High neuraxial block is a common cause of anesthesia-related maternal death. Within the American Society of Anesthesiologists Closed Claims Database, two-thirds of high blocks were caused by accidental intrathecal injection through a presumed epidural catheter.1 The epidural test dose is designed to minimize the risk of accidental intrathecal injection. The test dose should clearly demonstrate that the catheter is not intrathecal, or demonstrate that it is intrathecal, without posing a threat to the patient. Ideally this would mean that no sensory or motor block would occur if the test dose were injected into the epidural space, and an obvious, but safe, block would occur if the same dose were injected into the intrathecal space. Lidocaine 45 mg with epinephrine 15 μg (1.5% lidocaine with epinephrine 5 μg/mL: 3 mL) is a common epidural test dose. The sensitivity and specificity of this dose of lidocaine have been determined in nonpregnant patients.2 Although commonly recommended for use in obstetrics, the safety of this dose has been questioned, because high,3-5 prolonged,5 and total6 spinal blocks have been reported. Richardson et al.3 found that all 5 patients in whom an intrathecal catheter was identified by a positive test dose with 45 mg of lidocaine after a negative aspiration developed a high sensory level (range T6 to C6) and significant hypotension in the labor room.
Pregnant patients are more sensitive to the effects of local anesthetics,7 and intrathecal local anesthetic dose requirements are reduced in pregnancy.8,9 Thus, a smaller dose of lidocaine might be effective at identifying accidental intrathecal placement while producing less-extensive and profound block and a lower risk of high block. Abraham et al.10 found that lidocaine 30 mg produced rapid, objective evidence of spinal blockade in 15 pregnant women. The average cephalad extent of the sensory block was T9. However, the authors did not determine the sensitivity or specificity of this dose, nor compare it to the more commonly administered 45-mg dose. We undertook this study to determine whether lidocaine 30 mg was as effective as lidocaine 45 mg in creating objective and rapid evidence of a sensory or motor block diagnostic of intrathecal injection.
The IRB approved this randomized, double-blind, prospective trial. One hundred healthy parturients who presented for elective cesarean delivery with an indication for combined spinal epidural (e.g., repeat, previous abdominal surgery) were enrolled after written informed consent. Subjects were randomized by computer-generated table. Group assignment was maintained in an opaque, sealed envelope, which was opened after enrollment. Subjects were assigned to 1 of 4 groups:
- Group IT30 received a spinal injection of 3 mL 1% lidocaine (30 mg) with epinephrine 5 μg/mL (15 μg) (Abbott Laboratories, North Chicago, IL).
- Group IT45 received a spinal injection of 3 mL 1.5% lidocaine (45 mg) with epinephrine 5 μg/mL (15 μg) (Astra USA, Westborough, MA).
- Group EP30 received an epidural injection of 3 mL 1% lidocaine (30 mg) with epinephrine 5 μg/mL (15 μg).
- Group EP45 received an epidural injection of 3 mL 1.5% lidocaine (45 mg) with epinephrine 5 μg/mL (15 μg).
All patients received 1 L of IV lactated Ringer’s solution before administration of the anesthetic. The neuraxial anesthetic was performed at the L2-3 or L3-4 interspace using loss-of-resistance to saline with the patient in the sitting position. The amount of saline injected was not measured. In the intrathecal groups (IT30 and IT45), the study drug was administered through a 24-gauge Sprotte needle as part of a needle-through-needle combined spinal–epidural technique. After the spinal drug was administered, a multiport, 20-gauge, nylon epidural catheter (Smith Medical, Portex, Keene, NH) was placed in the epidural space, and the patient positioned supine with 30° left uterine displacement. In the epidural groups (EP30 and EP45), the epidural catheter was placed without a spinal injection, and the study drug was administered through the catheter with the patient seated. The patient was placed supine with 30° left uterine displacement after study drug injection. A blinded assessor entered the operating room after the patient was in position. Both the assessor and the patient were blinded to study drug dose and route of administration. After completion of all study assessments, additional local anesthetic was administered through the epidural catheter as needed to obtain adequate surgical anesthesia. Subjects were excluded if cerebrospinal fluid could not be identified through the Sprotte needle in the spinal groups or if the epidural catheter failed to produce adequate anesthesia for surgery in the epidural groups.
