The most common regional anesthetic in children is the caudal approach to the epidural space (“the caudal block”). The caudal block can be used as an adjunct to general anesthesia (1), administered at the completion of surgery to provide postoperative analgesia (2) or, in specific circumstances, chosen as an alternative to general anesthesia in the high-risk infant (3). With the increased use of caudal epidural anesthesia, there has also been an increased appreciation and recognition of the potential for adverse effects related to the technique of placement or because of systemic toxicity related to the local anesthetic (4,5). One of the time-honored methods of identifying accidental systemic injection is the “test dose.” However, its validity has been questioned in the literature and several centers have abandoned its use. This article reviews the current medical literature available regarding the test dose as it pertains to the use of regional anesthesia in infants and children.
One of the primary means of limiting toxicity from inadvertent systemic injection is the development of a technique that promptly and reliably identifies systemic injection. Although aspiration and return of blood is definite evidence for intravascular needle/catheter placement, a negative aspiration for blood lacks sensitivity in preventing systemic administration. As a result of a lack of complete ossification of the sacral vertebral bodies in neonates and young infants, direct intraosseous injection can also occur; it perhaps accounts for many reports of systemic toxicity despite a negative aspiration. Because of the failure of aspiration of blood for predicting the potential for systemic injection of local anesthetics, other means of preventing or identifying systemic injection of local anesthetic solutions are necessary.
In common clinical practice, the identification of systemic injection has relied on “the test dose,” which contains an amount of epinephrine that theoretically, if injected systemically, will cause tachycardia and hypertension. The time-honored criteria for identifying intravascular injection have recently been questioned and/or modified. The initial criteria used for suggesting a positive response to test dosing during placement of a caudal block in children were extrapolated from the adult literature. Moore and Batra (6) defined a positive response in adults as an increase in heart rate (HR) of 30 bpm or more. Although these data were initially extrapolated to the pediatric population, their validity was questioned by Desparmet et al. in 1990 (7). These investigators evaluated the HR response after the IV administration (simulated intravascular test dose) of 0.1 mL/kg of 0.1% lidocaine with epinephrine 1:200,000 (epinephrine dose of 0.5 μg/kg) to children anesthetized with 1 MAC halothane in a 50% nitrous oxide and oxygen. In 19 of 20 patients who received atropine (10 μg/kg), there was an increase in the HR of 10 bpm after the intravascular administration of epinephrine whereas only 16 of 21 patients who had not received atropine had an increase in the HR of 10 bpm or more. This study clearly demonstrated that the criteria suggested by Moore and Batra (6) for adults could not be applied to pediatric-aged patients. Desparmet et al. (7) also demonstrated that there was an increase in systolic blood pressure of a similar magnitude, regardless of whether atropine was given; however, they reported their blood pressure data as the mean ± sds, and therefore the number of patients who had a specific increase in blood pressure was not presented, and no threshold was identified that could be used to define a positive response. The data of Desparmet et al. (7) demonstrating that the “test dose” and HR response was not 100% sensitive in identifying intravascular injection was further supported by clinical reports of toxicity in pediatric-aged patients receiving caudal epidural blocks with epinephrine containing local anesthetic solutions (8,9).
As a follow-up to their 1990 study (7), Desparmet et al. evaluated HR responses after the administration of either IV epinephrine or isoproterenol with and without the administration of atropine in halothane-anesthetized (1 MAC), 2-wk-old lambs (10). Each of the eight lambs received IV (simulated intravascular test doses) epinephrine (0.25, 0.5, 0.75, and 1 μg/kg), isoproterenol (0.05, 0.1, 0.15, 0.2 μg/kg) and then lidocaine (1, 2, 3, and 4 mg/kg) with epinephrine (0.5, 1, 1.5, and 2 μg/kg) respectively. The study protocol was then repeated after the administration of atropine 10 μg/kg. When atropine was not given, there was a transient increase in HR after epinephrine, but this did not occur in all of the animals, regardless of the dose. Although the administration of atropine resulted in a more sustained response to the epinephrine test doses, it did not occur in all of the animals. When isoproterenol was used, regardless of whether atropine was administered, there was a sustained response and the HR changed in all of the animals. No difference was noted when lidocaine was added to the test dose solution.
