Unusual
Life-Threatening Hypercarhia During Cardiac Anesthesia and Cardiopulmonary Bypass D. Kubiak, MD,T.K. Keane, MD, G. Made,
A
N INCREASE IN the arterial carbon dioxide tension (P,CO,) during cardiopulmonary bypass (CPB) is an anesthetic emergency. The anesthesiologist should consider many causes including equipment failure and malignant hyperthermia.’ An unusual cause for a marked increase in P,CO, because of a unique coincidence is reported. The investigation of the cause led the anesthesiologist to a faulty gas supply source originating from medical equip-
ment outside the operating room. CASE
REPORT
A 70-year-old woman, weighing 67 kg, was scheduled for an elective coronary artery bypass graft procedure. After premedication with 10 mg of diazepam, anesthesia was induced with sufentanil, 350 pg, midazolam, 2 mg, and pancuronium bromide, 6 mg. Anesthesia was maintained with a continuous midazolami sufentanil infusion, and 50% air/oxygen. Hemodynamic monitoring included central venous, radial artery and pulmonary artery pressure recording; and temperatures from the nasopharynx, rectum, and left thenar site were measured. The patient was ventilated with a Servo 900B ventilator (Siemens, Elema, Sweden). A Siemens 940 ventilation computer and a Siemens 930 mainstream infra-red CO, analyzer (Siemens, Elema, Sweden) were used to monitor ventilatory parameters. Pressures, pulse oximetry, and capnography were registered every minute, while temperatures, inspiratory oxygen concentration, and ventilation parameters were recorded every 15 minutes with an automatic documentation system? During perfusion a Cobe Excel membrane oxygenator was used. A capnograph (Gould Godart, Mark II; Bilthoven, The Netherlands) analyzed CO, excretion in the oxygenator exhaust gas flow. Acid-base control was managed according to the alpha-stat principle. The case was uneventful until 2 minutes before incision of the pericardium, when an abrupt drop in end-tidal CO, from 46 mm Hg to 0 was noted. This was attributed to a temporary equipment malfunction, because the lungs were being ventilated, and the peak pressure and expired minute volume remained normal. Shortly thereafter, the pulmonary artery pressure (PAP) suddenly increased from 1618 mm Hg to 40/18 mm Hg. Three minutes later, systolic and mean arterial pressures also increased 19% and 6%, respectively. Diastolic pressure decreased 17%. Pulse oximetry dropped slightly, but remained satisfactory. This hemodynamic situation was puzzling. Insufficient anesthesia depth was considered and was treated by an additional bolus dose of 100 ug of sufentanil, together with repeated bolus doses of nitroglycerin (175 p,g in total). Further examination of the CO, analyzer was postponed as the increased pressures demanded attention and the patient was ready to go on CPB. However, the anesthesiologist did notice hyperemia of the face and left arm, and an increase of the peripheral temperature by 0.9”C to 33.8”C, which is unusual at this time. The nasopharyngeal temperature remained 352°C. The oxygen blender was set to 100% 0, for the transition to bypass. This resulted in a tremendous increase in the end-tidal CO,. This phenomenon went unnoticed, but was recorded on the automatic documentation system. CPB commenced with a stable mean arterial pressure of 50 mm Hg. However, a high CO, content in the exhaust gas from the oxygenator was immediately noticed by the perfusionist and an arterial blood gas sample was obtained, while the flow through the gas inlet of the oxygenator was maximized to 15 L/min. The result
CP, R. Huet, MD, PhD