Assessments of the subjects were performed by the blinded assessor at 2, 3, 5, and 10 minutes after the lidocaine dose. Assessments included sensory block to cold (alcohol swab) and to pinprick (patient asked if a plastic needle felt sharp) starting at L5 and moving cephalad until no sensory level could be identified, the patient’s subjective sense that her legs felt heavy or warm, and the presence and degree of motor block using the modified Bromage scale11:
- 1 = Complete block (unable to move feet or knees).
- 2 = Almost complete block (able to move feet only).
- 3 = Partial block (just able to move knees).
- 4 = Detectable weakness of hip flexion while supine (full flexion of knees).
- 5 = No detectable weakness of hip flexion while supine.
- 6 = Able to perform partial knee bend (not performed for this study).
The order of the assessments was sensory block to cold, followed by pinprick, followed by motor block assessment. The subjective degree of sensory changes was asked during the objective assessments. If all assessments could not be completed in the allotted time, the assessor skipped noncompleted assessments and proceeded to the next time period. Data analysis was performed on the completed assessment at teach time interval. In addition, the assessor was asked to make a clinical judgment about the route of administration of the test dose.
The primary end point was evidence of a subjective sense of heaviness or warmth in the parturient’s legs 3 minutes after injection. The 3-minute interval has been used by others10 and is a clinically practical period. We used the objective presence of a motor block or objective evidence of a sensory block to cold or pinprick as secondary end points to help differentiate between spinal and epidural injection in patients who could not be clearly differentiated by the subjective changes alone. We compared subjects who received the same lidocaine dose to determine whether subjective sensory changes, objective sensory blockade, or motor weakness were sensitive indicators of spinal administration of medication. We also compared the IT and EP groups with each other to determine whether there was a significant difference between doses. Secondary outcomes included the degree of sensory and motor blockade and the side effects associated with the test dose in each group. Arterial blood pressure measurement was assessed continually for 3 minutes after the study drug was administered and then every 3 minutes or as deemed appropriate by the anesthesia providers. The anesthesia providers caring for the patient were not blinded to the study group. Hypotension was defined as a decrease in systolic blood pressure of >20% from baseline, any systolic blood pressure <90 mm Hg, or a requirement for a vasopressor (most commonly ephedrine) as determined by the anesthesiologist caring for the patient. A high sensory block was defined as a sensory level to cold or pinprick above T6, identified by the xiphoid process.3 The study ended after the 10-minute assessment.
The rate of accidental intrathecal catheter placement using stiff and styletted catheters was historically as high as 1:50.12 When assessed in research conditions, the likelihood of an intrathecal catheter after a negative aspiration is between 1 in 1750 (0.06%) and 1 in 126,000 (0.0008%).13 However, an intrathecal catheter after negative aspiration may be more common in clinical practice, with a rate between 1:380 and 1:1000.3,4,14 We used the most conservative data, 1:380, to calculate the positive predictive value (PPV) and negative predictive value (NPV). Sample size analysis was performed on the basis of an assumption that intrathecal injection would be detected in all patients in both IT groups. On the basis of this presumed 100% detection rate, sample size was arbitrarily calculated to ensure that the 95% confidence interval (CI) of the detection rate was no worse than 85%, with an α = 0.05. We determined that we would need at least 22 patients per group. We planned to enroll 100 women to ensure that group sizes were achieved.
The Shapiro–Wilk test was used to test for a normal distribution of continuous variables (all P values >0.05). Comparisons between normally distributed variables were performed using 2-sided t test for 2 groups; the Levine test was used to ensure equal variances. Analysis of variance with Bonferroni correction was used for multiple-group comparisons of demographics (age, height, weight, body mass index); the denominator of 6 was calculated for 4 variables (n × (n − 1)/2). Variables that were not normally distributed, or interval scale data (i.e., modified Bromage motor scale), were analyzed using Mann–Whitney test or Kruskall–Wallis test, for 2 or multiple comparisons, respectively. Incidences were compared using Fisher exact test. Exact CIs were calculated using the Newcombe–Wilson method with continuity correction, except when the ratio was equal to, or close to, 0% or 100%, in which case the Clopper–Pearson exact values were used. Comparison of sensory block over time was performed with linear repeated-measures analysis, with post hoc testing using the Bonferroni correction. The data were consistent with a normal distribution (all P values >0.05) when combined by dose (30 and 45 mg) and by route (IT and EP), and the sphericity by Mauchly test. The Greenhouse–Geisser correction was used for lack of sphericity. Not all assessments were completed at each time interval, and thus data were analyzed only for completed data at each time period for each test. Statistical analysis was performed using SPSS for Windows 12.0 (Chicago, IL). P ≤ 0.05 was considered significant.
