| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Departments of *Anesthesiology and
Medical Information, CHU Montpellier, France
Address correspondence and reprint requests to Olivier Raux, MD, Département d’Anesthésie Réanimation A, Hôpital Lapeyronie, 371, Avenue du Doyen Gaston Giraud, 34295 Montpellier Cedex 5, France. Address email to o-raux{at}chu-montpellier.fr
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
IMPLICATIONS: Epidural anesthesia may induce significant hemodynamic changes, well documented in adults. Using noninvasive hemodynamic monitoring in children, we reported an increase in cardiac output and a decrease in arterial blood pressure only when epinephrine was added to epidurally-injected local anesthetics.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Conversely, numerous invasive hemodynamic studies have described the modifications of CO related to EA in adults. Even when sympathetic cardiac innervation is not involved in the blocked area, CO may decrease as much as 20% and be corrected by IV fluid administration (7). Dilation of venous vessels by sympathetic blockade provokes blood pooling, which may explain a decrease in venous return and thereby in CO (7,8). Moreover, association of GA to EA can lead to a large decrease in CO, associated with bradycardia and substantial hypotension (9). The addition of adrenaline to local anesthetics (LA) worsens hypotension through the systemic absorption of adrenaline that leads to a vasodilator ß effect (10,11). Then, contrary to plain LA, CO is markedly enhanced by ß-adrenergic stimulation (10).
We conducted this study to investigate the hemodynamic alterations, particularly in CO, related to EA by the caudal or thoracolumbar approach, in anesthetized young children and infants. CO was minimally invasively monitored by transesophageal Doppler (TED). The role of adrenaline addition to LA was studied in separate groups, comparatively with plain LA.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Fasting time was 6 h for solids and formula and 2 h for clear liquids. All patients received 0.4 mg/kg of midazolam and 20 µg/kg of atropine rectally 30 min before anesthesia. Inhaled GA was induced with 50% nitrous oxide and 7% sevoflurane in oxygen. An IV cannula was inserted and fentanyl (1 to 2 µg/kg) + atracurium (0.5 mg/kg) were administered before insertion of laryngeal mask or tracheal tube, as appropriate according to intraoperative requirements. Anesthesia was maintained with 50% nitrous oxide, oxygen, and sevoflurane (1 MAC end-tidal concentration, according to age). Mechanical ventilation was adjusted to ensure end-tidal CO2 to 30–35 mm Hg. One percent dextrose in Ringer’s lactate solution was given for intraoperative fluid maintenance using standard maintenance guidelines (4 mL · kg-1 · h-1 for infants weighing <10 kg, and 40 mL/h plus an additional 2 mL · kg-1 · h-1 in those heavier than 10 kg). Rectal temperature was monitored and maintained in a normal range with a forced-air warmer.
EA was performed 20 min after GA induction, either by the caudal or thoracolumbar approach, according to the extension and duration of the sensory block and postoperative analgesia required for the planned surgery (caudal and thoracolumbar block groups, respectively). Patients were placed in the lateral position for the puncture. A 22-gauge short-beveled, styletted needle was used for single-shot caudal blocks. Epidural thoracolumbar blocks were performed with an 18- or 19-gauge Tuohy needle according to age. A catheter was then advanced 3–5 cm into the epidural space. Adequate placement of the catheter was verified with an aspiration test and a test dose of 0.1 mL/kg of 1% lidocaine with adrenaline 1:100,000. Patients were then placed in supine position. In both groups, after a negative aspiration test, patients were randomly allocated to receive 1 of the 2 mixtures for induction of EA: 0.8 mL/kg of 1% lidocaine + 0.25% bupivacaine (50/50) + 1 µg/mL of fentanyl (P) or 1 mL/kg of the same mixture with 5 µg/mL of adrenaline (A). Epidural injection was completed over 1 min. The 0.1 mL/kg test dose was deducted from the induction dose in the thoracolumbar block groups. Four groups were therefore constituted: caudal block with adrenaline added solution (CA) or with plain solution (CP) and epidural thoracolumbar block with adrenaline added solution (TLA) or with plain solution (TLP).