Because of the unreliability of epinephrine and after the animal study of Desparmet et al. (7) suggesting the superiority of isoproterenol (10), Perillo et al. (11) evaluated isoproterenol to detect inadvertent intravascular administration in anesthetized children. Anesthesia included 1.2 MAC halothane in 50% nitrous oxide and oxygen. Atropine was not administered. There was an increase in HR of ≥10 bpm in 17 of 21 patients that received 0.05 μg/kg and in 22 of 23 patients who received 0.075 μg/kg of isoproterenol administered IV (simulated intravascular test dose). The dose of isoproterenol was administered in a 0.5% lidocaine solution (0.5 mg/kg lidocaine). However, because of the lack of available information concerning the safety in regards to neurotoxicity of isoproterenol if administered into the epidural space, the authors could not advocate the use of isoproterenol in caudal epidural anesthesia.
A subsequent study also evaluated the potential of isoproterenol for use as the test dose (12). The authors defined a positive response as a HR increase of ≥20 bpm as suggested by Guinard et al (13). Saline control or 1 of 3 doses of isoproterenol (0.05, 0.075, and 0.1 μg/kg) were administered IV (simulated intravascular test dose) in the awake state and then during anesthesia with 1.2 MAC halothane in 70% nitrous oxide and oxygen. All of the test doses, including the saline control, were administered in a solution that contained 0.25 mg/kg bupivacaine. No atropine was administered. No changes were noted with the saline/bupivacaine control. In the awake state, all three doses of isoproterenol resulted in a HR increase of ≥20 bpm in all patients. However, after the induction of anesthesia, the sensitivity was 79%, 89%, and 100% respectively with doses of 0.05, 0.075, and 0.1 μg/kg. Transient ventricular arrhythmias were noted in one patient with the administration of 0.075 μg/kg of isoproterenol. No data were specifically given regarding the sensitivity if the criteria of ≥10 bpm increase in HR was used. The authors reported the time required for the HR to increase by 10 bpm or more after isoproterenol. The average time (in seconds) to achieve a HR increase of 10 bpm or more was significantly shorter with all isoproterenol doses in the awake state when compared with the anesthetized state (14 versus 24 s, 12 versus 19 s, and 9 versus 13 s respectively, when comparing awake versus anesthetized after 0.05, 0.075, and 0.1 μg/kg isoproterenol). These data may suggest that all of the patients achieved an increase in HR of ≥10 bpm in both the awake and anesthetized state, regardless of the dose of isoproterenol used. If that is the case, the sensitivity was 100% for all doses of isoproterenol in the awake and anesthetized state. The authors did not recommend the use of epidural isoproterenol because of the lack of information concerning its safety for neuraxial administration.
With the increased use of sevoflurane for pediatric anesthesia, Tanaka and Nishikawa (14) assessed the reliability of an epinephrine test dose (0.5 μg/kg) in 1% lidocaine (1 mg/kg), but this time during 1 MAC sevoflurane anesthesia in 60% nitrous oxide and oxygen. The test dose was administered IV (simulated intravascular test dose). The authors again noted that a positive response of an increase in HR of ≥20 bpm was not adequate in detecting inadvertent systemic injection. Only 8 of 15 patients pretreated with atropine and 10 of 15 who did not receive atropine developed an increase in HR of ≥20 bpm after the test dose. Using a criterion of an increase of ≥10 bpm, they noted a positive response in all children in the test dose (15/15) and test dose plus atropine (15/15) groups. When considering an increase in systolic blood pressure of ≥15 mm Hg, 15 of 15 patients that received atropine and the test dose and 10 of 15 who received only the test dose met the criteria. Patients who had received atropine had a more rapid increase of HR (15–60 s versus 15–90 s) and a more sustained HR and systolic blood pressure response.
Sethna et al. (15) evaluated the reliability of test dosing with epinephrine 0.5 μg/kg or 0.75 μg/kg in 1% lidocaine (1 mg/kg) after atropine (10 μg/kg) administration to children anesthetized with 1 MAC isoflurane in 100% oxygen. In their report, the authors explain that atropine was given to all patients based on their previous study (published only in abstract 1 form) and found that the larger dose of epinephrine (0.75 μg/kg) did not reliably increase HR in patients who had not received atropine. The test dose included 0.1 mL/kg of 1% lidocaine with one of two doses of epinephrine: 0.5 μg/kg or 0.75 μg/kg administered IV (simulated intravascular test dose). The median maximum increase in HR was similar with the two doses of epinephrine. However, after epinephrine 0.5 μg/kg, a HR increase of ≥10 bpm occurred in 19 of 21 patients, whereas a systolic blood pressure increase of ≥15 mm Hg occurred in 17 of 21 patients. After 0.75 μg/kg of epinephrine, a HR increase of ≥10 bpm occurred in 21 of 21 whereas a systolic blood pressure increase of ≥15 mm Hg occurred in 19 of 21 patients. The authors again demonstrated the unreliability of looking for a HR increase of ≥20 bpm as only 10 of 21 who received 0.5 μg/kg epinephrine and 12 of 21 who received 0.75 μg/kg epinephrine had a positive response.