of the blood gas analysis showed a severe acidosis caused by a marked increase of P,CO, with a normal P,O, (Table 1). The CO, level seemed disproportionately high in relation to the temperature; therefore, the anesthesiologist doubted whether malignant hyperthermia had ensued, and no dantrolene was given. Increased intracranial pressure was feared, and the patient was put in reverse Trendelenburg position to promote cerebral venous drainage; 100 mL of 20% mannitol was given. In addition, THAM, 250 mL, was administrated to improve the acidosis. The perfusionist checked the heart-lung machine for an external cause of CO, contamination and analyzed the inlet gas flow. The air/oxygen mixture after the blender contained 15% CO,. Subsequently, the medical gas supply of the hospital pipeline system was identified as the source, and was immediately shut off from the heart-lung machine and the ventilator. The gas inlet of the heart-lung machine was switched to a small 2.5-L oxygen cylinder, which was available on the heart-lung machine. Thereafter, the P,CO, gradually dropped to acceptable levels (Table I), with maximum gas inflow. The operation was quickly and uneventfully finished and the patient was weaned from CPB as soon as the acid-base state had normalized an hour later. A dopamine infusion was necessary to improve cardiac contractility. The patient was ventilated with a portable ventilator (Oxylog, Drager Zoetermeer, The Netherlands) on 50% oxygen/air (cylinder). The postoperative course in the intensive care unit was uncomplicated, the patient was extubated the following morning, and transferred to the ward. No neurological sequelae occurred. Multiple measurements of the medical gas supply showed a diminishing CO, concentration coming from the air pipeline sockets in the surrounding operating rooms. Because the hospital was undergoing construction, it was initially suspected that accidental interference with the gas supply had occurred. However, investigation by the hospital technical service disclosed no manipulation with the pipelines that day, and no registration of alarms for pressure drops or change in oxygen availability. The only external source of the CO, supply initially identified was a 2.5-L CO, cylinder mounted on the heart-lung machine. The CO, cylinder was equipped with a pressure reduction valve set to 59 psi and connected to the oxygenator gas inlet via a ffowmeter and a Siemens membrane gas blender (Siemens, Elema, Sweden). However, no CO, was used for this patient and the CO, cylinder was full. To exclude this source as a cause, the CO, cylinder was opened and challenged for CO, leakage hack to the air pipeline system. For this purpose, the gas outlet of the oxygenator had to be artificially occluded, but no substantial leakage could be established. Finally, it was recalled that during the period of the accident, the assistance of an anesthesiologist in the adjacent cardiac catheterization laboratory was requested for the management of a patient who had clinically deteriorated. It turned out that this patient needed an intra-aortic balloon pump (IABP, system 83; Datascope, Horvelaken, The Netherlands), which held a 10-L cylinder of CO,. The
From the Depanment of Anesthesiology, Perfusion Section, and the Technical Department, State University Hospital, Groningen, The Netherlands. Address reprint requests to R. Huet, MD, PhD, Departmenf of Anesthesiology, State University Groningen, PO Box 30.001, 9700 RR Groningen, The Netherlands. Copyright 0 1992 by W.B. Saunders Company 1053-0770/92/0601-0018$03.00/0
Journalof Cardiothoracic and VascularAnesthesia, Vol6. No 1 (February), 1992: pp 73-75
73
74
KUBIAK ET AL
Table 1. Blood Gas Analysis A 7.53
PH P&D, (kPal
El
C
D
E
F
6.67
7.04
7.40
7.45
7.35
4.4
42.0
11.6
6.0
5.3
6.4
33.0
315.0
88.0
45.0
39.8
48.0
29.8
28.0
50.4
60.6
61.5
47.1
(mm Hg) 0, saturation (%)
223.4
210.0
378.0
454.5
461.3
353.3
Bicarb. (mmol/L)
23
(mm Hg)
service and a seal in the mechanical switch was found to be was postulated that this defective component could have the leak. Technical drawings of the hospital documented medical gas supply pipelines travelled directly from the catheterization laboratory to the operating rooms.
worn. It caused that the cardiac
DISCUSSION
P,O* (kPa)
Hb. (g/dL)
0.97 11.8
0.94 34 6.8
0.94 29 -
0.95 28 4.8
0.94 27
0.97 25
5.5
12.8
NOTE. Arterial blood gas analysis (1 kPa = 7.5 mm Hg). A, after induction of anesthesia;
6, 9 minutes after initiation of bypass, 25
minutes after incident; C, 12 minutes after B; D, 10 minutes after discovery of cause, 87 minutes after incident; E, before end of bypass; F, 15 minutes after bypass.
