Mixed Venous Oximetry During Automatic Implantable Cardioverter-Defibrillator Placement Daniel F. Van Riper, MD, Jan Charles Horrow, Daniel McCormick,
DO, and Scott M. Goldman,
Mixed venous oxygen saturation (SVO,) monitoring was used to assess tissue and circulatory recovery following induced ventricular tachycardia or fibrillation in 17 patients undergoing surgery for automatic implantable cardioverter-defibrillator (AICD) placement. Return of systemic arterial pressure conventionally determines adequate recovery. The duration of circulatory arrest during defibrillator threshold (DFT) testing, measured from the moment of absent phasic pressure at the radial artery until its return, was 18 2 8 seconds (mean + SD, n = 118 episodes). The absolute decrease in SVO, from baseline to nadir for these 118 episodes was 14% f 8% absolute. and correlated well with the duretion of circulatory arrest (r = 0.757, P = O.DWl). The time from onset of phasic arterial blood pressure to the nadir of SGO,,
available for 46 episodes, was 28 t 14 seconds, and did not correlate with the duration of arrest. The time from onset of phasic pressure to the return of SVO, to within 1% (absolute) of baseline saturation, available for 84 episodes, was 52 + 32 seconds and, in the aggregate, correlated poorly (r = 0.401) with duration of arrest. Simultaneous recording of arterial pressure end SVO, (n = 41) showed that arterial recovery (8 + 3 seconds) occurred long before SVO, recovery (48 k 18 seconds, P = O.ooOl). The authors interpret these data as showing that mixed venous oximetry, compared to arterial blood pressure, provides a more sensitive indicator of tissue recovery following periods of circulatory arrest during DFT testing of AICDs. 0 1990 by W.B. Saunders Company.
HE TREATMENT of patients, who suffer from refractory malignant dysrhythmias, with the automatic implantable cardioverterdefibrillator (AICD) has progressed well beyond clinical trials to become an important tool for treating sustained ventricular dysrhythmias.’ The patient undergoing AICD implantation requires repeated induction of ventricular tachycardia (VT) or ventricular fibrillation (VF) while under general anesthesia. A majority of these patients have ischemic heart disease with poor left ventricular function, and the induced VT or VF places them at additional risk for acute ischemia and infarction.2*3 Little has been written regarding the physiological effects of these short periods of induced circulatory arrest.2 Anesthetic and surgical strategies during AICD placement may benefit from SVO, data, which gauge metabolic recovery, and may provide an additional margin of safety over hemodynamic monitors alone. This study explored the relationship among the duration of circulatory arrest and the magnitude of SVO, decrease, the time to SVO, nadir, and the time to SVO, recovery in an effort to optimize the interval between repeated DFT testing episodes.
sternotomy and preceded extracorporeal circulation. A left anterolateral thoracotomy was used in the 11 patients having AICD implantation without coronary artery surgery. All patients received anterior and posterior patch leads. Standard protocols governed defibrillation threshold settings.* Before the induction of anesthesia, each patient had a 20-gauge radial artery catheter inserted, and a pulmonary arterial (PA) catheter with oximetry capability (Oximetrix, Abbott Critical Care Systems, Mountain View, CA) placed through the right internal jugular vein. The necessary electrophysiologic monitors were also applied. Intravenous fentanyl, 20 pg/kg, in the thoracotomy patients, and 50 to 100 pg/kg in the sternotomy patients, or equipotent doses of sufentanil, supplemented when necessary by a volatile anesthetic in oxygen, provided anesthesia. Intravenous (IV) vecuronium or pancuronium, 0.1 mg/kg, produced muscle relaxation. Normothermia was maintained by using an airway heat and moisture exchanger, warmed IV solutions, low fresh gas flows, a warming blanket, and ambient operating room temperatures above 22’C. Bladder, esophageal, and pulmonary arterial temperatures were monitored. During the study period, PA diastolic pressure guided administration of
With institutional approval, 17 consecutive patients scheduled for AICD lead implantation and testing were studied. Six of these 17 patients received concomitant aortocoronary grafts. In these patients, AICD testing followed