Kozek-Langenecker et al. (16) compared the HR responses to isoproterenol in children during anesthesia with 1 MAC of either sevoflurane or halothane. Incremental doses of isoproterenol were administered IV (simulated intravascular test dose) to achieve an increase in HR of ≥20 bpm. The minimal effective dose of isoproterenol to increase HR by ≥20 bpm was 55 ng/kg (42–72 ng/kg; 95% confidence intervals) during sevoflurane anesthesia and 32 ng/kg (26–38 ng/kg; 95% confidence intervals) during halothane anesthesia. The authors concluded that larger doses of isoproterenol are needed during sevoflurane anesthesia and suggested that this may be related to sevoflurane’s effect on β-adrenoceptor signal transduction.
Although the initial practices focused on HR or blood pressure changes to detect inadvertent systemic injection, Freid et al. (17) in 1993 reported electrocardiographic changes (ST-T wave changes) and bradycardia with inadvertent systemic injection in five neonates and infants during caudal epidural block. Although they noted ST-T wave changes, no tachycardia occurred in any of the five patients. The changes were reproducible in two patients with repeated dosing, and in two patients blood was noted in the tubing or hub of the needle after administration of the test dose. In the fifth patient, the serum bupivacaine level was 2.6 μg/mL after the administration of 1 mg/kg of bupivacaine. The authors referred to bupivacaine toxicity data from animal studies (18,19) and suggested that the electrocardiogram (ECG) changes were related to the local anesthetic, bupivacaine, and not epinephrine. They also observed HR slowing and increased blood pressure in some of the infants after the ECG changes. They attributed these hemodynamic changes to epinephrine with an increased blood pressure and reflex HR slowing. Support for the suggestion that the ECG changes are related, at least in part, to the local anesthetic is provided by the recent report of Tanaka et al. (20) of a 2-mo-old infant who developed increases in T-wave amplitude during the caudal injection of a mixture of 0.25% bupivacaine and 1% lidocaine without epinephrine.
Fisher et al. (21) reported changes in T-wave amplitude (increase by ≥25%) in 25 of 30 patients with known intravascular injection in a prospective, observational study of epidural anesthesia over a 12-mo period. No attempt was made to control the anesthetic care so that the test dose varied significantly although most patients received approximately 0.5 μg/kg of epinephrine in either 0.25% bupivacaine solution or 1–2% lidocaine solution. They noted an incidence of systemic injection of 5.6% (42 of 742 blocks). ECG recordings were available from 30 of these patients. Twenty-five (83%) had an increase in T-wave amplitude of ≥25%, whereas 29 (97%) had either a change in T-wave amplitude or alteration in the rhythm (bradycardia or nodal rhythm). The authors suggested adding these criteria as a means of identifying inadvertent systemic injection because an increase in HR may not always occur because of HR slowing effects of halothane, baroreceptor reflexes after an increase in arterial pressure with epinephrine, or an autonomic balance favoring parasympathetic tone in infants and children.
After these studies demonstrating changes in the T wave in response to inadvertent systemic injection, Tanaka and Nishikawa (22) evaluated the sensitivity and specificity of changes of HR (≥ 10 bpm), systolic blood pressure (≥ 15 mm Hg), and T-wave amplitude (≥ 25%) in children anesthetized with 1 MAC sevoflurane in 67% nitrous oxide and oxygen. All patients received atropine before the IV administration of the solution (simulated intravascular test dose), which included 0.5 μg/kg of epinephrine in 1% lidocaine (1 mg/kg) or saline placebo. No changes were noted in the 16 children that received placebo. Of the 16 children that received the test dose, there was a positive response in 16/16, 13/16, and 16/16 respectively using the previously described heart rate, systolic blood pressure, and T-wave criteria. These changes occurred earliest in the T wave (20 ± 5 s), next in the HR (30 ± 7 s), and last in the systolic blood pressure (70 ± 31 s).