One hundred women were enrolled between January 1998 and September 2001. Four were excluded for protocol violations, leaving 96 for analysis (Fig. 1). We performed 5 assessments on each patient at each of 4 time periods (1920 total assessments). We were unable to perform 47 assessments (2.4%) because of clinical conditions (patient vomiting, time needed between assessments): 15 at 2 minutes, 7 at 3 minutes, 15 at 5 minutes, and 10 at 10 minutes. There were no differences in the rates of inability to collect data among the groups. The groups were similar in demographics and obstetric characteristics (Table 1).
All patients with intrathecal injection reported either warmth or heaviness or both in their legs at 3 minutes (primary outcome), whereas with epidural injection, 26% (6 of 23) of the EP30 and 41% (11 of 27) of the EP45 patients described changes (Table 2). The rate of subjective changes in both intrathecal doses was significantly greater than the respective EP dose (P < 0.001); the difference between the EP45 and EP30 groups was 15% (95% CI, −14% to 40%; P = 0.23).
By the 3-minute assessment, 100% of women in both of the IT groups also had objective evidence of a sensory block to cold or pinprick. Subjects in the EP groups were less likely to have objective sensory block than the IT groups (P < 0.01 for both). Evidence of any motor block was found in 83% (20 of 24) in IT30 and 100% (21 of 21) of IT45 patients (17% difference, 95% CI, −6% to 38%; P = 0.11) at the 3-minute mark, but was demonstrable in 100% in both groups at the 5- and 10-minute assessments. No patient in the EP30 group had a motor block at 3 minutes, but 2 subjects in the EP45 group did (difference = 7%, 95% CI, −12% to 26%; P = 0.50). By 10 minutes, this increased to 9 of 27 patients (33%, 95% CI, 17% to 54%) in the EP45 group.
The sensitivity and specificity for the 30-mg and 45-mg doses assessed at 3 minutes are shown in Table 3. At a presumed aspiration-negative intrathecal catheter prevalence of 1:380, the observed NPV was 100% for both subjective and objective assessments of a sensory block (95% CI, 99.95%–100% and 99.93%–100% when using subjective sensory changes for 30 mg and 45 mg, respectively); however, the PPVs were very low, suggesting that there could be a high false-positive rate.
Examining the change in sensory level over the 10-minute evaluation, we noted cephalad progression of block to both temperature and pinprick in all groups, with the IT groups being significantly higher than either EP group at all times (F2.348, 204.239 = 110.104, P < 0.001; Fig. 2). The incidence of any demonstrable motor block increased over the 10 minutes of the study, but the EP groups remained statistically similar (10-minute difference 15%, 95% CI, −13% to 39%; P = 0.33). We found no statistical differences in sensory levels between the IT30 and IT45 groups, nor the EP30 and EP45 groups, over the course of the 10 minutes of assessments (P > 0.05 for all).
The blinded observer correctly identified intrathecal injection in 23 of 24 patients (96%) in the 30-mg group, and 22 of 22 patients (100%) in the 45-mg group (P = 1.0, 95% CI for difference: −15% to 23%). The one incorrectly identified patient in the 30-mg intrathecal group had subjective sensation of heavy and warm legs, but only a L1 level to cold and pinprick and no motor block at 3 minutes. The observer correctly identified epidural injection in 23 of 23 patients in the 30-mg group and 25 of 27 patients (93%) in the 45-mg group (P = 0.49, 95% CI for difference: −11% to 26%).
There was a high incidence of side effects in both IT groups, but we did not detect a difference in the rates between the 2 IT doses (Table 4). All women in the IT45 group developed a block above T8 by 10 minutes, and 4 reached a level of T1. In the IT30 group, all women reached a final block height above T11, and 2 reached T1. A cesarean delivery was completed in 1 subject in the IT45 group without the need for additional epidural medication.
We demonstrated that both 30 mg and 45 mg lidocaine with 5 μg/mL epinephrine are very effective at detecting intrathecal injection in the parturient, and also confer high NPVs. Women who received either dose reported subjective sensation of warmth or heaviness in their legs 100% of the time when administered intrathecally. A hypothetical flow chart of 10,000 patients receiving a 30-mg test dose is shown in Figure 3. On the basis of our results, lack of subjective sense of warmth or heaviness after either lidocaine epidural test dose should rule out intrathecal placement of the catheter (box A in Fig. 3 yellow); however, because of the sample size and wide CIs for sensitivity of both doses, we cannot categorically claim that lack of sensory changes precludes intrathecal placement for either dose.