*Physician’s manual for the CPI automatic implantable cardioverter defibrillator. Document No. 1650155, Cardiac Pacemakers, Inc, St Paul, MN 55 164. From the Departments of Anesthesiology and Cardiothoracic Surgery, and the Division of Cardiology, Department of Medicine, Hahnemann University School of Medicine, Philadelphia, PA. Address reprints requests to Jan C. Harrow, MD. Department of Anesthesiology, Hahnemann University School of Medicine, Broad and Vine Sts, Philadelphia, PA 19102-1192. 0 1990 by W.B. Saunders Company. O&88-6290/90/0404-0006$03.00/0
Journalof Cardiothoracic Anesthesia, Vol4, No 4 (August), 1990: pp 453-457
VAN RIPER ET AL
small volumes (about 200 mL) of Plasmalyte-A (Travenol, Deerfield, IL), in an effort to maintain constant preload. Just before the initiation of the series of defibrillation threshold (DFT) testing, additional vecuronium, 0.1 mg/kg, IV, was administered to assure complete paralysis (Fig 1). The duration of arrest, which was measured from the absence of phasic pressure at the radial artery until its return, was measured for each episode of DFT testing. The SVO, display (Oximetrix 3, Abbott Critical Care Systems) was continuously observed beginning with the onset of arrest and continuing until return of SvOt to within 1% (absolute) of the prearrest value. The SVO, display was updated every 2 seconds, and the values displayed represented a running average of the last three values sampled. Excessive vessel wall reflectance artifact or deposits over the optics possibly provide false SSO, data. Continuous monitoring and graphical display of the intensity of the reflectance signal prevented any such misinterpretation. In 13 patients, data acquisition occurred by continuous observation of the SirOr display screen and stopwatch or by continuous recording of intravascular pressures and the running average S702 values via chart recorder (Dash IV, Astro-Med, Inc, Providence, RI) in 4 patients. Recorded information provided the following measured variables: (1) absolute decrease in SiiO, from baseline to nadir, in percent saturation; (2) time from onset of phasic arterial blood pressure following circulatory arrest to nadir of SFO, (“nadir time”), in seconds; (3) time from onset of phasic pressure to the return of SuO, to within 1% (absolute) of baseline saturation (“recovery time”), in seconds; (4) time from onset of phasic pressure to the return of systolic arterial pressure to within 95% of its baseline; and (5) absolute decrease in SiiOt from baseline at the moment when arterial pressure recovered to 95% of its baseline. Subsequent episodes of DFT testing were initiated no sooner than when the SVO, had returned to within 1% of the prearrest SuO, value. Correlation coefficients determined the degree of closeness of duration of circulatory arrest to the magnitude of decrease in SijO,, the time to SVO, nadir, and the SiiO, recovery time. Statistical significance required P -c 0.05.
provided 126 episodes of arrest (range, 3 to 18 epi-
sodes per patient). Eight of these 126 episodes were excluded from analysis due to factors such as spontaneous reversion of VT or a prolonged induction phase (>5 seconds) of VT or VF. In these instances, arrest duration was too short or baseline hemodynamics were affected by prolonged attempts to trigger the dysrhythmia. Arrest duration was 18 f 8 seconds (mean f SD) with a range of 4 to 43 seconds. Recovery time, recorded for 12 of the 17 patients, was 52 f 32 seconds (n = 84; range, 16 to 154 seconds). Nadir time, recorded for 5 of the 17 patients, was 28 r 14 seconds (n = 46; range, 10 to 77 seconds). Simultaneous arterial pressure and SVO, data were available for 4 of the 17 patients (41 episodes). No patient experienced a dysrhythmia or complete heart block during placement of the PA catheter. The absolute decrease in SvO, from baseline to nadir was 14 + 6% (range, 4% to 28%). The SSO, decrease correlated well with the duration of circulatory arrest (n = 118, r = 0.757, P = 0.0001; Fig 2). In contrast, nadir time did not correlate with the duration of arrest (n = 46, r = 0.026, P = NS). In the aggregate, recovery time correlated poorly with the duration of circulatory arrest (n = 84, r = 0.401). TO explore the possibility that a patient-specific factor confounded any possible correlation between recovery time and arrest duration, these data were analyzed separately for each patient. Of the 12 patients for whom recovery data were available, 7 had 5 or more data pairs. Individual correlations of recovery time with circulatory arrest duration for these 7 patients appear in Table 1. Correlations attain significance in only 4 of these 7 patients. Data for patient M appear in Fig 3.