A similar study sought to identify the positive predictive response of T-wave amplitude (increase of ≥25%), HR (increase of ≥10 bpm), and systolic blood pressure (increase of ≥15 mm Hg) in 42 children anesthetized with either 1 MAC halothane or sevoflurane in 70% nitrous oxide and oxygen (23). Because atropine administration may not be part of routine pediatric anesthesia practice in older infants and children, the authors opted to exclude it from their study. After the IV administration of epinephrine (0.5 μg/kg) with no local anesthetic, the investigators noted a positive response rate (T-wave amplitude, HR, blood pressure) in 100%, 95%, and 71% of children anesthetized with sevoflurane and 90%, 71%, and 71% of children anesthetized with halothane. They concluded a change in T-wave amplitude was superior to increases in either HR or blood pressure and that the sensitivity of blood pressure and T-wave amplitude changes was greater with sevoflurane than halothane.
Tanaka et al. (24) provided insight into the mechanisms accounting for the T-wave changes. They evaluated the positive predictive efficacy of HR (≥10 bpm) and T-wave amplitude (≥25% increase) after the IV (simulated intravascular test dose) administration of 0.5 μg/kg epinephrine with 0.25 mg/kg bupivacaine or isoproterenol 0.1 μg/kg with 1 mg/kg lidocaine (previous studies have demonstrated changes in T-wave amplitude with epinephrine 0.5 μg/kg with 1 mg/kg lidocaine). Anesthesia for the study included 1 MAC sevoflurane in 67% nitrous oxide and oxygen. All patients received atropine (10 μg/kg). Both epinephrine and isoproterenol were 100% accurate by HR criteria; however, no T-wave changes occurred with isoproterenol whereas T-wave changes occurred in 100% of patients receiving epinephrine.
Tanaka and Nishikawa (25) sought to determine the dose response curve for epinephrine used in the test dose. Anesthesia included 1 MAC sevoflurane in 67% nitrous oxide/oxygen with atropine. The test dose included 0.1, 0.05, or 0.025 mL/kg of a 1% lidocaine with epinephrine 1:200,000 solution to give an amount of epinephrine in the different test doses of 0.5, 0.25, or 0.125 μg/kg. When considering T-wave criteria of an increase in T-wave amplitude of ≥25%, 20/20 in both the 0.25 and 0.5 μg/kg epinephrine dose developed a positive response. When using HR criteria (increase of ≥10 bpm), 20/20 with the 0.5 μg/kg dose and 17/20 with the 0.25 μg/kg dose demonstrated a positive response. No criterion was consistently altered with the 0.125 μg/kg dose.
During the performance of regional anesthesia in children, the inadvertent intravascular injection of local anesthetics can be catastrophic. This complication may occur in up to 0.4% of caudal epidural blocks in pediatric-aged patients (2). The initial technique suggested to recognize this complication was the use of a test dose with criteria extrapolated from the adult literature suggesting a HR increase of 20–30 bpm as the identifying feature. Subsequent studies in infants and children demonstrated that the use of the adult criteria resulted in a low sensitivity for the test dose (Table 1). These studies suggested new pediatric criteria including an increase of HR of ≥10 bpm or systolic blood pressure of ≥15 mm Hg and also demonstrated that the hemodynamic changes do not always occur early, with some patients developing HR or blood pressure changes 60 to 90 s after injection. During halothane or isoflurane anesthesia, but not during sevoflurane anesthesia, the sensitivity of the hemodynamic criteria is increased with the administration of atropine or the use of a larger dose of epinephrine (0.5 versus 0.75 μg/kg). Although larger doses of epinephrine may increase the sensitivity of the test dose, there is also a concern that these larger doses may be associated with ventricular arrhythmias. Atropine premedication with sevoflurane anesthesia prolongs the duration of the tachycardia or the hypertension with the test dose.
Subsequent observations led to the suggestion of changes in T-wave amplitude or the ST segment as more sensitive means of identifying inadvertent systemic injection. Additionally, HR slowing or the development of nodal or sinus bradycardia were uncommon but specific signs of systemic injection. Sub-sequent analysis has demonstrated that changes in T-wave amplitude occur earliest, followed by changes in HR and then changes in systolic blood pressure. These ECG changes may be related primarily to the local anesthetic (bupivacaine) itself and not necessarily the epinephrine.