insertion of the IABP coincided precisely in time with the documen-
tation of the life-threatening increase of CO,. The anesthesiologist in the catheterization laboratory could not use the ventilator because the medical air supply socket was in use for the IABP. Therefore, the patient was ventilated by hand using 100% 0,, and arterial blood gas analysis showed no CO2 accumulation in that patient. The IABP was designed to be powered either by an electric motor or, for short interventions and transportation, by CO, from a 10-L cylinder. However, it had been modified to allow use of the medical compressed air supply from the hospital pipeline system, a less noisy power source for routine use. Air or CO, as the power source for the 1ABP was selected by means of a mechanical switch that controlled inflow from both sources (Fig 1). When the IABP was investigated, it was set to be operated by air from the pipeline. However, an interview with attendants showed that it was initially set to be operated by CO,, but that it had worked unsatisfactorily. No malfunction of the switch was found, but if the mechanical switch of the IABP was hand-held in the mid-position, a massive backflow of CO2 gas into the air pipeline system was confirmed at other sockets. The measured percentage at adjacent air outlets from the pipeline varied between 12% and 15% CO2 during this experiment. The CO, pressure reduction valve on the IABP was found to be set at 103 psi, while the hospital pressure supply was kept at 59 to 74 psi. The IABP was examined by the technical DRIVE GAS
mech.
cAR6oNDIOXIDE CYLINDERS
switch
to
wall
pressure
The P,CO, levels encountered in this case are rare. Heart rate increased SO%, but no dysrhythmias were observed. The pulmonary artery pressure also increased lOO%, and the effect of CO, on pulmonary vascular resistance was spectacular. Hypercapnia increases cerebral blood flow and intracranial pressure may increase, causing impaired cerebral perfusion. Malignant hyperthermia as the cause was rejected bccause the temperature change was not dramatic and was considered to be unrelated to the tremendous CO? increase. D’Ambra et al’ reported a hyperthermic cardiac patient during CPB with an increase of P,CO, of only one-fifth of this case. This case report demonstrates one of the advantages of an automatic data logging system. The exact time-related sequence of operative events facilitated interpretation of the data. An unobserved incident could be traced and clarified at a later stage. The infrared CO, analyzer (Siemens 930) used in this case has one flaw; it calibrates during inspiration. This principle of capnography causes the system to reset to a faulty end-tidal CO,, if CO, is present during inspiration. This was misleading, and the increase of P,CO, was not suspected before bypass. The CO, intoxication would have been shown, if a calibration was used independent of the respiratory cycle, such as during CPB when a continuous gas flow exists. In retrospect, the patient was intoxicated for approximately 10 minutes, until the blender was set to 100% oxygen before bypass. Datascope, the firm that markets the IABP, changed the design of the optional pneumatic switch. Additionally, a pneumatic diode was fitted in the air inlet of the IABP. Oxygen, medical air, and nitrous oxide are supplied to all
low
pressure
manifold
Fig 1. Position of the optional pneumatic mechanical switch on the IASP, Datascope System 83.
