- Email: [email protected]
Accuracy and performance of a modified continuous intravascular blood gas monitoring device during thoracoscopic surgery
Accuracy and Performance of a Modified Continuous Intravascular Blood Gas Monitoring Device During Thoracoscopic Surgery Michael T. Ganter, MD,* Christoph K. Hofer, MD,† Andreas Zollinger, MD,† Thierry Spahr, MD,* Thomas Pasch, MD,* and Marco P. Zalunardo, MD* Objective: The only commercially available continuous intravascular blood gas monitoring system for adults, the Paratrend (Diametrics Medical Inc, High Wycombe, UK), was modified by the manufacturer to the Paratrend 7ⴙ (PT7ⴙ) in 1999. The aim of this study was to evaluate the modified probe over a wide range of blood gas and pH values during thoracoscopic surgery in a similar setup as done with the previous model. Design: Prospective methods comparison study. Setting: University hospital. Participants: Twenty-three patients. Interventions: Elective thoracoscopic surgery. Measurements and Main Results: One hundred thirteen PT7ⴙ readings and their corresponding arterial blood gas and pH measurements (ABGA) were evaluated. The ranges for ABGA values were 50 to 474 mmHg for PO2, 29 to 58 mmHg for PCO2, and 7.28 to 7.49 for pH. Bland and Altman
C
ONTINUOUS INTRAVASCULAR BLOOD gas monitoring (CIBM) is applied by clinicians in selected patient groups. In the operating room, CIBM is used during surgery when blood gas values change rapidly and unexpectedly (eg, during thoracic surgery with 1-lung ventilation [OLV], cardiovascular surgery, and organ transplantation). Furthermore, CIBM is used in critically ill patients needing a considerable number of blood gas determinations over a long period of time (eg, premature infants with severe cardiopulmonary disease and patients with acute respiratory distress syndrome, sepsis, or severe trauma). The only commercially available CIBM system for adults, Paratrend 7 (Diametrics Medical Inc, High Wycombe, UK), was modified by the manufacturer to the Paratrend 7⫹ (PT7⫹) in 1999. PT7⫹ is a multiparameter CIBM system incorporating 4 different sensors. To measure PO2, the previously used Clark electrode was replaced with an optical PO2 sensor (fluorescent quenching sensor, ruthenium dye in silicone matrix). Accordingly, the length of the sensor tip has been halved. The other sensors and the main characteristics of the probe remained unchanged including optodes to determine PCO2 and pH (absorbance sensors, phenol red in bicarbonate solution) and a thermocouple (copper, constantan) to measure temperature and allow temperature correction of blood gas values. The sensors are housed in a heparin-coated microporous polyethylene tube that is permeable to the analytes to be measured. The probe is 0.5 mm in diameter, suitable for insertion through a 20-G or larger size catheter. Further details of the probe have been described elsewhere.1,2 Supported by the local distributor of the manufacturer, clinical experiences from PT7⫹ with excellent results were recently published in neurosurgical patients in the operating room (OR) and the intensive care unit (ICU).3,4 However, no sponsor-independent study on this modified CIBM system, which is being used clinically in most countries, has been published so far. The aim of the study was to evaluate the accuracy and performance of the modified CIBM probe PT7⫹ over a wide range of blood gas and pH values in thoracoscopic surgery during OLV.
analysis revealed a bias ⴞ 2 standard deviation of ⴚ20 ⴞ 86 mmHg for PO2, 3 ⴞ 9 mmHg for PCO2, and ⴚ0.01 ⴞ 0.06 for pH. No specific complications attributable to the probe were observed. Conclusion: In patients undergoing thoracoscopic surgery with rapidly changing blood gas parameters, the PT7ⴙ device is a valuable trend indicator and hence may be helpful for clinical decision making. However, the underestimation of PO2 values by 20 mmHg on average and the wide limits of agreement documented in this study must be regarded as limiting factors. © 2004 Elsevier Inc. All rights reserved. KEY WORDS: anesthesia, blood, gas analysis, equipment, monitors, intensive care, monitoring, continuous intravascular blood gas, multiparameter intravascular sensor, electrodes, optodes METHODS With local ethics committee approval and written informed consent, 23 consecutive patients undergoing elective thoracoscopic surgery were enrolled in the study. The study population constitutes a representative segment of patients presenting for thoracoscopic surgery at the University Hospital Zurich. On the day of surgery, after arriving in the OR, a radial artery was cannulated (left/right ⫽ 12/11) using a 20-G catheter (Insyte, Becton Dickinson Vascular Access Inc., Sandy, UT). Then, a previously calibrated PT7⫹ probe was introduced through this cannula according to the manufacturer’s instructions. No in vivo “recalibration” (adjusting the original calibration curve using in vitro laboratory blood gas determinations) was done. Anesthesia and monitoring during surgery (Hellige monitor; VICOM-SM SMU 612; PPG Hellige, Freiburg, Germany) were performed according to institutional standards, including monitoring of invasive arterial pressure and central venous pressure. General anesthesia was induced intravenously using propofol (n ⫽ 14) or etomidate (n ⫽ 9) combined with fentanyl. Anesthesia was maintained with intravenous propofol (n ⫽ 13) or isoflurane (n ⫽ 10) combined with intermittent doses of fentanyl or a continuous remifentanil infusion depending on clinical needs. Atracurium (n ⫽ 17) or rocuronium (n ⫽ 6) was used for neuromuscular blockade. For postoperative pain management, thoracic epidural anesthesia was instituted in 6 patients using ropivacaine. PT7⫹ values were not used to guide patients’ therapy. Left-sided double-lumen endobronchial tubes (BRONCHOPART; Ruesch GmbH, Kernen, Germany, placed under fiberoptic control) were used for
From the *Institute of Anesthesiology, University Hospital Zurich, Zurich, Switzerland; and †Institute of Anesthesiology and Intensive Care Medicine, Triemli City Hospital Zurich, Zurich, Switzerland. Presented in part at the annual meeting of the American Society of Anesthesiologists, San Francisco, CA, October 11-15, 2003. Address reprint requests to Marco P. Zalunardo, MD, Institute of Anesthesiology, University Hospital Zurich, Raemistrasse 100, Ch8091 Zurich, Switzerland. E-mail: [email protected] © 2004 Elsevier Inc. All rights reserved. 1053-0770/04/1805-0008$30.00/0 doi:10.1053/j.jvca.2004.07.017
Journal of Cardiothoracic and Vascular Anesthesia, Vol 18, No 5 (October), 2004: pp 587-591
587
588
GANTER ET AL
Table 1. Demographic and Surgical Characteristics Age (y) Female/male ratio (n) ASA status; I/II/III (n) Duration of anesthesia (min) Duration of one-lung ventilation (min) Surgical procedures (n) Lung volume reduction surgery Lung resection Pleurectomy Sympathectomy Thymectomy
55 ⫾ 15 (19-74) 10/13 5/9/9 168 ⫾ 48 (110-285) 51 ⫾ 29 (22-137) 8 8 4 2 1
NOTE. Data are presented as mean ⫾ SD (min/max).