Heart rate (HR, bpm) was continuously recorded from the electrocardiograph, and mean arterial blood pressure (MAP, mm Hg) was measured by automated blood pressure cuff every 5 min. As described in adults (12), a 6-mm diameter TED probe (ODM 2; Abbott, Maidenhead, UK), inserted through the mouth after induction of GA, allowed continuous monitoring of CO, which was recorded every 5 min. The probe was connected to a monitor displaying the blood flow velocity wave form after spectral analysis of the reflected Doppler-shift signal. After oral introduction, the probe was advanced gently until its tip was located in the mid-esophagus and then rotated so that a characteristic aortic blood flow signal was obtained. Probe position was optimized by slow rotation and alteration of the depth of insertion to generate a clear signal. Both an audible, maximal pitch and a visual signal that produced a sharply defined velocity wave form with minimal spectral dispersion defined the right probe position. Gain setting was adjusted to obtain the best outline of the aortic velocity waveform. Before each measurement, probe position was adjusted to ensure optimal acquisition of the maximal velocity signal. Failure to obtain a good quality and stability of the signal or the loss of the signal during the study resulted in exclusion of the patient from the study. The monitor was preset to calculate CO by averaging 5 consecutive measurements. CO is obtained from the product of HR and stroke volume (SV), which was calculated as follows:
|
|
where VAo (t) represents instantaneous maximum aortic velocity, T is the cardiac ejection time, CSAAo is the cross-sectional area of the descending aorta, and K is a correcting factor (= 1.43) the purpose of which is to transform the blood flow measured in the descending thoracic aorta into global CO, assuming that a constant fraction (70%) of the total blood flow passes through the descending aorta. CSAAo is estimated from a normogram based on the patient’s age, weight, and height. Because of the lack of such normograms for children younger than 10 yr inside the ODM 2 software, we arbitrarily used this 10-yr normogram for all patients. Systemic vascular resistance (SVR) is calculated as the quotient of MAP and CO, assuming arbitrarily the right atrial pressure equal to zero.
The anesthesiologist who recorded hemodynamic variables was blinded to the kind of epidurally injected mixture. After induction of GA, data recording began for 20 min to assess hemodynamic stability (T-20 to T-5), and continued during the 30 min after induction of EA (T5 to T30). Surgery was allowed to begin after completion of the study. An increase of more than 20% in HR or MAP at surgical incision was considered as a failure of EA and resulted in exclusion of the patient from the study.
Hemodynamic results are expressed as percentages of the value obtained at the first measure (T-20). Within each group, searching a significant modification for each hemodynamic variable was made with a Friedman test followed by Wilcoxon’s test for paired samples with Bonferroni correction when appropriate. The value of each hemodynamic variable obtained at the end of the initial 20-min period, just before induction of the epidural block (T-5), was considered as the reference value for paired comparisons. A preliminary analysis, obtained from the first 29 included patients, showed that the number of patients required to assert a 20% change in CO with a ß risk of 10% must be at least 5 patients per group. The Mann-Whitney U-test and Wilcoxon’s test were used to compare caudal to thoracolumbar groups receiving the same mixture and plain to adrenaline groups receiving the same level of block. The data are given as median (10%; 90% confidence intervals). Statistical significance was accepted for P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
|
|
|
|
By contrast, CO increased significantly in A groups respectively by 20% (-7; +32) and 34% (+13; +71) (P = 0.012) in the CA and TLA groups. However, the Friedman test, showing overall group differences, was the only one to be significant in the CA group (P = 0.015) even though the individual post hoc comparisons were not. SV also increased significantly in the 2 adrenaline groups, respectively, by 26% (+6; +54) (P = 0.0174) and 45% (+23; +58) (P = 0.012) in the CA and TLA groups. Concurrently, MAP and SVR decreased significantly, respectively by 14% (-21; +9) (P = 0.012) and 24% (-47; 0) (P = 0.012) in the CA group, and 17% (-20; 0) (P = 0.0468) and 40% (-56; +18) (P = 0.012) in the TLA group.