75
LIFE-THREATENING CO, INCIDENT DURING CPB
hospital facilities via pipelines. The medical air supply, delivered by compressors, and the oxygen supply are set at a pressure of 74 psi. The medical gas supply may drop to a minimum of 59 psi when hospital usage is high. Surprisingly, it was found that the CO, cylinder on the IABP was set to 103 psi and only was separated from the hospital pipe lines via a switch. Equipment failure can thus unknowingly pollute a remote work area through the gas supply system. The danger of pipelines without one-way valves connected to external power sources is reported here. This case report
illustrates the apparent need to understand the hospital medical gas distribution, and underscores the importance of portable oxygen cylinders. REFERENCES
1. D’Ambra MN, Donovan C, Harrell G, et al: Case Conference number 3. J Cardiothorac Anesth 4:386-399,199O 2. Karliczek GF, de Gem AF, Wiersma G, et al: Carola, a computer system for automatic documentation in anesthesia. J Clin Monit 4:211-221,1987
Life-Threatening Hypercarhia During Cardiac Anesthesia and Cardiopulmonary Bypass D. Kubiak, MD,T.K. Keane, MD, G. Made,
A
N INCREASE IN the arterial carbon dioxide tension (P,CO,) during cardiopulmonary bypass (CPB) is an anesthetic emergency. The anesthesiologist should consider many causes including equipment failure and malignant hyperthermia.’ An unusual cause for a marked increase in P,CO, because of a unique coincidence is reported. The investigation of the cause led the anesthesiologist to a faulty gas supply source originating from medical equip-
ment outside the operating room. CASE
REPORT
A 70-year-old woman, weighing 67 kg, was scheduled for an elective coronary artery bypass graft procedure. After premedication with 10 mg of diazepam, anesthesia was induced with sufentanil, 350 pg, midazolam, 2 mg, and pancuronium bromide, 6 mg. Anesthesia was maintained with a continuous midazolami sufentanil infusion, and 50% air/oxygen. Hemodynamic monitoring included central venous, radial artery and pulmonary artery pressure recording; and temperatures from the nasopharynx, rectum, and left thenar site were measured. The patient was ventilated with a Servo 900B ventilator (Siemens, Elema, Sweden). A Siemens 940 ventilation computer and a Siemens 930 mainstream infra-red CO, analyzer (Siemens, Elema, Sweden) were used to monitor ventilatory parameters. Pressures, pulse oximetry, and capnography were registered every minute, while temperatures, inspiratory oxygen concentration, and ventilation parameters were recorded every 15 minutes with an automatic documentation system? During perfusion a Cobe Excel membrane oxygenator was used. A capnograph (Gould Godart, Mark II; Bilthoven, The Netherlands) analyzed CO, excretion in the oxygenator exhaust gas flow. Acid-base control was managed according to the alpha-stat principle. The case was uneventful until 2 minutes before incision of the pericardium, when an abrupt drop in end-tidal CO, from 46 mm Hg to 0 was noted. This was attributed to a temporary equipment malfunction, because the lungs were being ventilated, and the peak pressure and expired minute volume remained normal. Shortly thereafter, the pulmonary artery pressure (PAP) suddenly increased from 1618 mm Hg to 40/18 mm Hg. Three minutes later, systolic and mean arterial pressures also increased 19% and 6%, respectively. Diastolic pressure decreased 17%. Pulse oximetry dropped slightly, but remained satisfactory. This hemodynamic situation was puzzling. Insufficient anesthesia depth was considered and was treated by an additional bolus dose of 100 ug of sufentanil, together with repeated bolus doses of nitroglycerin (175 p,g in total). Further examination of the CO, analyzer was postponed as the increased pressures demanded attention and the patient was ready to go on CPB. However, the anesthesiologist did notice hyperemia of the face and left arm, and an increase of the peripheral temperature by 0.9”C to 33.8”C, which is unusual at this time. The nasopharyngeal temperature remained 352°C. The oxygen blender was set to 100% 0, for the transition to bypass. This resulted in a tremendous increase in the end-tidal CO,. This phenomenon went unnoticed, but was recorded on the automatic documentation system. CPB commenced with a stable mean arterial pressure of 50 mm Hg. However, a high CO, content in the exhaust gas from the oxygenator was immediately noticed by the perfusionist and an arterial blood gas sample was obtained, while the flow through the gas inlet of the oxygenator was maximized to 15 L/min. The result
CP, R. Huet, MD, PhD