volume-controlled mechanical ventilation during 2- and 1-lung ventilation (Siemens Servo 900D ventilator; Siemens Life Support Systems, Sona, Sweden). Initial (T1) data were obtained after induction of anesthesia under 2-lung ventilation. After positioning of the patients according to the surgical requirements and before the introduction of the trocars through the intercostal spaces, OLV was started and maintained during the thoracoscopic procedure. Five minutes (T2), 15 minutes (T3), and 25 minutes (T4) after the onset of OLV, further data were acquired. Last measurements (T5) were performed 10 minutes after 2-lung ventilation was re-established. Provided that system stability was indicated by the PT7⫹ computer, parameters of the PT7⫹ were stored at the measurement time points. Simultaneously, a 1-mL arterial blood sample was obtained for in vitro blood gas analysis from the radial artery cannula using a preheparinized syringe (QS 50, Radiometer Medicals, Copenhagen, Denmark). In vitro ABGA measurement was performed in a point-of-care laboratory blood gas analyzer (Ciba-Corning series 800, Chiron Diagnostics, Emeryville, CA) within 3 minutes by a laboratory technician who was unaware of the in vivo PT7⫹ values. According to the operator’s manual, daily quality control procedures were done to guarantee recommended accuracy and performance of the laboratory blood gas analyzer.5,6 PT7⫹ and ABGA parameters were both determined at 37°C. Statistical analysis was performed using StatView for Windows (Version 5.0.1, SAS Institute Inc, Cary, NC). Results are presented as mean ⫾ standard deviation (SD). According to Bland and Altman, bias (mean difference between PT7⫹ and ABGA parameters) ⫾ 2 SD (limits of agreement) with their 95% confidence intervals were calculated.7 The repeated measures design was accommodated over time by analysis of variance for repeated measures with post-hoc Bonferroni/ Dunn test. A p value ⬍0.05 was considered to be statistically significant. Hydrogen ion concentration (H⫹) was calculated from the pH values measured by both systems to allow calculating correlation coefficients with their 95% confidence intervals. RESULTS
Twenty-three patients were enrolled and analyzed. Patient demographic and surgical characteristics are presented in Table 1. A total of 113 PT7⫹ readings with corresponding ABGA values were obtained. Two sets of data were incomplete; as operation time was too short in 1 patient, and malfunctioning of the probe occurred in another patient after 100 minutes. No further specific complications attributable to the probe (eg, dampening of the arterial pressure catheter, clot formation, ischemic events) were observed. Blood gas and pH values varied extensively during the measurement period, showing a range of 50 to 474 mmHg for PO2, 29 to 58 mmHg for PCO2, and 7.28 to 7.49 for pH.
Bland and Altman analysis (Fig 1) revealed a bias ⫾ 2 SD of ⫺20 ⫾ 86 mmHg for PO2, meaning that PT7⫹ showed on average 20 mmHg lower PO2 values than those received by ABGA. In the clinically most relevant lower range of PO2 ⬍100 mmHg, bias ⫾ 2 SD was ⫺14.7 ⫾ 25.4 mmHg (PO2 mean, min/max ⫽ 83.0, 49.5/99.8 mmHg, n ⫽ 17). Regarding the performance of PCO2 and pH measurement, bias ⫾ 2 SD was 3 ⫾ 9 mmHg and ⫺0.01 ⫾ 0.06, respectively. Correlation between PT7⫹ and ABGA showed r2 values of 0.84 for PO2, 0.61 for PCO2, and 0.57 for pH (Table 2). The performance of measured values over time remained unchanged; bias ⫾ 2 SD from each measurement time point compared with the others showed no clinically or statistically relevant difference. Furthermore, influence of different hemoglobin levels (mean ⫾ 2 SD ⫽ 12.5 ⫾ 2.0 g/dL) and patients’ rectal temperatures (mean ⫾ 2 SD ⫽ 35.7° ⫾ 0.4°C) were studied regarding accuracy of CIBM measurements. There was no correlation between the bias ⫾ 2 SD of CIBM measurements and these variables. DISCUSSION
Currently, there is only 1 CIBM system commercially available for adults. To measure intravascular PO2 more precisely, the manufacturer modified the former hybrid probe PT7 a few years ago, replacing the Clark electrode by an optode. The main characteristics of the new pure optode-based probe, the PT7⫹, remained unchanged. Theoretical advantages of the modified design are that the PO2 optode is not consumed, resulting in a high stability over time and minimal drift. Furthermore, PO2 optodes enable further miniaturization of the probe (ie, the current length of the sensor tip could be reduced from 4 to 2 cm). By minimizing the intravascular part of the sensor, blood flow is better preserved; dampening of the arterial pressure waveform as well as clot formation may occur less frequently, and the sensor may be less susceptible to bending and kinking. The present study evaluated the modified CIBM probe PT7⫹ during thoracoscopic surgery. The PT7⫹ underestimated arterial PO2 by 20 mmHg, on average. However, limits of agreement were found to be poor, ranging from ⫺106 to 66 mmHg (Table 2, Fig 1). Malfunctioning of PT7⫹ occurred in 1 patient after 100 minutes, displaying no reliable blood gas values anymore. No further specific complications attributable to the PT7⫹ probe (eg, dampening of the arterial pressure tracing, clot formation, ischemic events) were observed in the studied period (mean anesthesia time 168 minutes, Table 1). Looking at the clinically important range of PO2 below 100 mmHg, much better results were obtained, but there was still a bias of ⫺15 mmHg and limits of agreement of ⫺40 to 10 mmHg. The promising in vitro data presented in the technical specification sheet of the PT7⫹ were not fulfilled; using gas-tonometered solutions, 95% confidence limits were ⫾5% or ⫾3 mmHg (whichever is greater) of the actual values for PO2 between 20 to 120 mmHg.2 Regarding measurement of oxygen saturation in this lower PO2 range, pulse oximetry may theoretically yield more accurate data than that obtained with the PT7⫹ in the present study. Pulse oximeters are found to be accurate within ⫾4% (2 SD) between 70 and 100% and ⫾6% (2 SD) between 50% to 70% SpO2, respectively.8,9 However, most of these data were collected in highly controlled nonclinical situations with
BLOOD GAS MONITORING DEVICE
589
Fig 1. Bland and Altman analysis showing bias (mean of differences) ⴞ 2 SD (upper and lower limit of agreement) versus mean values of arterial PO2, PCO2, and pH obtained by Paratrend 7ⴙ (PARA) compared with values obtained by in vitro blood gas analysis using a co-oximeter (ABGA) (n ⴝ 113).