Because of issues related to the standard test dose incorporating 0.5 μg/kg of epinephrine, isoproterenol has been suggested as being more sensitive. The sensitivity of the isoproterenol test dose is increased with larger doses (0.05 versus 0.075 versus 0.1 μg/kg) without an increased incidence of arrhythmias in preliminary studies. Smaller doses of isoproterenol are needed during halothane anesthesia compared with sevoflurane anesthesia. Although the initial studies are encouraging and suggest that isoproterenol may be more sensitive than epinephrine during isoflurane or halothane anesthesia, there are no data regarding the potential for neurotoxicity with isoproterenol, thereby limiting its clinical utility.
In conclusion, during the performance of regional blockade in children, several steps should be taken to limit the potential for inadvertent systemic injection. Given its lack of adverse effects, the current literature (Table 1) supports the use of an epinephrine test dose (0.5 μg/kg) with the above mentioned qualifications. Although the sensitivity of any means of test dosing does not reach 100%, the specificity in all of the studies reviewed has been 100%. There have been no false positive (HR, blood pressure) results with the injection of saline or local anesthetic, suggesting that any significant electrocardiographic (HR increase of ≥10 bpm) or hemodynamic (increase of systolic blood pressure of ≥15 mm Hg) change occurring with the administration of the test should be taken as indicating inadvertent systemic injection. Observation of not only the HR and systolic blood pressure, but also the T-wave amplitude should increase the sensitivity of the test dose and aid in the recognition of inadvertent systemic injection. T-wave changes occur earliest, followed by HR changes and then by blood pressure changes. However, T-wave changes do not occur with isoproterenol, suggesting that the mechanism is not only a β-adrenergic receptor effect. The mechanisms responsible for these ECG changes have not been clearly delineated. T-wave changes have been described when only epinephrine is given, when only the local anesthetic is given, and when both agents are administered together.
The indicators of inadvertent systemic injection may be delayed for up to 60–90 s after the test dose, suggesting an appropriate observation period of 90 s after the test dose before delivery of the remainder of the local anesthetic solution. The sensitivity of the test dose during halothane or isoflurane anesthesia can be increased by the prior administration of atropine.
Another factor to consider is that most of the studies reviewed in this article administered the entire test dose IV. This may not always be the case in clinical practice, as only a fraction of the test dose may gain access to the systemic circulation. Several of the studies have demonstrated that the response to the test dose, whether epinephrine or isoproterenol, is dose-related, so that if only half of the test dose gains access to the systemic circulation, the sensitivity can be expected to decrease. This factor may account for some of the false negative results that occur clinically. As no method will be 100% sensitive, one should, whenever possible, consider fractionating the dose of the local anesthetic. Even if the test dose is negative, it is not suggested that the entire volume of the local anesthetic solution be delivered rapidly. The dose should be administered incrementally in 0.1–0.2 mL/kg aliquots with an observation time of 60–90 s after each injection. Additional factors that may limit the risks of systemic toxicity include limitation of the total dose of local anesthetic used by using the minimal effective concentration and volume, as well as the use of newer local anesthetics with the potential for decreased cardiac toxicity (26). Future studies will need to consider the accuracy of test dosing during total IV anesthesia with newer agents such as propofol or remifentanil that can have significant effects on heart rate. Additionally, the differential effects of the local anesthetics including newer drugs such as ropivacaine and levobupivacaine and their modulation of the adrenergic drugs’ effects on HR, blood pressure, and ECG changes need further evaluation and clarification.
1 Sang C, Sethna N, Sullivan L, Berde C. Evaluation of epinephrine epidural test dose in children under isoflurane anesthesia [abstract]. Anesthesiology 1994;81:A1344. Cited Here...
1. Tobias JD, Rasmussen GE, Holcomb GW III, et al. Continuous caudal anaesthesia with chloroprocaine as an adjunct to general anesthesia in neonates. Can J Anaesth 1996; 43: 69–72.
2. Dalens B, Hansanoui A. Caudal anesthesia in pediatric surgery: success rate and adverse effects in 750 consecutive patients. Anesth Analg 1989; 68: 83–9.
3. Spear RM, Deshpande JK, Maxwell LG. Caudal anesthesia in the awake, high-risk infant. Anesthesiology 1988; 69: 407–9.
4. Flandin-Blety C, Barrier G. Accidents following extradural analgesia in children: the results of a retrospective study. Paediatr Anaesth 1995; 5: 41–6.
