypoxemia during apnea in normal subjects: mechanisms and impact of lung volume. J Appl Physiol. 1983;55:1777–83
16. Schebesta K, Hüpfl M, Ringl H, Machata AM, Chiari A, Kimberger O. A comparison of paediatric airway anatomy with the SimBaby high-fidelity patient simulator. Resuscitation. 2011;82:468–72
17. Schebesta K, Hüpfl M, Rössler B, Ringl H, Müller MP, Kimberger O. Degrees of reality: airway anatomy of high-fidelity human patient simulators and airway trainers. Anesthesiology. 2012;116:1204–9
18. Wong DT, Lai K, Chung FF, Ho RY. Cannot intubate-cannot ventilate and difficult intubation strategies: results of a Canadian national survey. Anesth Analg. 2005;100:1439–46
19. Price JW, Price JR, Pratt DD, Collins JB, McDonald J. High-fidelity simulation in anesthesiology training: a survey of Canadian anesthesiology residents’ simulator experience. Can J Anaesth. 2010;57:134–42
20. Borges BC, Boet S, Siu LW, Bruppacher HR, Naik VN, Riem N, Joo HS. Incomplete adherence to the ASA difficult airway algorithm is unchanged after a high-fidelity simulation session. Can J Anaesth. 2010;57:644–9
21. Park CS, Rochlen LR, Yaghmour E, Higgins N, Bauchat JR, Wojciechowski KG, Sullivan JT, McCarthy RJ. Acquisition of critical intraoperative event management skills in novice anesthesiology residents by using high-fidelity simulation-based training. Anesthesiology. 2010;112:202–11
22. Hardman JG, Wills JS. The development of hypoxaemia during apnoea in children: a computational modelling investigation. Br J Anaesth. 2006;97:564–70
23. Nimmagadda U, Chiravuri SD, Salem MR, Joseph NJ, Wafai Y, Crystal GJ, El-Orbany MI. Preoxygenation with tidal volume and deep breathing techniques: the impact of duration of breathing and fresh gas flow. Anesth Analg. 2001;92:1337–41
24. Tanoubi I, Drolet P, Donati F. Optimizing preoxygenation in adults. Can J Anaesth. 2009;56:449–66
25. Vadodaria BS, Gandhi SD, McIndoe AK. Comparison of four different emergency airway access equipment sets on a human patient simulator. Anaesthesia. 2004;59:73–9
© 2013 International Anesthesia Research Society