healthy volunteers, and noninvasive standard monitoring (pulse oximetry, capnometry) did not accurately reflect the profound fluctuations in blood gas and pH values during thoracoscopic surgery with OLV.10 Comparison of blood gas variables from continuous intravascular sensors with those from laboratory blood gas analyz-
ers is a controversial issue. The accuracy of laboratory blood gas analyzers can be quantified in the laboratory where the values to be measured are known. However, this is not the case in clinical studies. The clinical performance of an optode-based intravascular probe must be judged in comparison with an electrode-based blood gas analyzer, for which clinical perfor-
590
GANTER ET AL
Table 2. Comparison of Measured Parameters by Paratrend 7ⴙ and In Vitro Blood Gas Analysis
PO2 (mmHg) PCO2 (mmHg) pH [H⫹] (nmol/L)
Mean ⫾ SD
Range
r (95% CI)
181.8 ⫾ 92.0 40.1 ⫾ 6.2 7.40 ⫾ 0.05 40.5 ⫾ 4.5
49.5/474.0 29.3/57.8 7.28/7.49 32.4/52.5
0.92 (0.88/0.94) 0.78 (0.70/0.85) - (-/-) 0.76 (0.67/0.83)
Bias (95% CI)
2 SD (95% CI)
⫺20.4 (⫺27.1/⫺13.7) 2.6 (1.9/3.3) ⫺0.01 (⫺0.02/⫺0.01) 1.4 (0.9/1.9)
85.8 (74.2/97.4) 8.9 (7.7/10.1) 0.05 (0.05/0.07) 6.5 (5.6/7.3)
NOTE. Values for mean ⫾ SD and range are those determined by in vitro arterial blood gas analysis. [H⫹] was calculated from the pH to allow correlation analysis (r ⫽ correlation coefficient; n ⫽ 113).
comparable to the given recommendations. There was no drift in accuracy of measured parameters over time. In contrast, the only 2 published clinical trials on PT7⫹ in neurosurgical patients showed excellent performance, even for PO2 measurements (Table 3).3,4 Several issues have to be considered comparing results of different studies. Both prior studies were performed by the same group, and the second paper was supported by the local distributor. The 2 studies “recalibrated” their probes every 12 hours as recommended, but nothing was stated about other “recalibrations.” By doing an in vivo “recalibration” immediately after insertion of the probe, an eventual offset of the probe could be lessened, hence improving the bias. However, repeatability (limits of agreement) would not improve. The present authors used a previously ex vivo calibrated PT7⫹ probe and did no in vivo “recalibration” (adjusting the original calibration curve using in vitro laboratory blood gas determinations). Blood gas variations in patients undergoing neurosurgery or staying in the neurosurgical ICU are much smaller compared with patients undergoing thoracoscopic surgery. Reaction time might not be sufficient in clinical situations in which blood gas values change rapidly and unex-
mance cannot be specifically quantified. Laboratory blood gas analyzers (even between individual analyzers of the same manufacturer) also have inconsistencies. Therefore, even with an accurate laboratory blood gas analyzer, the measured blood gas values may differ somewhat from the real blood gas values in vivo. Additionally, PT7⫹ measures blood gas and pH values at patient’s body core (intravascular) temperature and values are adjusted for 37°C, as opposed to laboratory blood gas analyzers measuring at 37°C. Because intermittently drawn blood samples analyzed by laboratory blood gas analyzers are the clinical standard of care, this procedure was used as a reference method to assess the accuracy of CIBM.2 No official recommendations concerning performance of continuous intravascular blood gas devices exist. However, several guidelines for laboratory blood gas analyzers have been published by the Clinical Laboratory Improvement Amendment/Health Care Finance Administration,11 the College of American Pathologists,6 and the Emergency Care Research Institute.2,5 If they are applied to the authors’ data, PO2 measurements failed to meet these recommendations, even with the better performance in the lower PO2 range. However, values for PCO2 and pH were acceptable and
Table 3. Evaluation Studies of the Paratrend 7 and Its Current Modification, the Paratrend 7ⴙ on Adult Patients PCO2 (mmHg)
PO2 (mmHg)
Studies with Paratrend 7 (hybrid probe) Clutton-Brock et al16 Venkatesh et al17 Venkatesh et al18 Abraham et al19 Nunomiya et al13 Liem et al20 Venkatesh et al21
Venkatesh et al22 Myles et al23 Endoh et al24 Zollinger et al1 Ishikawa et al25 Zaugg et al10 Myles et al12 Studies with Paratrend 7⫹ (pure optode probe) Menzel et al2 Menzel et al3 Present study
pH
Setting
Range
Bias
2 SD
Range
Bias
2 SD
ICU ICU ICU ICU ICU OR: LAP OR: CPB OR: CPB OR: CPB OR: ORTHO OR: CPB OR: CVS OR: OLV OR: OLV OR: OLV OR: OLV OR: OLV
-/93/202 60/447 -/45/542 -/130/464 137/510 61/300 -/-/26/120 46/458 47/449 72/255 37/625 -/-
2 6 3 ⫺2 ⫺7 3 4 1 ⫺9 ⫺1 3 ⫺1 0 2 ⫺22
55 41 51 40 68 90 54 39 86 9 73 80 42 83 108
-/25/51 26/52 -/29/68 -/24/41 21/41 21/50 -/-/28/72 31/71 25/57 30/47 27/56 -/-
1 2 2 1 1 ⫺2 2 1 2 1 ⫺1 0 2 1 1 0 ⫺2
ICU: NS OR: NS OR: OLV
53/486 75/375 50/474
⫺2 2 ⫺20
27 31 86
22/60 -/29/58
0 2 3
Range
Bias
2 SD
11 25 10 5 4 4 4 4 8 4 6 7 6 6 6 4 12
-/⫺0.01 7.21/7.53 0.01 7.31/7.61 0.01 -/0.01 7.19/7.60 0.00 -/0.00 7.36/7.57 0.02 7.28/7.53 0.01 7.10/7.55 0.02 -/0.02 -/0.02 7.12/7.59 0.01 7.19/7.50 ⫺0.02 7.30/7.49 0.00 7.34/7.47 ⫺0.01 7.24/7.51 0.01 -/0.01
0.12 0.14 0.12 0.05 0.04 0.05 0.10 0.12 0.12 0.06 0.07 0.07 0.07 0.04 0.08 0.04 0.10
7 4 9
7.19/7.85 0.00 -/⫺0.02 7.28/7.49 ⫺0.01
0.08 0.04 0.06
Abbreviations: ICU, intensive care unit; OR, operating room; LAP, laparoscopic surgery; CPB, cardiopulmonary bypass; ORTHO, orthopedic surgery; CVS, cardiovascular surgery; OLV, 1-lung ventilation; NS, neurosurgery.