5. Ved SA, Pinosky M, Nicodemus H. Ventricular tachycardia and brief cardiovascular collapse in two infants after caudal anesthesia using a bupivacaine-epinephrine solution. Anesth Analg 1993; 79: 1121–3.
6. Moore DC, Batra MS. The components of an effective test dose prior to epidural block. Anesthesiology 1981; 55: 693–6.
7. Desparmet J, Mateo J, Ecoffey C, Mazoit X. Efficacy of an epidural test dose in children anesthetized with halothane. Anesthesiology 1990; 72: 249–51.
8. Matsumiya N, Dohi S, Takahashi H, et al. Cardiovascular collapse in an infant after caudal anesthesia with a lidocaine-epinephrine solution. Anesth Analg 1986; 65: 1074–6.
9. Eyres RL, Brown TCK, Hastings C. Plasma level of bupivacaine during convulsions. Anaesth Intens Care 1983; 2: 385.
10. Desparmet JF, Berde CB, Schwartz DC, Lunn RJ. Efficacy of adrenaline, lignocaine-adrenaline and isoprenaline as a test dose in halothane-anesthetized lambs. Eur J Anesth 1991; 8: 123–8.
11. Perillo M, Sethna NF, Berde CB. Intravenous isoproterenol as a marker for epidural test-dosing in children. Anesth Analg 1993; 76: 178–81.
12. Kozek-Langenecker S, Chiari A, Semsroth M. Simulation of an epidural test dose with intravenous isoproterenol in awake and in halothane-anesthetized children. Anesthesiology 1996; 85: 277–80.
13. Guinard J, Mulroy MF, Carpenter RL, Knopes KD. Test dose: optimal epinephrine content with and without acute beta-adrenergic blockade. Anesthesiology 1990; 73: 386–92.
14. Tanaka M, Nishikawa T. Simulation of an epidural test dose with intravenous epinephrine in sevoflurane-anesthetized children. Anesth Analg 1998; 86: 952–7.
15. Sethna NF, Sullivan L, Retik A, et al. Efficacy of simulated epinephrine-containing epidural test dose after intravenous atropine during isoflurane anesthesia in children. Reg Anesth Pain Med 2000; 25: 566–72.
16. Kozek-Langenecker SA, Marhofer P, Krenn CG, et al. Simulation of an epidural test dose with intravenous isoproterenol in sevoflurane- and halothane- anesthetized children. Anesth Analg 1998:549–52.
17. Freid EB, Bailey AG, Valley RD. Electrocardiographic and hemodynamic changes associated with unintentional intravascular injection of bupivacaine with epinephrine in infants. Anesthesiology 1993; 79: 394–8.
18. Heavener JE. Cardiac dysrhythmias induced by infusion of local anesthetics into the lateral cerebral ventricle of cats. Anesth Analg 1986; 65: 133–8.
19. Denson DD, Behbehani MM, Gregg RV. Enantiomer specific effects of intravenously administered arrythmogenic dose of bupivacaine on neurons of the nucleus tractus solitarius and the cardiovascular system in the anesthetized cat. Reg Anesth 1992; 17: 311–6.
20. Tanaka M, Nitta R, Nishikawa T. Increased T-wave amplitude after accidental intravascular injection of lidocaine plus bupivacaine without epinephrine in sevoflurane-anesthetized child. Anesth Analg 2001; 92: 915–7.
21. Fisher QA, Shaffner DH, Yaster M. Detection of intravascular injection of regional anaesthetics in children. Can J Anaesth 1997; 44: 592–8.
22. Tanaka M, Nishikawa T. Evaluating T-wave amplitude as a guide for detecting intravascular injection of a test dose in anesthetized children. Anesth Analg 1999; 88: 754–8.
23. Kozek-Langenecker S, Marhofer P, Jones K, et al. Cardiovascular criteria for epidural test dosing in sevoflurane and halothane-anesthetized children. Anesth Analg 2000; 90: 579–83.
24. Tanaka M, Kimura T, Goyagi T, et al. Evaluating hemodynamic and T wave criteria of simulated intravascular test doses using bupivacaine or isoproterenol in anesthetized children. Anesth Analg 2000; 91: 567–72.
25. Tanaka M, Nishikawa T. The efficacy of a simulated test dose in sevoflurane-anesthetized children: a dose response study. Anesth Analg 1999; 89: 632–7.
26. Khalil S, Campos C, Faraq AM, et al. Ropivacaine compared with bupivacaine in children. Anesthesiology 1999; 91: 1279–84.