| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Section of Cardiothoracic Anesthesia and the
Department of Cardiothoracic Surgery, St. Elisabeth Heart Center, University Hospital of Trondheim, Trondheim, Norway, and the
Unit for Applied Clinical Research, Faculty of Medicine, Norwegian University of Science and Technology, Trondheim, Norway
Address correspondence and reprint requests to Idar Kirkeby-Garstad, MD, Section of Cardiothoracic Anesthesia, St. Elisabeth Heart Centre, Trondheim University Hospital, Hans Nissens gt 3 N 7018 Trondheim, Norway. Address email to Idar.Kirkeby-Garstad{at}medisin.ntnu.no
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
IMPLICATIONS: During early mobilization after aortic valve replacement, a marked and consistent reduction in mixed venous oxygen saturation to 35% and mixed venous oxygen partial pressure to 3 kPa was observed.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The patients followed our standard regimen. Medication except acetyl salicylic acid was continued up to surgery. Premedication was morphine 10 mg and scopolamine 0.4 mg IM. Anesthesia was induced with diazepam 50 µg/kg, thiopental 1–3 mg/kg, fentanyl 5–6 µg/kg, and pancuronium 100 µg/kg. Isoflurane was given for maintenance of anesthesia with supplements of fentanyl up to a total of 10 µg/kg, and midazolam 1–5 mg. Cold crystalloid cardioplegia was given using a modified St. Thomas solution (4). Extracorporeal circulation (ECC) was performed with a standard membrane oxygenator, alpha-stat blood gas control, and a pump flow of 1.5–2.0 L/m-2 at a venous temperature of 32°C. Mean arterial blood pressure during ECC was maintained between 40 and 60 mm Hg. The patients were tracheally extubated in the intensive care unit (ICU) when they were cooperative and with stable hemodynamic status, insignificant bleeding, and arterial oxygen saturation more than 95% at FIO2 
0.5. Morphine (in doses of 1–3 mg IV) as needed with paracetamol 4 g daily supplementation was given for postoperative pain relief. Preoperative medication, including ß-blocking drugs, was resumed on day 1.
A standardized 15–20 min mobilization procedure was performed on the morning of postoperative days 1 and 2 (hereafter: day 1, day 2), conducted by one of the authors (IKG or OFMS) with assistance of ICU nurses and a physiotherapist. Hemodynamic and oxygen measurements were obtained at the end of the following sequential steps:
Under stable conditions, both SvO2 and cardiac index (CI) fluctuate by approximately 10%. As a measure of variability, the standard deviation is roughly estimated as one-quarter of this range (5), i.e., approximately 2.5%. It would be of clinical interest to detect a 10% change in SvO2 or CI, equivalent to a standardized difference of the order of 3 to 4%. A paired Student’s t-test (each patient was his own control) with a power of 0.9 and two-sided 
= 0.05 will require only 3 to 5 observations (5). Accommodating for some uncertainty and potentially improving precision, a total of 15 patients were included.
The oximetry catheter was calibrated in vitro before insertion using a standard procedure. Recalibration was performed in the morning of days 1 and 2 immediately before the mobilization sequence. The standard in vivo recalibration protocol was used with SvO2 and hemoglobin (Hgb) obtained from an ABL300 blood gas analyzer (Radiometer, Denmark) as reference. This one-point calibration is claimed by the manufacturer to yield a measurement accuracy of ±2% for a range of SvO2 from 30 to 100% (Baxter Healthcare Corporation, Irvine, CA).
The pressure transducers were zeroed before each mobilization, and repositioned before each measurement, aligning the zero point to the fifth intercostal space in the mid-axillary line. The VigilanceTM system with continuous CO (CCO) measurements was used in 12 patients, and the ExplorerTM system was used in 3 patients (Baxter Healthcare Corporation, Irvine, CA). Both systems use Baxter’s 2-Sat system displaying SvO2 continuously. If an acceptable Signal Quality Index (SQI) was obtained (SQI < 3), SvO2 was recorded directly, otherwise the catheter was repositioned.
The VigilanceTM CCO monitor measures changes in pulmonary artery temperature in response to variations in heat produced by a low-powered filament located in the right atrium and ventricle (6). CI was recorded from the system’s STAT mode updating every 30–60 s as the mean value of 3 stable measurements. When using the non-CCO catheters, the mean value of 3 stable readings from 10 mL saline injection at room temperature was recorded. During mobilization, the PAC pressure curve was continuously observed to ensure correct position of the catheter.