BLOOD GAS MONITORING DEVICE
591
pectedly despite indicated system stability by the PT7⫹ monitor. Furthermore, the site of insertion was different in the studies and may influence the performance of the CIBM system. The most common site for CIBM measurement is the radial artery in adults (present study), but it has been shown that insertion of the probe into a femoral artery yields more reliable results compared with those from the radial artery, especially during low-flow states.2 The setting of the present study was similar to previous studies in thoracoscopic patients from the authors’ group (Table 3).1,10 The only modification in the study setup was an exchange of the laboratory blood gas analyzer, which hardly explains the large bias between CIBM and ABGA in PO2 measurement. A similar offset in PO2 between the 2 measurement techniques was also found in 1 study during lung transplantation.12 With the former PT7 device, a similar tendency of better performance for PO2 ⬍100 mmHg was observed in thoracoscopic (bias ⫺4 mmHg, limits of agreement ⫺21 to 13 mmHg)10 and critically ill patients (bias ⫺2 mmHg, limits of agreement ⫺15 to 11 mmHg).13 Interestingly, in the study on
neurosurgical ICU patients with the PT7⫹, the opposite was found; performance of PO2 measurement increasingly worsened below 100 mmHg PO2. The underestimation of PO2 values by the PT7⫹ may be explained partly by the “wall effect,” observed in early CIBM devices using photochemical sensors for PO2 measurement.14,15 If the sensor becomes attached to the wall of the vessel, thus measuring a combined blood and tissue PO2, the displayed oxygen value may be falsely low compared with the true arterial PO2. Hybrid probes (eg, PT7) were less susceptible to this artifact because the PO2 electrode had a larger surface area.2 Clinical indications for CIBM, in terms of evidence-based medicine, remain unclear because the cost-benefit ratio and the impact on patient outcome are unknown. In patients with rapidly changing blood gas parameters, the PT7⫹ device is a valuable trend indicator and hence may be helpful for clinical decision making. However, the underestimation of PO2 values and the wide limits of agreement documented in this study must be regarded as limiting factors.
REFERENCES 1. Zollinger A, Spahn DR, Singer T, et al: Accuracy and clinical performance of a continuous intra-arterial blood-gas monitoring system during thoracoscopic surgery. Br J Anaesth 79:47-52, 1997 2. Ganter M, Zollinger A: Continuous intravascular blood gas monitoring: development, current techniques, and clinical use of a commercial device. Br J Anaesth 91:397-407, 2003 3. Menzel M, Henze D, Soukup J, et al: Experiences with continuous intra-arterial blood gas monitoring. Minerva Anestesiol 67:325331, 2001 4. Menzel M, Soukup J, Henze D, et al: Experiences with continuous intra-arterial blood gas monitoring: Precision and drift of a pure optode system. Intensive Care Med 29:2180-2186, 2003 5. ECRI (Emergency Care Research Institute): Blood gas/pH analyzers. Health Devices 24:208-243, 1995 6. Shapiro BA: Evaluation of blood gas monitors: performance criteria, clinical impact, and cost/benefit. Crit Care Med 22:546-548, 1994 7. Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307-310, 1986 8. Jensen LA, Onyskiw JE, Prasad NG: Meta-analysis of arterial oxygen saturation monitoring by pulse oximetry in adults. Heart Lung 27:387-408, 1998 9. Soubani AO: Noninvasive monitoring of oxygen and carbon dioxide. Am J Emerg Med 19:141-146, 2001 10. Zaugg M, Lucchinetti E, Zalunardo MP, et al: Substantial changes in arterial blood gases during thoracoscopic surgery can be missed by conventional intermittent laboratory blood gas analyses. Anesth Analg 87:647-653, 1998 11. Medicare MaCp: Regulations implementing the Clinical Laboratory Improvement Amendments of 1988 (CLIA)—HCFA. Final rule with comment period. Fed Regist 57:7002-7186, 1992 12. Myles PS, Buckland MR, Weeks AM, et al: Continuous arterial blood gas monitoring during bilateral sequential lung transplantation. J Cardiothorac Vasc Anesth 13:253-257, 1999 13. Nunomiya S, Toshihide T, Masaru T, et al: Clinical evaluation of a new continuous intraarterial blood gas monitoring system in the intensive care setting. J Anesth 10:163-169, 1996
14. Mahutte CK: On-line arterial blood gas analysis with optodes: Current status. Clin Biochem 31:119-130, 1998 15. Wahr JA, Tremper KK: Continuous intravascular blood gas monitoring. J Cardiothorac Vasc Anesth 8:342-353, 1994 16. Clutton-Brock TH, Hendry SP, Fink S: Preliminary clinical evaluation of the Paratrend 7 intravascular blood gas monitoring system. Intensive Care Med 18:S154, 1992 17. Venkatesh B, Clutton-Brock TH, Hendry SP: Continuous measurement of blood gases using a combined electrochemical and spectrophotometric sensor. J Med Eng Technol 18:165-168, 1994 18. Venkatesh B, Clutton-Brock TH, Hendry SP: A multiparameter sensor for continuous intra-arterial blood gas monitoring: a prospective evaluation. Crit Care Med 22:588-594, 1994 19. Abraham E, Gallagher TJ, Fink S: Clinical evaluation of a multiparameter intra-arterial blood-gas sensor. Intensive Care Med 22:507-513, 1996 20. Liem MS, Kallewaard JW, de Smet AM, et al: Does hypercarbia develop faster during laparoscopic herniorrhaphy than during laparoscopic cholecystectomy? Assessment with continuous blood gas monitoring. Anesth Analg 81:1243-1249, 1995 21. Venkatesh B, Clutton-Brock TH, Hendry SP: Evaluation of the Paratrend 7 intravascular blood gas monitor during cardiac surgery: Comparison with the C4000 in-line blood gas monitor during cardiopulmonary bypass. J Cardiothorac Vasc Anesth 9:412419, 1995 22. Venkatesh B, Pigott DW, Fernandez A, et al: Continuous measurement of arterial blood gas status during total hip replacement: A prospective study. Anaesth Intensive Care 24:334-41, 1996 23. Myles PS, Story DA, Higgs MA, et al: Continuous measurement of arterial and end-tidal carbon dioxide during cardiac surgery: Pa-ETCO2 gradient. Anaesth Intensive Care 25:459-463, 1997 24. Endoh H, Honda T, Oohashi S, et al: Continuous intrajugular venous blood-gas monitoring with the Paratrend 7 during hypothermic cardiopulmonary bypass. Br J Anaesth 87:223-238, 2001 25. Ishikawa S, Makita K, Nakazawa K, et al: Continuous intraarterial blood gas monitoring during oesophagectomy. Can J Anaesth 45:273-276, 1998