of the blood gas analysis showed a severe acidosis caused by a marked increase of P,CO, with a normal P,O, (Table 1). The CO, level seemed disproportionately high in relation to the temperature; therefore, the anesthesiologist doubted whether malignant hyperthermia had ensued, and no dantrolene was given. Increased intracranial pressure was feared, and the patient was put in reverse Trendelenburg position to promote cerebral venous drainage; 100 mL of 20% mannitol was given. In addition, THAM, 250 mL, was administrated to improve the acidosis. The perfusionist checked the heart-lung machine for an external cause of CO, contamination and analyzed the inlet gas flow. The air/oxygen mixture after the blender contained 15% CO,. Subsequently, the medical gas supply of the hospital pipeline system was identified as the source, and was immediately shut off from the heart-lung machine and the ventilator. The gas inlet of the heart-lung machine was switched to a small 2.5-L oxygen cylinder, which was available on the heart-lung machine. Thereafter, the P,CO, gradually dropped to acceptable levels (Table I), with maximum gas inflow. The operation was quickly and uneventfully finished and the patient was weaned from CPB as soon as the acid-base state had normalized an hour later. A dopamine infusion was necessary to improve cardiac contractility. The patient was ventilated with a portable ventilator (Oxylog, Drager Zoetermeer, The Netherlands) on 50% oxygen/air (cylinder). The postoperative course in the intensive care unit was uncomplicated, the patient was extubated the following morning, and transferred to the ward. No neurological sequelae occurred. Multiple measurements of the medical gas supply showed a diminishing CO, concentration coming from the air pipeline sockets in the surrounding operating rooms. Because the hospital was undergoing construction, it was initially suspected that accidental interference with the gas supply had occurred. However, investigation by the hospital technical service disclosed no manipulation with the pipelines that day, and no registration of alarms for pressure drops or change in oxygen availability. The only external source of the CO, supply initially identified was a 2.5-L CO, cylinder mounted on the heart-lung machine. The CO, cylinder was equipped with a pressure reduction valve set to 59 psi and connected to the oxygenator gas inlet via a ffowmeter and a Siemens membrane gas blender (Siemens, Elema, Sweden). However, no CO, was used for this patient and the CO, cylinder was full. To exclude this source as a cause, the CO, cylinder was opened and challenged for CO, leakage hack to the air pipeline system. For this purpose, the gas outlet of the oxygenator had to be artificially occluded, but no substantial leakage could be established. Finally, it was recalled that during the period of the accident, the assistance of an anesthesiologist in the adjacent cardiac catheterization laboratory was requested for the management of a patient who had clinically deteriorated. It turned out that this patient needed an intra-aortic balloon pump (IABP, system 83; Datascope, Horvelaken, The Netherlands), which held a 10-L cylinder of CO,. The
From the Depanment of Anesthesiology, Perfusion Section, and the Technical Department, State University Hospital, Groningen, The Netherlands. Address reprint requests to R. Huet, MD, PhD, Departmenf of Anesthesiology, State University Groningen, PO Box 30.001, 9700 RR Groningen, The Netherlands. Copyright 0 1992 by W.B. Saunders Company 1053-0770/92/0601-0018$03.00/0
Journalof Cardiothoracic and VascularAnesthesia, Vol6. No 1 (February), 1992: pp 73-75
73
74
KUBIAK ET AL
Table 1. Blood Gas Analysis A 7.53
PH P&D, (kPal
El
C
D
E
F
6.67
7.04
7.40
7.45
7.35
4.4
42.0
11.6
6.0
5.3
6.4
33.0
315.0
88.0
45.0
39.8
48.0
29.8
28.0
50.4
60.6
61.5
47.1
(mm Hg) 0, saturation (%)
223.4
210.0
378.0
454.5
461.3
353.3
Bicarb. (mmol/L)
23
(mm Hg)
service and a seal in the mechanical switch was found to be was postulated that this defective component could have the leak. Technical drawings of the hospital documented medical gas supply pipelines travelled directly from the catheterization laboratory to the operating rooms.
worn. It caused that the cardiac
DISCUSSION
P,O* (kPa)
Hb. (g/dL)
0.97 11.8
0.94 34 6.8
0.94 29 -
0.95 28 4.8
0.94 27
0.97 25
5.5
12.8
NOTE. Arterial blood gas analysis (1 kPa = 7.5 mm Hg). A, after induction of anesthesia;
6, 9 minutes after initiation of bypass, 25
minutes after incident; C, 12 minutes after B; D, 10 minutes after discovery of cause, 87 minutes after incident; E, before end of bypass; F, 15 minutes after bypass.