A surprising percentage of parturients who receive epidural lidocaine reported subjective changes or evidence of objective sensory blockade only 3 minutes after injection. Even with the lidocaine 30-mg dose, 1 in 4 developed subjective sensory changes, and 1 in 3 developed objective evidence; these rates were slightly higher with the 45-mg dose. Because the incidence of aspiration-negative intrathecal catheter is low, the vast majority of patients with sensory changes at 3 minutes will have an epidural catheter. Unfortunately, the sensory examination of the IT and EP groups showed sufficient overlap that the objective examination is nondiscriminatory (Fig. 2).
Evidence of a motor block at 3 minutes was 100% specific for the IT30 group: no patient who received this dose via epidural catheter developed a motor block by 3 minutes. For women with subjective changes, the presence of motor block effectively rules in intrathecal placement (box D in Fig. 3 red) if a 30-mg test dose was used. However, 7% of the EP45 group developed a motor block, which potentially confounds the ability to discriminate the position of the catheter, because these women might have been construed as having an intrathecal catheter when they did not. Using the lower bounds of the 95% CI, a clinician might encounter >80 patients with an epidural catheter for every true intrathecal catheter. We were unable to determine whether the lidocaine 30-mg dose is a better discriminator because of our small sample size.
We must also caution that because of the sensitivity of motor block in the IT30 group (and lower-bound 95% CI in the IT45 group), a small number of patients could describe subjective changes with no motor block, but potentially have an intrathecal catheter (B in Fig. 3 green). Reaspiration of the catheter is likely to identify the intrathecal catheters in this group. Safety cannot be attributed solely to the test dose, but must include safe doses of medications and a vigilant clinician. From the American Society of Anesthesiologists Closed Claims Database, 10 of 15 women who died because of a high level with an epidural catheter had a negative test dose. With the exception of low doses administered via patient-controlled epidural analgesia regimens, every dose of local anesthesia administered through an epidural should be treated as a test dose; doses should be fractionated (3–5 mL at a time with aspiration in between), and possible intrathecal placement assessed after each dose. In addition, the anesthesia care provider should be prepared to manage cardiovascular collapse and airway compromise due to a high spinal level whenever local anesthetics are administered.
Because pregnancy is associated with increased sensitivity to local anesthetics, the proper dose required for nonpregnant individuals may not be appropriate in the parturient. Colonna-Romano et al2. studied the ability of 45 mg lidocaine to detect intrathecal injection in comparison with saline. They determined that the inability to maintain a straight leg raise had a sensitivity of 100% and a specificity of 93% for intrathecal placement, and also that the NPV and PPV for this dose were very high. However, they did not have a control group that received epidural medication, so the specificity data cannot be explicitly translated to clinical practice. Abraham et al.10 found that evidence of an S2 sensory block before 3 minutes using 2 mL of 1.5% lidocaine was 100% sensitive and specific for intrathecal placement. However, they did not evaluate the effects of the more commonly used 3-mL dose of 1.5% lidocaine administered intrathecally, and thus were not able to compare the sensitivities between the doses. Studying a nonpregnant population, Poblete et al.15 found that 3 mL (60 mg) of 2% lidocaine administered intrathecally produced a motor block in all patients after 6 minutes, in comparison with none who received it via an epidural catheter. This dose is likely to be unsafe in the labor room, and the need to wait 6 minutes to determine the efficacy of the test is impractical.
The lower dose in our study did not decrease the incidence of side effects, though the study was not powered to detect a difference, and the CIs of the differences between groups were wide. For example, when administered intrathecally, the incidence of sensory level above T6 at 10 minutes was 91% in group IT45 versus 75% in group IT30 (95% CI for difference, −10% to 39%). Post hoc power analysis suggests that we would have to study 78 women per group to detect a reduced rate of high block with 80% power.
One may want to decrease the lidocaine dose administered during the test dose for reasons other than decreasing the likelihood of high spinal anesthesia. The test dose has been shown to have a negative effect on motor function and ambulation during the administration of low-dose labor epidural analgesia. Cohen et al.16 found that the administration of a test dose of 3 mL of 1.5% lidocaine increased the likelihood that the parturient would complain of weak legs and decreased the likelihood she could ambulate. The same test dose has been shown to impair motor function in other studies.17,18 We also found that one-third of subjects who received 45 mg of epidural lidocaine had a significant motor block.