A printout of the measurements of CCO and SvO2 was obtained in each case. A standard three-lead electrocardiogram (ECG) was continuously observed for ST deviations or arrhythmia. At time points T1, T5, and T6 (Fig. 1) arterial and mixed venous blood was drawn and analyzed with the ABL300 blood gas analyzer. Patients demanding pacemaker were managed with ventricular pacing, and the pacemaker frequency was kept constant during mobilization.
|
|
|
using oxygen partial pressure (kPa) and saturation from the blood gas samples.
The results are presented as median, mean with 95% confidence interval (CI), mean ± 1 SD, or range, as appropriate. Analysis of variance for repeated measurements was used to assess changes during mobilization. Pairwise Student’s t-tests were done for post hoc analyses, with Bonferroni correction according to the number of comparisons done. These analyses were supplemented with the nonparametric Friedman’s test and Wilcoxon’s signed rank test, respectively. Linear regression with backward selection (P to remove at least 0.10) was performed to identify possible predictor variables for mean SvO2 at maximum exercise as the outcome variable.
A P value <0.05 was considered significant. SPSS version 11 (SPSS, Chicago, IL) was used.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
|
The median time from ICU admission to mobilization on day 1 was 20.8 h (range, 18.5–22.3 h) and 44.5 h (range, 42.5–46 h) on day 2. The median amount of morphine given IV was 0.15 mg/kg (range, 0.08–0.30 mg/kg) on day 1 and 0.13 mg/kg (range, 0.03–0.39 mg/kg) on day 2. At the time of mobilization all patients had good pain relief, were alert, generally content, and ready to participate. They also felt comfortable during the procedure; two were dizzy when standing up, but recovered without medication. Ischemic events were not detected, nor was any new arrhythmia. During the second postoperative night two PACs were removed for patient comfort. Between-days comparison is thus based on 13 patients.
A typical recording from the VigilanceTM monitor during mobilization is shown in Figure 1. A marked reduction in SvO2 was seen when the patient was sitting up, followed by progressive decrease in SvO2 as long as the patient was walking in place (T3 and T5). Relaxing in a chair at T4 led to a temporary increase in SvO2, and the SvO2 reached premobilization levels quickly after the patient returned to bed. This train of events produced a characteristic and consistent W pattern in every patient (Fig. 2).
|
SvO2) was calculated as the difference between mean SvO2 values at rest and during exercise. These measures are presented in Table 1.
|
The oxygen consumption during exercise increased by 64% ± 41% and 58% ± 33% on days 1 and 2, respectively. No change in mean arterial blood pressure, PAC wedge pressure, or oxygen delivery index was seen during mobilization or between day 1 and day 2. Hemodynamic and oxygen transport variables are given in Table 2. During mobilization arterial PCO2 was significantly reduced but still within normal limits, and no significant change occurred in pH or base excess. Ventilation was thus slightly increased, but there was no metabolic acidosis.
Patient number 4, who initially recovered well, developed multiple organ failure after 1 wk and died 4 wk after the operation. All other patients were alive after 1 mo, giving 30-day mortality comparable to the average 3.6% for AVR patients at our institution (unpublished data). No other case of major organ failure or prolonged hospital stay was noted.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The SvO2 results from mixing of desaturated blood from all organs. Different organs have different venous oxygen content. A change in the distribution of blood and magnitude of extraction may change SvO2. The SvO2 reflects the overall balance between oxygen delivery and consumption and varies in healthy subjects from approximately 70% at rest to <20% during maximal exercise (7).
In this group of patients with AVR, low resting SvO2 values and considerable variations between patients were found during mobilization (Fig. 2, Table 1). Three factors were found to influence the SvO2 during exercise, the Hgb level, the use of ß-blockers, and the operation type. Low Hgb and ß-blockers reduced oxygen delivery by reducing arterial oxygen content and CO, respectively. The effect of operation type may indicate a slower recovery in patients with combined procedures than for those subjected to AVR alone. Our results suggest that higher postoperative levels of Hgb as well as inotropic support may be possible counter-measures for low SvO2 values. Svedjeholm et al. (2) demonstrated that low SvO2 values at patient arrival in the ICU negatively affected outcome. The low SvO2 values may indicate tissue hypoxia (1). Although of interest, these results may not be relevant as regards the short time desaturation during mobilization. The clinical impression was that mobilization was well tolerated by our patients. Further, we found no association between the degree of desaturation and outcome. We therefore continue to mobilize patients early after cardiac valve surgery.