insertion of the IABP coincided precisely in time with the documen-
tation of the life-threatening increase of CO,. The anesthesiologist in the catheterization laboratory could not use the ventilator because the medical air supply socket was in use for the IABP. Therefore, the patient was ventilated by hand using 100% 0,, and arterial blood gas analysis showed no CO2 accumulation in that patient. The IABP was designed to be powered either by an electric motor or, for short interventions and transportation, by CO, from a 10-L cylinder. However, it had been modified to allow use of the medical compressed air supply from the hospital pipeline system, a less noisy power source for routine use. Air or CO, as the power source for the 1ABP was selected by means of a mechanical switch that controlled inflow from both sources (Fig 1). When the IABP was investigated, it was set to be operated by air from the pipeline. However, an interview with attendants showed that it was initially set to be operated by CO,, but that it had worked unsatisfactorily. No malfunction of the switch was found, but if the mechanical switch of the IABP was hand-held in the mid-position, a massive backflow of CO2 gas into the air pipeline system was confirmed at other sockets. The measured percentage at adjacent air outlets from the pipeline varied between 12% and 15% CO2 during this experiment. The CO, pressure reduction valve on the IABP was found to be set at 103 psi, while the hospital pressure supply was kept at 59 to 74 psi. The IABP was examined by the technical DRIVE GAS
mech.
cAR6oNDIOXIDE CYLINDERS
switch
to
wall
pressure
The P,CO, levels encountered in this case are rare. Heart rate increased SO%, but no dysrhythmias were observed. The pulmonary artery pressure also increased lOO%, and the effect of CO, on pulmonary vascular resistance was spectacular. Hypercapnia increases cerebral blood flow and intracranial pressure may increase, causing impaired cerebral perfusion. Malignant hyperthermia as the cause was rejected bccause the temperature change was not dramatic and was considered to be unrelated to the tremendous CO? increase. D’Ambra et al’ reported a hyperthermic cardiac patient during CPB with an increase of P,CO, of only one-fifth of this case. This case report demonstrates one of the advantages of an automatic data logging system. The exact time-related sequence of operative events facilitated interpretation of the data. An unobserved incident could be traced and clarified at a later stage. The infrared CO, analyzer (Siemens 930) used in this case has one flaw; it calibrates during inspiration. This principle of capnography causes the system to reset to a faulty end-tidal CO,, if CO, is present during inspiration. This was misleading, and the increase of P,CO, was not suspected before bypass. The CO, intoxication would have been shown, if a calibration was used independent of the respiratory cycle, such as during CPB when a continuous gas flow exists. In retrospect, the patient was intoxicated for approximately 10 minutes, until the blender was set to 100% oxygen before bypass. Datascope, the firm that markets the IABP, changed the design of the optional pneumatic switch. Additionally, a pneumatic diode was fitted in the air inlet of the IABP. Oxygen, medical air, and nitrous oxide are supplied to all
low
pressure
manifold
Fig 1. Position of the optional pneumatic mechanical switch on the IASP, Datascope System 83.
75
LIFE-THREATENING CO, INCIDENT DURING CPB
hospital facilities via pipelines. The medical air supply, delivered by compressors, and the oxygen supply are set at a pressure of 74 psi. The medical gas supply may drop to a minimum of 59 psi when hospital usage is high. Surprisingly, it was found that the CO, cylinder on the IABP was set to 103 psi and only was separated from the hospital pipe lines via a switch. Equipment failure can thus unknowingly pollute a remote work area through the gas supply system. The danger of pipelines without one-way valves connected to external power sources is reported here. This case report
illustrates the apparent need to understand the hospital medical gas distribution, and underscores the importance of portable oxygen cylinders. REFERENCES
1. D’Ambra MN, Donovan C, Harrell G, et al: Case Conference number 3. J Cardiothorac Anesth 4:386-399,199O 2. Karliczek GF, de Gem AF, Wiersma G, et al: Carola, a computer system for automatic documentation in anesthesia. J Clin Monit 4:211-221,1987