We can identify several limitations to our study design and conclusions. We administered the intrathecal medications through a spinal needle rather than through a catheter. A caudal orientation of an intrathecally placed catheter has the potential to affect the spread of local anesthetics.19,20 The degree to which this would affect the accuracy of our results is not known. Similarly, our study drug is slightly hypobaric in comparison with cerebrospinal fluid; thus, the height of the block could be influenced by patient position.21 All patients were kept seated for nearly 2 minutes after the spinal injection while the epidural catheter was placed. We believe this reflects clinical practice, in which the catheter would be secured and the patient might be asked to lie down with the assumption that the test dose will be negative.
Additionally, because of the sample sizes, the 95% CIs were wide. Figure 3 demonstrates an idealized flow of 10,000 patients receiving lidocaine 30 mg that is based on our actual results. There would be a different flow diagram for the 45-mg dose. Most notably, more women with an epidural catheter would have both sensory and motor changes (false positives). Using the upper or lower bounds of the CIs for each test at each decision point of this diagram could produce dozens of variations. For instance, using the lower bounds of the 95% CI for sensitivities, box A would include 3 intrathecal catheters for either dose that would demonstrate no subjective changes. Conversely, using the upper bounds of the rate of motor block, up to 24% of women receiving the 45-mg dose would develop a motor block despite an epidural location (box C orange), decreasing the PPV of a motor block among women with subjective changes even lower than the 3.4% value that we found (box D). On the basis of our findings, the difference in sensitivity for intrathecal injection between lidocaine 30 mg and 45 mg is likely small; a large study to elucidate the safety and efficacy of these doses is warranted. Such a study would (1) better identify the sensitivities of each dose with more narrow CIs, (2) identify potential risks associated with each dose (e.g., unsafe high levels from the higher dose balanced against the possibility of rarely missing an intrathecal catheter with the lower dose), and (3) identify the impact of each dose on the safety and efficacy when administered epidurally (e.g., increased motor block or hypotension with the higher dose).
Our study was underpowered to detect catastrophic complications. Total spinal anesthesia or even a spinal level high enough to cause respiratory compromise or cardiovascular collapse is very rare after a test dose. In our study, 6 women had a sensory level to T1 (2 IT30 and 4 IT45). While we did not measure the height of motor block, on the basis of the spread of sensory block, a spread of just a few more centimeters in the intrathecal space could have led to diaphragmatic paralysis. This is especially relevant because the test dose may rarely miss an intrathecal catheter. A much larger study would be needed to determine whether the potential risk of missing these intrathecal catheters is offset by a decreased rate of catastrophic events created by a lower dose. Finally, 1 patient in the IT30 group was misidentified by the blinded observer. This may indicate that 30 mg lidocaine is less effective than 45 mg at detecting intrathecal placement of the drug. Using the presence of subjective sensory changes, this patient would have been correctly identified as having a potential intrathecal placement. Using our algorithm, the lack of motor block categorized this as a probable epidural catheter.
In conclusion, the test dose is designed to improve the safety of an epidural catheter by minimizing the risk of intrathecal injection of large doses of local anesthetics. However, the test dose itself may have side effects, ranging from increased motor block to high or total spinal anesthesia. The correct dose should balance the risk of missing an intrathecal catheter against the side effects caused by the test dose. Though the CIs were wide and there is a strong possibility of a type 2 statistical error, we found that 3 mL of lidocaine 1.0% with 1:200,000 epinephrine was not less effective than 3 mL of lidocaine 1.5% at ruling out intrathecal placement. Most parturients will have no subjective changes, proving that the catheter is not intrathecal. The 30-mg dose is also very effective at ruling in an intrathecal catheter for most patients, but may rarely miss an intrathecal catheter. Whether the lower dose would be less sensitive, and whether this lower sensitivity would be offset by a lower rate of dangerously high spinal anesthesia, could not be addressed by this study. A much larger study is required to corroborate these finding and better elucidate the side effects and complications associated with each dose.
Name: Stephen Pratt, MD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Stephen Pratt has seen the original study data, reviewed the analysis of the data, approved the final manuscript, and is the author responsible for archiving the study files.
Name: Philip Hess, MD.
Contribution: This author helped design the study, conduct the study, analyze the data, and write the manuscript.
Attestation: Philip Hess has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
Name: Anasuya Vasudevan, MD, FRCA.