Episodes of induced arrest
fl Non-depolarizing relaxant anesthetic PRN
catheter B Oximetry PA Catheter
w Normothermia w Preload
Fig 1. The experimental protocol. The investigation commences at the left side of the figure and progresses to the right. Many episodes of circulatory arrest occur, as indicated by the upright shaded rectangles.
MIXED VENOUS OXIMETRY DURING AICD PLACEMENT
Fig 2. Scattergram of the decrease in SgO, in absolute percent saturation versus duration of circulatory arrest in seconds; n = 118 data pairs, r = 0.757. P = 0.0001. By least regression, squares AsGo, (%I = 0.55 (% per 8) - arrest duration (8) + 3.4%.
Arrest Time (seconds)
Figure 4 displays a representative episode in one patient. The simultaneous chart recordings of arterial pressure and SVO, data (n = 41) showed that arterial blood pressure recovered long before EGO,. Systolic arterial pressure recovered to 95% of its baseline value in 6 f 3 seconds (n = 41), whereas SVO, recovery time was 48 + 16 seconds (two-tailed paired t test, P = 0.0001). The displayed SVO,, internally computed using a running average formula, produced delays of up to 6 seconds and does not nearly account for this difference. Indeed, arterial pressure recovered before SVO, even reached its nadir. SVO, had only begun its descent and was 4% + 3% (absolute) below baseline level at 95% recovery of arterial pressure, compared with the nadir SVO, of 16% + 6% below baseline (P = 0.0001, two-tailed paired t test, n = 41).
What period of time should elapse following an episode of arrest for DFT testing before proceeding with the next episode? SVO, monitoring appears to provide information on when the
Table 1. Individual Correlations of Recovery Time With Arrest Time Patient
No. of Episodes Analyzed
Arrest Time (Seconds) Fig 3. Scattergram of recovery time versus duration of circulatory arrest (both in seconds) for patient M; n = 13 data pairs, r = 0.844, P = 0.0003. The least squares regression line is: Recovery time (8) = 2.3 . arrest duration (8) + 1 .O s.
VAN RIPER ET AL
Radial Arterial Pressure (mmW
Fig 4. Compressed tracing of SgO,, arterial pressura, and pulmonary arterial pressure during a typical episode of DFT testing. The upper panel graphs GO,, the central panel tran8duced radial artery pressure, and the lower panel pulmonary arterial pressure. Note that arterial pressure recovers to prearrest values before SgO, even reaches its nadir. Pulmonary arterial pressure remains unaffected by DFT testing.
patient has recovered from one episode and is able to withstand another. With other influences controlled and constant, changes in SVO, reflect changes in oxygen extraction or flow. The correlation observed in the current study between S ~0, decrease and the duration of circulatory arrest substantiates this expectation. Return of SVO, to prearrest levels indicates sufficient tissue recovery from the period of oxygen depletion, induced by circulatory standstill. The SSO, recovery time correlated poorly with the duration of circulatory arrest in the aggregate. The apparent variability in recovery times may differ from patient to patient due to hemoglobin concentration, cardiac output, metabolic rate, or the effects of preoperative medications such as P-blockers or calcium channel inhibitors. Preoperative ejection fraction did not predict correlation of recovery time with duration of circulatory arrest. Individual patient correlations of SVO, recovery time with circulatory arrest time were significant in only four of seven patients (Table 1). Data from the three remaining patients did not exhibit any consistent error,
for example, a progressive worsening of recovery rate with time. Two of these three patients presented only five data pairs for analysis. The remaining patient sustained the most episodes of arrest (n = 18, one excluded from analysis), and the longest arrest times (24 f 8 seconds; range, 12 to 43 seconds). Perhaps of greater significance, that patient differed from other study patients because he experienced dysrhythmias throughout the immediate recovery period following each episode of DFT, thus unpredictably altering cardiac output. These factors may explain the lack of correlation between recovery and arrest times. Nadir time and recovery time are both dependent on flow to wash out desaturated capillary blood. Therefore, patient factors varying flow during the recovery