C
ONTINUOUS INTRAVASCULAR BLOOD gas monitoring (CIBM) is applied by clinicians in selected patient groups. In the operating room, CIBM is used during surgery when blood gas values change rapidly and unexpectedly (eg, during thoracic surgery with 1-lung ventilation [OLV], cardiovascular surgery, and organ transplantation). Furthermore, CIBM is used in critically ill patients needing a considerable number of blood gas determinations over a long period of time (eg, premature infants with severe cardiopulmonary disease and patients with acute respiratory distress syndrome, sepsis, or severe trauma). The only commercially available CIBM system for adults, Paratrend 7 (Diametrics Medical Inc, High Wycombe, UK), was modified by the manufacturer to the Paratrend 7⫹ (PT7⫹) in 1999. PT7⫹ is a multiparameter CIBM system incorporating 4 different sensors. To measure PO2, the previously used Clark electrode was replaced with an optical PO2 sensor (fluorescent quenching sensor, ruthenium dye in silicone matrix). Accordingly, the length of the sensor tip has been halved. The other sensors and the main characteristics of the probe remained unchanged including optodes to determine PCO2 and pH (absorbance sensors, phenol red in bicarbonate solution) and a thermocouple (copper, constantan) to measure temperature and allow temperature correction of blood gas values. The sensors are housed in a heparin-coated microporous polyethylene tube that is permeable to the analytes to be measured. The probe is 0.5 mm in diameter, suitable for insertion through a 20-G or larger size catheter. Further details of the probe have been described elsewhere.1,2 Supported by the local distributor of the manufacturer, clinical experiences from PT7⫹ with excellent results were recently published in neurosurgical patients in the operating room (OR) and the intensive care unit (ICU).3,4 However, no sponsor-independent study on this modified CIBM system, which is being used clinically in most countries, has been published so far. The aim of the study was to evaluate the accuracy and performance of the modified CIBM probe PT7⫹ over a wide range of blood gas and pH values in thoracoscopic surgery during OLV.
analysis revealed a bias ⴞ 2 standard deviation of ⴚ20 ⴞ 86 mmHg for PO2, 3 ⴞ 9 mmHg for PCO2, and ⴚ0.01 ⴞ 0.06 for pH. No specific complications attributable to the probe were observed. Conclusion: In patients undergoing thoracoscopic surgery with rapidly changing blood gas parameters, the PT7ⴙ device is a valuable trend indicator and hence may be helpful for clinical decision making. However, the underestimation of PO2 values by 20 mmHg on average and the wide limits of agreement documented in this study must be regarded as limiting factors. © 2004 Elsevier Inc. All rights reserved. KEY WORDS: anesthesia, blood, gas analysis, equipment, monitors, intensive care, monitoring, continuous intravascular blood gas, multiparameter intravascular sensor, electrodes, optodes METHODS With local ethics committee approval and written informed consent, 23 consecutive patients undergoing elective thoracoscopic surgery were enrolled in the study. The study population constitutes a representative segment of patients presenting for thoracoscopic surgery at the University Hospital Zurich. On the day of surgery, after arriving in the OR, a radial artery was cannulated (left/right ⫽ 12/11) using a 20-G catheter (Insyte, Becton Dickinson Vascular Access Inc., Sandy, UT). Then, a previously calibrated PT7⫹ probe was introduced through this cannula according to the manufacturer’s instructions. No in vivo “recalibration” (adjusting the original calibration curve using in vitro laboratory blood gas determinations) was done. Anesthesia and monitoring during surgery (Hellige monitor; VICOM-SM SMU 612; PPG Hellige, Freiburg, Germany) were performed according to institutional standards, including monitoring of invasive arterial pressure and central venous pressure. General anesthesia was induced intravenously using propofol (n ⫽ 14) or etomidate (n ⫽ 9) combined with fentanyl. Anesthesia was maintained with intravenous propofol (n ⫽ 13) or isoflurane (n ⫽ 10) combined with intermittent doses of fentanyl or a continuous remifentanil infusion depending on clinical needs. Atracurium (n ⫽ 17) or rocuronium (n ⫽ 6) was used for neuromuscular blockade. For postoperative pain management, thoracic epidural anesthesia was instituted in 6 patients using ropivacaine. PT7⫹ values were not used to guide patients’ therapy. Left-sided double-lumen endobronchial tubes (BRONCHOPART; Ruesch GmbH, Kernen, Germany, placed under fiberoptic control) were used for
From the *Institute of Anesthesiology, University Hospital Zurich, Zurich, Switzerland; and †Institute of Anesthesiology and Intensive Care Medicine, Triemli City Hospital Zurich, Zurich, Switzerland. Presented in part at the annual meeting of the American Society of Anesthesiologists, San Francisco, CA, October 11-15, 2003. Address reprint requests to Marco P. Zalunardo, MD, Institute of Anesthesiology, University Hospital Zurich, Raemistrasse 100, Ch8091 Zurich, Switzerland. E-mail: [email protected] © 2004 Elsevier Inc. All rights reserved. 1053-0770/04/1805-0008$30.00/0 doi:10.1053/j.jvca.2004.07.017
Journal of Cardiothoracic and Vascular Anesthesia, Vol 18, No 5 (October), 2004: pp 587-591
587
588
GANTER ET AL
Table 1. Demographic and Surgical Characteristics Age (y) Female/male ratio (n) ASA status; I/II/III (n) Duration of anesthesia (min) Duration of one-lung ventilation (min) Surgical procedures (n) Lung volume reduction surgery Lung resection Pleurectomy Sympathectomy Thymectomy
55 ⫾ 15 (19-74) 10/13 5/9/9 168 ⫾ 48 (110-285) 51 ⫾ 29 (22-137) 8 8 4 2 1
NOTE. Data are presented as mean ⫾ SD (min/max).