It has been questioned whether reflectance oximetry catheters are accurate and reliable at values less than 50% (8). In an earlier study (9), we analyzed 352 paired reflectance oximetry catheter and bench hemoximetry measurements of SvO2 during mobilization (the present patient sample is included). A good overall agreement between the methods was documented. Only a small systematic bias was identified; the reflectance oximetry catheter underestimated bench hemoximetry SvO2 by 1.5% for each 10% decrease in SvO2 less than 65%. Simultaneous observations of low PvO2 values confirm the desaturation.
It has further been suggested that the CCO system is less accurate during abrupt changes in CO because of delayed time response (7). However, Singh et al. (10) found good correlation between CCO measurements (both in the trend and STAT mode) and intermittent measurements of CO during acute hemodynamic changes in off-pump CABG patients. Lazor et al. (11) studied the response to increasing the pacemaker rates and found delayed response time for both CCO modes compared with SvO2; the STAT mode updated the values faster than the trend mode. In our study, any bias attributable to delayed response time of the CCO method seems unlikely as only small alterations in CO were observed during the entire mobilization sequence.
In CABG patients Horiuchi et al. (12) found that turning patients to the lateral position in the early postoperative period increased CO from 7 L/min to more than 9 L/min with only a small decrease in SvO2 of approximately 4%. Viale et al. (1) found that increased oxygen consumption—associated with shivering—during the first 2 hours after CABG was compensated for by an increase in CI from 3.1 to 4.6 L · min-1 · m-2 and a decrease in SvO2 from 65.5% to 52.5%. In contrast to our results, the increased oxygen consumption was compensated for mainly through an increase in CO. An increment in oxygen consumption may be compensated for differently depending on its size (13), but oxygen consumption during rest and stress in these studies was of the same magnitude as in our patients. The above-mentioned patients were, however, sedated, mechanically ventilated, and resting in bed, whereas we studied patients who were standing up and adjusting to the standing posture.
Compiled physiological data from several studies on normal subjects indicate that standing up reduces the central venous pressure (CVP). Stroke volume is also reduced; this is presumably attributable to a Starling mechanism. Despite an increase in heart rate, CO is reduced (14). Gentle contractions in the calf muscles restore venous return to the heart, normalizing CVP and CO (14). Because activity increases oxygen consumption, these changes presumably cause a decrease in SvO2, but quantitative data are lacking. Left ventricular hypertrophy increases the AVR patients’ sensitivity to preload reductions. This may have contributed to the lack of increase in CO and the decrease in SvO2 in our patients when standing up. Increased postoperative oxygen consumption (15) and reduced arterial oxygen content attributable to the low Hgb may further have reduced SvO2.
Anesthesia, cardioplegic arrest, and cardiac manipulations may have induced cardiac dysfunction. Previous studies on CABG patients resting in bed concluded that cardiac dysfunction after surgery is normalized within the first postoperative morning (16,17). The results from studies performed at rest may, however, not be applicable whereas we tested the heart under stress, which may be a better way to monitor the patients’ adaptation to the postoperative situation. Wranne et al. (18) suggested that right ventricular dysfunction might last up to 6 months after cardiac surgery. We found no improvement from day 1 to day 2, which may support a prolonged cardiac dysfunction after AVR surgery.
This study has established that early mobilization of AVR patients consistently induces a reduction in SvO2 to values clinically used as markers of insufficient circulation and adverse patient outcome (1,2). No cardiac compensation and no improvement from day 1 to day 2 were observed. The small number of patients, the lack of data from preoperative mobilization in the study group, and the lack of a control group of subjects with normal cardiac function limit the interpretation of underlying mechanisms and possible clinical implications. Judged from the oxygen consumption, early mobilization represents a moderate workload. Nevertheless, the surprisingly low SvO2 values may indicate that some of our patients were pushed almost to the maximum of their actual working capacity. The clinical implications are unclear and the relation to patient outcome awaits further investigation.
|
Acknowledgments |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  I. Kirkeby-Garstad, U. Wisloff, E. Skogvoll, T. Stolen, A.-E. Tjonna, R. Stenseth, and O. F. Sellevold The marked reduction in mixed venous oxygen saturation during early mobilization after cardiac surgery: the effect of posture or exercise? Anesth. Analg., June 1, 2006; 102(6): 1609 - 1616. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