Contribution: This author helped enroll the patients, follow protocol, and collect the data.
Attestation: Anasuya Vasudevan has seen the original study data, reviewed the analysis of the data, and approved the final manuscript.
This manuscript was handled by: Cynthia A. Wong, MD.
1. Davies JM, Posner KL, Lee LA, Cheney FW, Domino KB. Liability associated with obstetric anesthesia: a closed claims analysis. Anesthesiology. 2009;110:131–9
2. ColonnaRomano P, Padolina R, Lingaraju N, Braitman LE. Diagnostic accuracy of an intrathecal test dose in epidural analgesia. Can J Anaesth. 1994;41:572–4
3. Richardson MG, Lee AC, Wissler RN. High spinal anesthesia after epidural test dose administration in five obstetric patients. Reg Anesth. 1996;21:119–23
4. Vallejo MC, Adler LJ, Finegold H, Mandell GL, Ramanathan S. Periosteal entrapment of an epidural catheter in the intrathecal space. Anesth Analg. 2001;92:1532–4
5. Kuczkowski KM. Prolonged spinal anesthesia in a parturient after administration of a standard epidural test dose with lidocaine and epinephrine. Acta Anaesthesiol Scand. 2003;47:1050
6. Palkar NV, Boudreaux RC, Mankad AV. Accidental total spinal block: a complication of an epidural test dose. Can J Anaesth. 1992;39:1058–60
7. Datta S, Lambert DH, Gregus J, Gissen AJ, Covino BG. Differential sensitivities of mammalian nerve fibers during pregnancy. Anesth Analg. 1983;62:1070–2
8. Abouleish EI. Postpartum tubal ligation requires more bupivacaine for spinal anesthesia than does cesarean section. Anesth Analg. 1986;65:897–900
9. James KS, McGrady E, Patrick A. Combined spinal–extradural anaesthesia for preterm and term caesarean section: is there a difference in local anaesthetic requirements? Br J Anaesth. 1997;78:498–501
10. Abraham RA, Harris AP, Maxwell LG, Kaplow S. The efficacy of 1.5% lidocaine with 7.5% dextrose and epinephrine as an epidural test dose for obstetrics. Anesthesiology. 1986;64:116–9
11. Breen TW, Shapiro T, Glass B, FosterPayne D, Oriol NE. Epidural anesthesia for labor in an ambulatory patient. Anesth Analg. 1993;77:919–24
12. Dawkins CJ. An analysis of the complications of extradural and caudal block. Anaesthesia. 1969;24:554–63
13. Guay J. The epidural test dose: a review. Anesth Analg. 2006;102:921–9
14. Hess PE, Pratt SD, Lucas TP, Miller CG, Corbett T, Oriol NE, Sarna MC. Predictors of breakthrough pain during labor epidural analgesia. Anesth Analg. 2001;93:414–8
15. Poblete B, Van Gessel EF, Gaggero G, Gamulin Z. Efficacy of three test doses to detect epidural catheter misplacement. Can J Anaesth. 1999;46:34–9
16. Cohen SE, Yeh JY, Riley ET, Vogel TM. Walking with labor epidural analgesia: the impact of bupivacaine concentration and a lidocaine–epinephrine test dose. Anesthesiology. 2000;92:387–92
17. Calimaran AL, StraussHoder TP, Wang WY, McCarthy RJ, Wong CA. The effect of epidural test dose on motor function after a combined spinal–epidural technique for labor analgesia. Anesth Analg. 2003;96:1167–72
18. Roelants F, MercierFuzier V, Lavand’homme PM. The effect of a lidocaine test dose on analgesia and mobility after an epidural combination of neostigmine and sufentanil in early labor. Anesth Analg. 2006;103:1534–9
19. Biboulet P, Capdevila X, Aubas P, Rubenovitch J, Deschodt J, d’Athis F. Causes and prediction of maldistribution during continuous spinal anesthesia with isobaric or hyperbaric bupivacaine. Anesthesiology. 1998;88:1487–94
20. Standl T, Beck H. Influence of the subarachnoid position of microcatheters on onset of analgesia and dose of plain bupivacaine 0.5% in continuous spinal anesthesia. Reg Anesth. 1994;19:231–6
21. Richardson MG, Wissler RN. Densities of dextrosefree intrathecal local anesthetics, opioids, and combinations measured at 37 degrees C. Anesth Analg. 1997;84:95–9