volume-controlled mechanical ventilation during 2- and 1-lung ventilation (Siemens Servo 900D ventilator; Siemens Life Support Systems, Sona, Sweden). Initial (T1) data were obtained after induction of anesthesia under 2-lung ventilation. After positioning of the patients according to the surgical requirements and before the introduction of the trocars through the intercostal spaces, OLV was started and maintained during the thoracoscopic procedure. Five minutes (T2), 15 minutes (T3), and 25 minutes (T4) after the onset of OLV, further data were acquired. Last measurements (T5) were performed 10 minutes after 2-lung ventilation was re-established. Provided that system stability was indicated by the PT7⫹ computer, parameters of the PT7⫹ were stored at the measurement time points. Simultaneously, a 1-mL arterial blood sample was obtained for in vitro blood gas analysis from the radial artery cannula using a preheparinized syringe (QS 50, Radiometer Medicals, Copenhagen, Denmark). In vitro ABGA measurement was performed in a point-of-care laboratory blood gas analyzer (Ciba-Corning series 800, Chiron Diagnostics, Emeryville, CA) within 3 minutes by a laboratory technician who was unaware of the in vivo PT7⫹ values. According to the operator’s manual, daily quality control procedures were done to guarantee recommended accuracy and performance of the laboratory blood gas analyzer.5,6 PT7⫹ and ABGA parameters were both determined at 37°C. Statistical analysis was performed using StatView for Windows (Version 5.0.1, SAS Institute Inc, Cary, NC). Results are presented as mean ⫾ standard deviation (SD). According to Bland and Altman, bias (mean difference between PT7⫹ and ABGA parameters) ⫾ 2 SD (limits of agreement) with their 95% confidence intervals were calculated.7 The repeated measures design was accommodated over time by analysis of variance for repeated measures with post-hoc Bonferroni/ Dunn test. A p value ⬍0.05 was considered to be statistically significant. Hydrogen ion concentration (H⫹) was calculated from the pH values measured by both systems to allow calculating correlation coefficients with their 95% confidence intervals. RESULTS
Twenty-three patients were enrolled and analyzed. Patient demographic and surgical characteristics are presented in Table 1. A total of 113 PT7⫹ readings with corresponding ABGA values were obtained. Two sets of data were incomplete; as operation time was too short in 1 patient, and malfunctioning of the probe occurred in another patient after 100 minutes. No further specific complications attributable to the probe (eg, dampening of the arterial pressure catheter, clot formation, ischemic events) were observed. Blood gas and pH values varied extensively during the measurement period, showing a range of 50 to 474 mmHg for PO2, 29 to 58 mmHg for PCO2, and 7.28 to 7.49 for pH.
Bland and Altman analysis (Fig 1) revealed a bias ⫾ 2 SD of ⫺20 ⫾ 86 mmHg for PO2, meaning that PT7⫹ showed on average 20 mmHg lower PO2 values than those received by ABGA. In the clinically most relevant lower range of PO2 ⬍100 mmHg, bias ⫾ 2 SD was ⫺14.7 ⫾ 25.4 mmHg (PO2 mean, min/max ⫽ 83.0, 49.5/99.8 mmHg, n ⫽ 17). Regarding the performance of PCO2 and pH measurement, bias ⫾ 2 SD was 3 ⫾ 9 mmHg and ⫺0.01 ⫾ 0.06, respectively. Correlation between PT7⫹ and ABGA showed r2 values of 0.84 for PO2, 0.61 for PCO2, and 0.57 for pH (Table 2). The performance of measured values over time remained unchanged; bias ⫾ 2 SD from each measurement time point compared with the others showed no clinically or statistically relevant difference. Furthermore, influence of different hemoglobin levels (mean ⫾ 2 SD ⫽ 12.5 ⫾ 2.0 g/dL) and patients’ rectal temperatures (mean ⫾ 2 SD ⫽ 35.7° ⫾ 0.4°C) were studied regarding accuracy of CIBM measurements. There was no correlation between the bias ⫾ 2 SD of CIBM measurements and these variables. DISCUSSION
Currently, there is only 1 CIBM system commercially available for adults. To measure intravascular PO2 more precisely, the manufacturer modified the former hybrid probe PT7 a few years ago, replacing the Clark electrode by an optode. The main characteristics of the new pure optode-based probe, the PT7⫹, remained unchanged. Theoretical advantages of the modified design are that the PO2 optode is not consumed, resulting in a high stability over time and minimal drift. Furthermore, PO2 optodes enable further miniaturization of the probe (ie, the current length of the sensor tip could be reduced from 4 to 2 cm). By minimizing the intravascular part of the sensor, blood flow is better preserved; dampening of the arterial pressure waveform as well as clot formation may occur less frequently, and the sensor may be less susceptible to bending and kinking. The present study evaluated the modified CIBM probe PT7⫹ during thoracoscopic surgery. The PT7⫹ underestimated arterial PO2 by 20 mmHg, on average. However, limits of agreement were found to be poor, ranging from ⫺106 to 66 mmHg (Table 2, Fig 1). Malfunctioning of PT7⫹ occurred in 1 patient after 100 minutes, displaying no reliable blood gas values anymore. No further specific complications attributable to the PT7⫹ probe (eg, dampening of the arterial pressure tracing, clot formation, ischemic events) were observed in the studied period (mean anesthesia time 168 minutes, Table 1). Looking at the clinically important range of PO2 below 100 mmHg, much better results were obtained, but there was still a bias of ⫺15 mmHg and limits of agreement of ⫺40 to 10 mmHg. The promising in vitro data presented in the technical specification sheet of the PT7⫹ were not fulfilled; using gas-tonometered solutions, 95% confidence limits were ⫾5% or ⫾3 mmHg (whichever is greater) of the actual values for PO2 between 20 to 120 mmHg.2 Regarding measurement of oxygen saturation in this lower PO2 range, pulse oximetry may theoretically yield more accurate data than that obtained with the PT7⫹ in the present study. Pulse oximeters are found to be accurate within ⫾4% (2 SD) between 70 and 100% and ⫾6% (2 SD) between 50% to 70% SpO2, respectively.8,9 However, most of these data were collected in highly controlled nonclinical situations with
BLOOD GAS MONITORING DEVICE
589
Fig 1. Bland and Altman analysis showing bias (mean of differences) ⴞ 2 SD (upper and lower limit of agreement) versus mean values of arterial PO2, PCO2, and pH obtained by Paratrend 7ⴙ (PARA) compared with values obtained by in vitro blood gas analysis using a co-oximeter (ABGA) (n ⴝ 113).
