Unintended hypocapnia during anesthesia is associated with the morbidity and mortality of patients and must be promptly diagnosed and treated.1 We present an unusual case of the Dräger Primus Anesthesia Workstation (Dräger Medical AG&Co KG, Lüberck, Germany) displaying false readings of low end-tidal carbon dioxide (Etco2) due to failure of the internal solenoid zero valve. We want to inform clinicians about this potential failure and include it in the differential diagnoses of hypocapnia.
An 81-year-old man (weight, 68 kg) underwent elective coronary artery bypass grafting surgery. Signed informed consent was obtained from the patient for this publication. The anesthetic workstation passed a routine self-test, which was performed immediately before surgery. Standard anesthesia monitors, a peripheral intravenous line, and a radial arterial line were placed before anesthesia induction.2 Induction of anesthesia was uneventful and an 8.0-mm, cuffed endotracheal tube was inserted into the trachea and secured at the 21-cm mark at the lips. After intubation, equal bilateral breath sounds were confirmed by auscultation. There was no evidence of cyanosis and the pulse-oximetry (Spo2) value was maintained at 100%. However, the Etco2 readings on the Dräger Primus anesthesia workstation patient gas monitor (PGM) were unusually low (11 mm Hg) immediately after intubation.
Consequently, a structured approach was used to identify the etiology of the hypocapnia. On reassessment, the patient appeared to be adequately anesthetized (heart rate, 55 bpm; blood pressure, 105/65 mm Hg; Spo2, 100%; temperature, 37°C) with a patent and correctly placed endotracheal tube. During the inspiratory ventilation cycle, there was a good expansion of the patient’s chest with equal breath sounds bilaterally without audible wheezing or crackles. The ventilator was set to volume control mode with the following: tidal volume, 500 mL; respiratory rate, 12 breaths/min; positive end-expiratory pressure, 4 cm H2O, and fresh gas flow rate, 2 L/min with a mixture of 60% oxygen (O2) in hospital air. The sevoflurane vaporizer was set to 1.5%. The Etco2gas sampling line appeared clear. However, it was flushed to eliminate potential obstruction due to condensation or the presence of internal debris. The water trap also appeared clear without any condensation. All connections within the circuit and gas sampling system were tightly secured without any line kinking. In addition, the components of the Etco2 gas sample line and water trap were replaced. Arterial blood gas examination was performed to determine potential causes of the observed hypocapnia. Arterial blood gas analysis demonstrated values within normal physiological ranges (pH, 7.42; arterial carbon dioxide pressure [Paco2], 41 mm Hg; arterial oxygen pressure [Pao2], 292 mm Hg). Despite normal physiological findings and stable equipment parameters, Etco2 readings on the gas monitor were persistently low (12–15 mm Hg). After eliminating all potential patient-, circuit-, or ventilator-related causes for the hypocapnia, another spirometry and gas analyzer module for the General Electric (GE) monitor (E-sCAIOV-00 Compact Airway Module, Carescape Modular Monitor B850, GE Health Care, Helsinki, Finland) was added to the workstation. Etco2 readings appeared normal (ie, 37–40 mm Hg). Moreover, the Etco2 readings corresponded with the results of arterial blood gas analysis. The surgery was completed successfully without further complications.
Subsequently, a biomedical engineer examined the malfunctioning Dräger Primus PGM equipment. The flow rate in the sample gas line was tested first, which demonstrated a reduced flow of approximately 60 mL/min (nominal flow rate, 200 mL/min).3 A calibration test was performed using a gas containing known concentrations of carbon dioxide (CO2), O2, nitrous oxide (N2O), and anesthetic agents (Calibration gas GE Healthcare, Part# 755571-HEL, GE Health Care). All results measured by the malfunctioning gas monitor were abnormally low. Based on the low sample flow rates and low gas calibration results, failure of the PGM sampling pump or the solenoid zero valve was suspected (Figure).3 Small taps were made on the pump and valve assembly with a screwdriver to release any potentially stuck valves. The PGM was then placed back into the Dräger-Primus unit and retested. Normal gas sample flows of 200 mL/min were observed, and the concentration readings of the calibration gas were within the manufacturer’s recommendations.
The potential deleterious effects of hypocapnia have been reviewed previously.4 Although hypocapnia under anesthesia may be therapeutic (ie, for acute treatment of increased intracranial pressure), its long-term effects are associated with complications and development of systemic diseases, including alkalosis, and cerebral and myocardial ischemia.4,5 Physiological or mildly elevated Etco2 levels (40–50 mm Hg) have been recommended during maintenance of anesthesia.6
A structured approach to intraoperative events enables logical deduction of the source of the problem, with the concurrent clinical management of the patient.7 A prioritized patient-circuit-machine approach helps to rule out acute clinical emergencies and illustrates the importance of considering all 4 components—the patient, circuit, anesthesia machine, and monitor—in an organized fashion. In the present case, potential life-threatening patient-related causes of hypocapnia were eliminated quickly (eg, esophageal intubation, cardiac arrest). Stepwise examination of the circuit, ventilator, and gas sampling lines for potential sources of the hypocapnia showed no evident fault.8 Hypocapnia in a stable anesthetized patient with normal arterial blood gas values and properly functioning circuit and ventilator led us to suspect the capnography monitor as the cause of the low Etco2 readings. This assumption was confirmed intraoperatively by adding a second gas analyzer. Moreover, postoperative failure analysis of the suspected defective device identified a fault in the working mechanism of the solenoid zero valve.