Inter-group comparisons, when restricted to groups that were comparable before EA, showed that the alterations of CO, SV, and SVR were greater in the TLA group than in the TLP and CA groups (Table 3).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In comparison with our results, other pediatric studies reported a lesser decrease of MAP. MAP is reported to decrease by <10%, or not vary at all, and the worsening effect of A has never been emphasized (1–5,13). Several methodological considerations may account for this discrepancy.
The extent of the block is the main factor of the intensity of hemodynamic modifications in adults (14). In children, such an impact has been reported only in the study by Murat et al. (1), where hypotension was deeper when the blocks rose up to T4-5 rather than T10-12. We determined the volumes and the concentrations of local anesthetics, as well as the height of epidural approach, to obtain a spread of block adjusted to the surgical procedure without exceeding admitted thresholds of systemic toxicity, in accordance with our usual practice (15). We did not try to measure the extent of the sensory block, because of methodological difficulties, owing to combined GA, principally (16). The extent of the sensory block related to EA is hazardous in children, depending on their age (2,3,17). No simple mathematical formula allows its precise calculation (2,3,17). On a weight basis, volume of LA required per spinal segment increases progressively when age decreases (17). Proportioning this volume to body surface area has been suggested (17). As our age range was relatively narrow, and in view of the lack of perioperative hemodynamic reaction, we assumed that extent of sensory block was located between T8 and T10 dermatomes in the caudal groups (2) and between T4 and T6 in the TL groups (3,17,18). Though we cannot exclude that some blocks reached above T4, especially in the thoracolumbar groups, the noteworthy stability of CO in TLP group does not support this hypothesis (19). Although not significant, the larger decrease in MAP and SVR in the TLP versus CP groups argues in favor of the occurrence, as in adults, of a vascular sympathetic block whose clinical consistency increases with the spread of block. Hence, the difference in the spread of the block may explain the discrepancy between our study and others.
Sevoflurane, our GA drug, decreases SVR to a larger extent than halothane (20). A sevoflurane-mediated greater inhibition of the well-documented compensatory vasoconstriction in the unblocked area may have unmasked hemodynamic effects of EA compared with those studies previously performed with halothane (1,4).
A moderate decrease of HR associated with lumbar EA has already been reported in children and adults while the cardiac sympathetic fibers are not blocked (1,8,13,19). In adults, this is attributed to a cardiopulmonary-mediated reflex increase in cardiac vagal activity related to the decrease in venous return (8,19). However, an echographic study failed to show a decrease in preload related to lumbar EA in children (13). Besides, in contrast to adults, a predominant arterial baroreflex-mediated withdrawal of cardiac vagal activity has been reported during spinal anesthesia in infants (21).
The stability of CO in the plain groups, already reported by Payen et al. (4) after caudal anesthesia, is consistent with this lack of preload decrease (13). Among other hypotheses, smaller concentrations of LA used in pediatric anesthesia in comparison with adults may account for a better stability of venous return in children through a weaker and/or less extensive sympathetic blockade (15,17). A lesser sensitivity of the small, poorly myelinated B-fibers to LA in comparison with the large myelinated A-fibers has been reported, leading to the concept of partial sympathetic blockade (22–24).
One major finding in our study is the effect of A addition to LA on the changes of both CO and SVR. In view of the weaker systemic absorption of A-added LA, we increased the dose of LA by 25% in the A as compared with plain groups, taking into account the thresholds of systemic toxicity (15,25). We cannot exclude that this plays a role in the greater decrease of SVR in the A as compared with the plain groups, through a more extensive sympathetic blockade. However, comparison of the CP to TLP groups did not elicit any significant hemodynamic difference, although extent of block was likely to be greater in TLP group. So we assume that the addition of A was the main factor of the hemodynamic changes, particularly the increase in CO. Cardiac and vascular ß effects of A related to its slow systemic absorption from the epidural space are well established in adults and have recently been suggested in children (5,10,11). A more profound sympathetic block related to the decrease of the systemic uptake of A-added LA has also been suggested (19). The greater hemodynamic alterations observed in the TLA group as compared with the CA group can be attributed to the addition of the sympathetic block, which was more extensive in the TLA group and whose intensity was potentially enhanced by A and the systemic ß effect of A. As compensatory vasoconstriction in the unblocked area should be greater in the case of more extensive and profound sympathetic block, its inhibition through the vascular ß effect of A is an interesting hypothesis to explain this additive effect (19).