healthy volunteers, and noninvasive standard monitoring (pulse oximetry, capnometry) did not accurately reflect the profound fluctuations in blood gas and pH values during thoracoscopic surgery with OLV.10 Comparison of blood gas variables from continuous intravascular sensors with those from laboratory blood gas analyz-
ers is a controversial issue. The accuracy of laboratory blood gas analyzers can be quantified in the laboratory where the values to be measured are known. However, this is not the case in clinical studies. The clinical performance of an optode-based intravascular probe must be judged in comparison with an electrode-based blood gas analyzer, for which clinical perfor-
590
GANTER ET AL
Table 2. Comparison of Measured Parameters by Paratrend 7ⴙ and In Vitro Blood Gas Analysis
PO2 (mmHg) PCO2 (mmHg) pH [H⫹] (nmol/L)
Mean ⫾ SD
Range
r (95% CI)
181.8 ⫾ 92.0 40.1 ⫾ 6.2 7.40 ⫾ 0.05 40.5 ⫾ 4.5
49.5/474.0 29.3/57.8 7.28/7.49 32.4/52.5
0.92 (0.88/0.94) 0.78 (0.70/0.85) - (-/-) 0.76 (0.67/0.83)
Bias (95% CI)
2 SD (95% CI)
⫺20.4 (⫺27.1/⫺13.7) 2.6 (1.9/3.3) ⫺0.01 (⫺0.02/⫺0.01) 1.4 (0.9/1.9)
85.8 (74.2/97.4) 8.9 (7.7/10.1) 0.05 (0.05/0.07) 6.5 (5.6/7.3)
NOTE. Values for mean ⫾ SD and range are those determined by in vitro arterial blood gas analysis. [H⫹] was calculated from the pH to allow correlation analysis (r ⫽ correlation coefficient; n ⫽ 113).
comparable to the given recommendations. There was no drift in accuracy of measured parameters over time. In contrast, the only 2 published clinical trials on PT7⫹ in neurosurgical patients showed excellent performance, even for PO2 measurements (Table 3).3,4 Several issues have to be considered comparing results of different studies. Both prior studies were performed by the same group, and the second paper was supported by the local distributor. The 2 studies “recalibrated” their probes every 12 hours as recommended, but nothing was stated about other “recalibrations.” By doing an in vivo “recalibration” immediately after insertion of the probe, an eventual offset of the probe could be lessened, hence improving the bias. However, repeatability (limits of agreement) would not improve. The present authors used a previously ex vivo calibrated PT7⫹ probe and did no in vivo “recalibration” (adjusting the original calibration curve using in vitro laboratory blood gas determinations). Blood gas variations in patients undergoing neurosurgery or staying in the neurosurgical ICU are much smaller compared with patients undergoing thoracoscopic surgery. Reaction time might not be sufficient in clinical situations in which blood gas values change rapidly and unex-
mance cannot be specifically quantified. Laboratory blood gas analyzers (even between individual analyzers of the same manufacturer) also have inconsistencies. Therefore, even with an accurate laboratory blood gas analyzer, the measured blood gas values may differ somewhat from the real blood gas values in vivo. Additionally, PT7⫹ measures blood gas and pH values at patient’s body core (intravascular) temperature and values are adjusted for 37°C, as opposed to laboratory blood gas analyzers measuring at 37°C. Because intermittently drawn blood samples analyzed by laboratory blood gas analyzers are the clinical standard of care, this procedure was used as a reference method to assess the accuracy of CIBM.2 No official recommendations concerning performance of continuous intravascular blood gas devices exist. However, several guidelines for laboratory blood gas analyzers have been published by the Clinical Laboratory Improvement Amendment/Health Care Finance Administration,11 the College of American Pathologists,6 and the Emergency Care Research Institute.2,5 If they are applied to the authors’ data, PO2 measurements failed to meet these recommendations, even with the better performance in the lower PO2 range. However, values for PCO2 and pH were acceptable and
Table 3. Evaluation Studies of the Paratrend 7 and Its Current Modification, the Paratrend 7ⴙ on Adult Patients PCO2 (mmHg)
PO2 (mmHg)
Studies with Paratrend 7 (hybrid probe) Clutton-Brock et al16 Venkatesh et al17 Venkatesh et al18 Abraham et al19 Nunomiya et al13 Liem et al20 Venkatesh et al21
Venkatesh et al22 Myles et al23 Endoh et al24 Zollinger et al1 Ishikawa et al25 Zaugg et al10 Myles et al12 Studies with Paratrend 7⫹ (pure optode probe) Menzel et al2 Menzel et al3 Present study
pH
Setting
Range
Bias
2 SD
Range
Bias
2 SD
ICU ICU ICU ICU ICU OR: LAP OR: CPB OR: CPB OR: CPB OR: ORTHO OR: CPB OR: CVS OR: OLV OR: OLV OR: OLV OR: OLV OR: OLV
-/93/202 60/447 -/45/542 -/130/464 137/510 61/300 -/-/26/120 46/458 47/449 72/255 37/625 -/-
2 6 3 ⫺2 ⫺7 3 4 1 ⫺9 ⫺1 3 ⫺1 0 2 ⫺22
55 41 51 40 68 90 54 39 86 9 73 80 42 83 108
-/25/51 26/52 -/29/68 -/24/41 21/41 21/50 -/-/28/72 31/71 25/57 30/47 27/56 -/-
1 2 2 1 1 ⫺2 2 1 2 1 ⫺1 0 2 1 1 0 ⫺2
ICU: NS OR: NS OR: OLV
53/486 75/375 50/474
⫺2 2 ⫺20
27 31 86
22/60 -/29/58
0 2 3
Range
Bias
2 SD
11 25 10 5 4 4 4 4 8 4 6 7 6 6 6 4 12
-/⫺0.01 7.21/7.53 0.01 7.31/7.61 0.01 -/0.01 7.19/7.60 0.00 -/0.00 7.36/7.57 0.02 7.28/7.53 0.01 7.10/7.55 0.02 -/0.02 -/0.02 7.12/7.59 0.01 7.19/7.50 ⫺0.02 7.30/7.49 0.00 7.34/7.47 ⫺0.01 7.24/7.51 0.01 -/0.01
0.12 0.14 0.12 0.05 0.04 0.05 0.10 0.12 0.12 0.06 0.07 0.07 0.07 0.04 0.08 0.04 0.10
7 4 9
7.19/7.85 0.00 -/⫺0.02 7.28/7.49 ⫺0.01
0.08 0.04 0.06
Abbreviations: ICU, intensive care unit; OR, operating room; LAP, laparoscopic surgery; CPB, cardiopulmonary bypass; ORTHO, orthopedic surgery; CVS, cardiovascular surgery; OLV, 1-lung ventilation; NS, neurosurgery.