period may also explain the lack of correlation between nadir time and arrest duration. Unfortunately, cardiac output could not be measured during recovery from DFT testing episodes, and thus, it is not possible to determine whether recovery time correlated with flow. Gaba et al4 recommend standard intraarterial catheters with either central venous or pulmonary artery pressure monitoring in AICD patients. Those authors provide no guidelines to gauge recovery from successive DFT tests, although a return of arterial pressure is implied. Deutsch et al5 also recommend timing the DFT tests by the blood pressure response. Data in the current study show that the return of baseline arterial pressure precedes recovery of SVO,. Using the return of arterial pressure alone, in order to determine recovery of the heart from episodes of circulatory arrest, may lead to unreliable and variable testing thresholds. Repeated episodes of arrest without sufficient tissue recovery time may predispose patients to an accumulation of detrimental metabolic products, myocardial ischemia, and possibly low output cardiac failure.’ These physiological trespasses may be more deleterious in patients with ischemic heart disease. Coronary sinus oximetry, during atria1 pacing in patients with ischemia, demonstrates that decreases in oxygen saturation precede other indicators of myocardial ischemia.6 Furthermore, the recovery of coronary sinus pH changes is more prolonged (3 to 5 minutes) in patients with clinical evidence of myocardial ischemia when compared with a nonischemic patient
MIXED VENOUS OXIMETRY DURING AICD PLACEMENT
group.’ Ischemic heart disease constitutes the predominant indication for AICD implantation. Because the ischemic heart takes longer to recover metabolically from stressful events, a more conservative approach towards recovery of multiple DFT tests is desirable in these patients. By gauging metabolic recovery, SVO, monitoring provides an additional margin of safety over hemodynamic monitors alone. Mixed venous oximetry presents several benefits to the patient undergoing AICD implantation and testing. First, SVO, monitoring allows multiple DFT testing to proceed with maximal efficiency. Fixed time regimens produce unnecessary delay between episodes of DFT testing, while recovery gauged by arterial pressure may yield testing intervals that are too brief.
SVO,-guided recovery can minimize surgical and anesthesia duration with potential savings in cost and patient morbidity. Second, by allowing adequate myocardial recovery from circulatory arrest, SVO, monitoring may diminish the potential for variability in defibrillation threshold determinations. Third, SVO, measurement takes into account patient factors that influence tissue recovery. These may be missed by estimates of recovery time based on the duration of arrest alone. The data show that the duration of circulatory arrest does not predict SVO, recovery time. ACKNOWLEDGMENT The authors appreciate the suggestions of Pietro Colonna-Romano, MD, and the assistance of Dawn Osborne, BS.
REFERENCES 1. Olinger GN, Chapman PD, Troup PJ, et al: Stratified application of the automatic implantable cardioverter-defibrillator. J Thorac Cardiovasc Surg 96:141-149, 1988 2. Antunes ML, Spotnitz HM, Livelli FD, et al: Effect of electrophysiological testing on ejection fraction during cardioverter/defibrillator implantation. Ann Thorac Surg 45315-318, 1988 3. Reid RR, Mirowski M, Mower MM, et al: Clinical evaluation of the internal automatic cardioverter-defibrillator in survivors of sudden cardiac death. Am J Cardiol 51:16081613,1983 4. Gaba DM, Wyner J, Fish KJ: Anesthesia and the automatic implantable cardioverter-defibrillator. Anesthesiology 62:786-792, 1985
5. Deutsch N, Hantler CB, Morady F, Kirsh M: Perioperative management of the patient undergoing automatic internal cardioverter-defibrillator implantation. J Cardiothorac Anesth 4:236-244, 1990 6. Poole-Wilson PA, Canepa-Anson R: Continuous measurement of coronary sinus oxygen saturation as a method for detecting coronary artery disease, in Schweiss JF (ed): Continuous Measurement of Blood Oxygen Saturation in the High-Risk Patient, vol 1. San Diego, CA, Beach, 1983, pp 59-65 7. Cobbe SM, Poole-Wilson PA: Continuous coronary sinus and arterial pH monitoring during pacing-induced ischaemia in coronary artery disease. Br Heart J 47:369-374, 1982