Despite small alterations in MAP and HR, we elicited, thanks to TED monitoring, important and unexpected changes of CO. In view of the difficulty of invasive monitoring, this emphasizes the usefulness of such a minimally invasive technique in children for hemodynamic assessment. However, some limitations must be kept in mind: CSAAo is estimated from normograms, and descending-aortic to total-blood-flow ratio is fixed equal to 70% (26). The actual CSAAo of our patients was certainly inferior to the estimated CSAAo, unsuited to the age and the morphology of our patients, and descending aortic to total blood flow ratio has been suggested to be <70% in children (5,27). Hence, absolute values of CO reported in Table 1 as SV and SVR are only indicative, and they must not be considered as exact. But because these two factors are constant in the calculation of CO from the measurement of aortic blood velocity by ODM 2 (see Methods), this does not impede following the changes of CO in the course of time in each subject. Indeed, TED has been favorably correlated to thermodilution monitoring of trends of CO in critically ill adults and children (12,28). It is the reason why we expressed our results as variations in percents of the first measure (T-20).
However, the assumption that CSAAo and descending-aortic to total-blood-flow ratio remain constant along the time in each subject is questionable. Indeed, Larousse et al. (5), using TED probe coupled with an echographic monitoring of aortic diameter in young children, showed an increase of aortic diameter by 15% after A-added LA caudal anesthesia. In view of the calculation formula of CSAAo from aortic diameter, this could have caused an underestimation of CO changes by approximately 30% in our study, at least in the A groups. It is striking that Larousse et al., taking changes of aortic diameter into account, reported an increase of descending aortic flow by 70%, thus grossly equivalent to the combination of CO changes reported in our A groups and the 30% CSAAo changes reported in their study. However, aortic diameter monitoring has been suggested to introduce significant error in TED monitoring of CO in children (29). Leather et al. (30) showed in adults that ODM 2 overestimates CO changes related to plain LA lumbar EA by comparison with thermodilution, probably owing to sympathetic blockade-induced redistribution of blood flow towards the lower part of the body, leading to an increase of descending aortic to total blood flow ratio. However, even though agreement between ODM 2 and thermodilution was acceptable in this study before EA, it became unacceptable after EA, possibly owing to the lack of probe repositioning before each measurement for signal optimization (12). Moreover, the increase in spread of block between the CP and TLP groups did not result in any difference in CO trends in our study, suggesting a weaker impact of sympathetic block on blood flow redistribution in children than in adults. Indeed, Payen et al. (4) found a quasi-stability (-6%) of CO after plain LA caudal anesthesia in children monitored with transcutaneous Doppler-echocardiography, a result comparable to ours, and failed to show any increase in femoral blood flow related to blood flow redistribution.
In conclusion, despite limitations of TED monitoring of CO, our study shows the stability of CO during EA with plain LA, and an important increase of CO when A is added. The associated vasodilation related to the sympathetic block is minimal, but seems to be influenced by the spread of block. Hypotension is worsened by A, and its use should be reconsidered in patients with high hemodynamic risk, such as neonates.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  A. Monsel, A. Salvat-Toussaint, P. Durand, V. Haas, C. Baujard, P. Rouleau, S. El Aouadi, D. Benhamou, and K. Asehnoune The Transesophageal Doppler and Hemodynamic Effects of Epidural Anesthesia in Infants Anesthetized with Sevoflurane and Sufentanil Anesth. Analg., July 1, 2007; 105(1): 46 - 50. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