Zupan et al9 described abnormal Etco2 waveforms caused by leaks in the gas sampling line. A long, initial plateau phase of the Etco2 waveform followed by a short-lasting peak and low Etco2 values were associated with a loose connection between the sampling line and CO2 analyzer. In their study, the values of O2 and N2O gases measured by the machine were a function of the size of the leak and the peak inspiratory pressure. We observed no such abnormal Etco2 waveform in our case, and, moreover, leakage or loose connections within the capnography line or water trap were eliminated by replacing both components.
In this case, it is unlikely that the presently identified mode of failure could have been prevented or detected earlier. Our anesthesia technicians routinely perform a full self-test daily on every anesthesia workstation. This Dräger-Primus passed all self-tests performed preoperatively. A crude preinduction exhalation test through a surgical mask into the circuit or sample line is not performed routinely at our institution for sanitary reasons. Furthermore, it only yields a qualitative trace presence of Etco2 because of the incomplete seal around the surgical mask. Our patient showed no loss of the Etco2 tracing; only the magnitude of the reading was erroneous.
Ultimately, the cause of hypocapnia in the present case was a malfunctioning solenoid zero valve in the PGM (Figure; component 7). The valve failed to close fully, and therefore, allowed entrainment of room air within the patient’s gas sample. This diluted the concentration of CO2 within the gas sample analyzed by the monitor resulting in the false readings of low Etco2. Incomplete closure of the zero valve also reduced the patient’s gas sampling flow rate by the same mechanism. The PGM has a pressure transducer (Figure; component P), which is located downstream to the zero valve, just before the pump. Therefore, the pump would be able to draw gas from both room air and patients’ gas sample to achieve its prescribed flow of 200 mL/min without displaying an error. In this case, a flow of only 60 mL/min was contributed by the patient’s gas sampling line; the remaining 140 mL/min was drawn in from the room air through the malfunctioning solenoid zero valve.
Other makes or models of side stream capnography monitors with a similar design are susceptible to this type of failure. As a design improvement, an internal fault alarm triggered by the monitor and another pressure transducer before the zero valve could detect the reduction in the flow rate of gas sampled from patients.
The manufacturer of the gas monitor module currently does not recommend changing the solenoid valves for preventive maintenance service. However, because of this case, the solenoid zero and occlusion valves were replaced in all Dräger Primus anesthesia workstations at our institution and will be changed every 2–3 years as part of our internal policy for preventive maintenance.
This report illustrates an unusual cause of erroneous gas monitor capnography readings due to mechanical malfunction of the solenoid zero valve of the PGM. Clinicians must be aware of this potential failure and include this cause in their differential diagnoses of low Etco2 readings.
Name: Victor M. Neira, MD, MAEd.
Contribution: This author helped with the case description, gathering clinical information, writing drafts of the paper, reviewing the literature, and editing the manuscript.
Name: Brian Fagan, BMET.
Contribution: This author helped with the biomedical engineering technical aspects of the case, the figure of the monitors, and editing the manuscript.
Name: Andrew D. Milne, MD, MSc, PEng, FRCPC.
Contribution: This author helped with gathering clinical information, writing and editing the draft of the manuscript, and creating figures.
This manuscript was handled by: BobbieJean Sweitzer, MD, FACP.
1. Dony P, Dramaix M, Boogaerts JG. Hypocapnia measured by end-tidal carbon dioxide tension during anesthesia is associated with increased 30-day mortality rate. J Clin Anesth. 2017;36:123–126.
2. Dobson G, Chow L, Flexman A, et al. Guidelines to the practice of Anesthesia - Revised Edition 2019. Can J Anaesth. 2019;66:75–108.
3. Dräger Medical. Patient gas module. In: Technical documentation. Primus anaesthetic workstation. Lübeck, Germany: Dräger Medical AG & Co. Revision 6.0. Chapter4: 2341.
4. Laffey JG, Kavanagh BP. Hypocapnia. N Engl J Med. 2002;347:43–53.
5. Dudaryk R, Bodzin DK, Ray JJ, Jabaley CS, McNeer RR, Epstein RH. Low end-tidal carbon dioxide at the onset of emergent trauma surgery is associated with nonsurvival: a case series. Anesth Analg. 2017;125:1261–1266.
6. Way M, Hill GE. Intraoperative end-tidal carbon dioxide concentrations: what is the target? Anesthesiol Res Pract. 2011;2011:271539.
7. Runciman WB, Merry AF. Crises in clinical care: an approach to management. Qual Saf Health Care. 2005;14:156–163.
8. Chitilian HV, Kaczka DW, Vidal Melo MF. Miller RD, Cohen NH, Eriksson LI, Fleisher L, Wiener-Kronish JP, Young WL. Respiratory monitoring. In: Miller’s Anesthesia. 2015:Philadelphia: Elsevier, 1541–1579.
9. Zupan J, Martin M, Benumof JL. End-tidal CO2 excretion waveform and error with gas sampling line leak. Anesth Analg. 1988;67:579–581.