BLOOD GAS MONITORING DEVICE
591
pectedly despite indicated system stability by the PT7⫹ monitor. Furthermore, the site of insertion was different in the studies and may influence the performance of the CIBM system. The most common site for CIBM measurement is the radial artery in adults (present study), but it has been shown that insertion of the probe into a femoral artery yields more reliable results compared with those from the radial artery, especially during low-flow states.2 The setting of the present study was similar to previous studies in thoracoscopic patients from the authors’ group (Table 3).1,10 The only modification in the study setup was an exchange of the laboratory blood gas analyzer, which hardly explains the large bias between CIBM and ABGA in PO2 measurement. A similar offset in PO2 between the 2 measurement techniques was also found in 1 study during lung transplantation.12 With the former PT7 device, a similar tendency of better performance for PO2 ⬍100 mmHg was observed in thoracoscopic (bias ⫺4 mmHg, limits of agreement ⫺21 to 13 mmHg)10 and critically ill patients (bias ⫺2 mmHg, limits of agreement ⫺15 to 11 mmHg).13 Interestingly, in the study on
neurosurgical ICU patients with the PT7⫹, the opposite was found; performance of PO2 measurement increasingly worsened below 100 mmHg PO2. The underestimation of PO2 values by the PT7⫹ may be explained partly by the “wall effect,” observed in early CIBM devices using photochemical sensors for PO2 measurement.14,15 If the sensor becomes attached to the wall of the vessel, thus measuring a combined blood and tissue PO2, the displayed oxygen value may be falsely low compared with the true arterial PO2. Hybrid probes (eg, PT7) were less susceptible to this artifact because the PO2 electrode had a larger surface area.2 Clinical indications for CIBM, in terms of evidence-based medicine, remain unclear because the cost-benefit ratio and the impact on patient outcome are unknown. In patients with rapidly changing blood gas parameters, the PT7⫹ device is a valuable trend indicator and hence may be helpful for clinical decision making. However, the underestimation of PO2 values and the wide limits of agreement documented in this study must be regarded as limiting factors.
REFERENCES 1. Zollinger A, Spahn DR, Singer T, et al: Accuracy and clinical performance of a continuous intra-arterial blood-gas monitoring system during thoracoscopic surgery. Br J Anaesth 79:47-52, 1997 2. Ganter M, Zollinger A: Continuous intravascular blood gas monitoring: development, current techniques, and clinical use of a commercial device. Br J Anaesth 91:397-407, 2003 3. Menzel M, Henze D, Soukup J, et al: Experiences with continuous intra-arterial blood gas monitoring. Minerva Anestesiol 67:325331, 2001 4. Menzel M, Soukup J, Henze D, et al: Experiences with continuous intra-arterial blood gas monitoring: Precision and drift of a pure optode system. Intensive Care Med 29:2180-2186, 2003 5. ECRI (Emergency Care Research Institute): Blood gas/pH analyzers. Health Devices 24:208-243, 1995 6. Shapiro BA: Evaluation of blood gas monitors: performance criteria, clinical impact, and cost/benefit. Crit Care Med 22:546-548, 1994 7. Bland JM, Altman DG: Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1:307-310, 1986 8. Jensen LA, Onyskiw JE, Prasad NG: Meta-analysis of arterial oxygen saturation monitoring by pulse oximetry in adults. Heart Lung 27:387-408, 1998 9. Soubani AO: Noninvasive monitoring of oxygen and carbon dioxide. Am J Emerg Med 19:141-146, 2001 10. Zaugg M, Lucchinetti E, Zalunardo MP, et al: Substantial changes in arterial blood gases during thoracoscopic surgery can be missed by conventional intermittent laboratory blood gas analyses. Anesth Analg 87:647-653, 1998 11. Medicare MaCp: Regulations implementing the Clinical Laboratory Improvement Amendments of 1988 (CLIA)—HCFA. Final rule with comment period. Fed Regist 57:7002-7186, 1992 12. Myles PS, Buckland MR, Weeks AM, et al: Continuous arterial blood gas monitoring during bilateral sequential lung transplantation. J Cardiothorac Vasc Anesth 13:253-257, 1999 13. Nunomiya S, Toshihide T, Masaru T, et al: Clinical evaluation of a new continuous intraarterial blood gas monitoring system in the intensive care setting. J Anesth 10:163-169, 1996
14. Mahutte CK: On-line arterial blood gas analysis with optodes: Current status. Clin Biochem 31:119-130, 1998 15. Wahr JA, Tremper KK: Continuous intravascular blood gas monitoring. J Cardiothorac Vasc Anesth 8:342-353, 1994 16. Clutton-Brock TH, Hendry SP, Fink S: Preliminary clinical evaluation of the Paratrend 7 intravascular blood gas monitoring system. Intensive Care Med 18:S154, 1992 17. Venkatesh B, Clutton-Brock TH, Hendry SP: Continuous measurement of blood gases using a combined electrochemical and spectrophotometric sensor. J Med Eng Technol 18:165-168, 1994 18. Venkatesh B, Clutton-Brock TH, Hendry SP: A multiparameter sensor for continuous intra-arterial blood gas monitoring: a prospective evaluation. Crit Care Med 22:588-594, 1994 19. Abraham E, Gallagher TJ, Fink S: Clinical evaluation of a multiparameter intra-arterial blood-gas sensor. Intensive Care Med 22:507-513, 1996 20. Liem MS, Kallewaard JW, de Smet AM, et al: Does hypercarbia develop faster during laparoscopic herniorrhaphy than during laparoscopic cholecystectomy? Assessment with continuous blood gas monitoring. Anesth Analg 81:1243-1249, 1995 21. Venkatesh B, Clutton-Brock TH, Hendry SP: Evaluation of the Paratrend 7 intravascular blood gas monitor during cardiac surgery: Comparison with the C4000 in-line blood gas monitor during cardiopulmonary bypass. J Cardiothorac Vasc Anesth 9:412419, 1995 22. Venkatesh B, Pigott DW, Fernandez A, et al: Continuous measurement of arterial blood gas status during total hip replacement: A prospective study. Anaesth Intensive Care 24:334-41, 1996 23. Myles PS, Story DA, Higgs MA, et al: Continuous measurement of arterial and end-tidal carbon dioxide during cardiac surgery: Pa-ETCO2 gradient. Anaesth Intensive Care 25:459-463, 1997 24. Endoh H, Honda T, Oohashi S, et al: Continuous intrajugular venous blood-gas monitoring with the Paratrend 7 during hypothermic cardiopulmonary bypass. Br J Anaesth 87:223-238, 2001 25. Ishikawa S, Makita K, Nakazawa K, et al: Continuous intraarterial blood gas monitoring during oesophagectomy. Can J Anaesth 45:273-276, 1998