Epidural test dose with levobupivacaine and ropivacaine: determination of ED50 motor block after spinal administration

M. Camorcia*,1, G. Capogna1, G. Lyons2 and M. Columb3

1 Department of Anaesthesia, Clinica Città di Roma, Roma, Italy. 2 Department of Anaesthesia, St James’ University Hospital, Leeds, UK. 3 Department of Anaesthesia and Intensive Care, South Manchester University Hospital, Wythenshawe, UK

*Corresponding author: Department of Anaesthesia Città di Roma Hospital Via Maidalchini 20, 00152 Roma, Italy. E-mail: michela_camorcia{at}yahoo.it

Accepted for publication: January 30, 2004


    Abstract
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Background. When a test is required to detect a possible intrathecal catheter, many would seek to use the same local anaesthetic as that used for epidural analgesia. The rapid onset of inappropriate motor block after a local anaesthetic administered epidurally implies intrathecal spread. Because of claims of greater sensory–motor separation, or because of reduced potency compared with bupivacaine, the efficacy of the new local anaesthetics in intrathecal testing has been questioned. The aim of this study was to establish the feasibility of a test dose for an inadvertent intrathecal catheter using ropivacaine and levobupivacaine, and to establish the dose required.

Methods. Sixty women undergoing elective Caesarean section with a combined spinal– epidural technique were enrolled into this prospective, double-blind sequential allocation study. The women were randomized to receive plain levobupivacaine 0.5% or ropivacaine 0.5% intrathecally. The dose was determined according to up–down sequential allocation. The end-point was any evidence of lower limb motor block within 5 min of injection.

Results. The ED50 motor block at 5 min was 4.8 mg (95% CI, 4.49, 5.28) for levobupivacaine and 5.9 mg (95% CI, 4.82, 6.98) for ropivacaine (95% CI difference, 0.052, 1.98) (P=0.04). The estimated ED95 motor block was 5.9 mg (95% CI 5.19, 6.71) for levobupivacaine and 8.3 mg (95% CI, 6.30, 10.44) for ropivacaine. The potency ratio between the two drugs was 0.83 (95% CI, 0.69, 0.99).

Conclusions. Both local anaesthetics produce evidence of motor block within 5 min of intrathecal injection and could serve as tests of intrathecal administration. Derived ED95 values suggest 10 mg doses should be effective, but this study did not measure predictive value. Ropivacaine is less potent for motor block than levobupivacaine by a factor of 0.83 (P<0.04).

Br J Anaesth 2004; 92: 850–3

Keywords: anaesthetics local, levobupivacaine; anaesthetics local, ropivacaine; analgesia, obstetric; analgesic techniques, epidural


    Introduction
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Each year in the UK approximately 170 000 women receive epidural analgesia. Of these, approximately 0.8% will experience unintentional dural puncture and subarachnoid catheter placement1 and approximately 0.2% will receive subarachnoid injection of local anaesthetic intended for the epidural space.2 Unexpected high blocks are a cause of maternal morbidity and have been implicated in mortality.3 4 Despite the absence of evidence for the use of a test dose to identify an undetected subarachnoid catheter, many use a small dose of epidural local anaesthetic for this purpose. An ideal test dose should allow early identification of epidural catheter misplacement rapidly but without a high block. An easy and reliable clinical sign, such as the occurrence of motor block within 5 min, will provide clear evidence of subarachnoid injection.5

The introduction of two new local anaesthetics into clinical practice leads us to revisit some old questions and consider some new ones. The issue of dose arises, and is complicated by the possibility of potency differences between ropivacaine, levobupivacaine and racemic bupivacaine, together with variations in sensory–motor separation.

The aim of this study was to determine the ED50 and estimate the ED95 for motor block, for spinal levobupivacaine 0.5% w/v and ropivacaine 0.5% w/v and to consider their place as tests of inadvertent intrathecal injection.


    Methods
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
After local institutional ethical approval and written informed consent, 60 parturients, ASA I or II, at more than 37 weeks of gestation, undergoing elective Caesarean section with a combined spinal–epidural technique were enrolled in this prospective, double-blind sequential allocation study. Subjects were randomized into two groups to receive either levobupivacaine or ropivacaine.

All women received Ringer’s lactate solution 700 ml i.v. before the block. The epidural space was located using a 16 gauge Tuohy needle and the loss of resistance to saline technique, at the L2–L3 or L3–L4 interspace with the parturient in the left lateral position. After lumbar puncture with a 26 gauge pencil-point needle passed through the Tuohy needle and the aspiration of cerebrospinal fluid, the study solution was injected into the subarachnoid space. The spinal needle was then withdrawn and an epidural catheter was threaded through the lumen of the Tuohy needle, which was withdrawn. The parturients were then placed supine, and the uterus was displaced to the left with a wedge under the right hip.

The intrathecal solutions were either the commercially available form of plain levobupivacaine 0.5% w/v (Chirocaina, Abbott) or plain ropivacaine 1.0% w/v (Naropina, Astra) diluted with 0.9% w/v saline to achieve a final concentration of 0.5% w/v. The dose administered was determined by the response of the previous parturient to the higher or lower dose, according to the up–down sequential allocation.6 7 The exception to this was the first woman in each group, for whom the starting dose was arbitrarily chosen to be 4 mg for both drugs. The dose interval was 1 mg. The study solutions were made up using a finely graduated 1 ml syringe by an anaesthetist who took no further part in the study. The anaesthetist performing the procedure and subsequent assessment was blinded to the concentration used and group allocation.

The end-point was the occurrence of any motor block in either lower limb within 5 min after the subarachnoid injection of the study solution. Time 0 was considered to be the end of the injection of the spinal local anaesthetic.

Motor block was assessed every minute for 5 min using the Bromage scale to evaluate motor function of the knees and feet, and an additional scale to evaluate hip flexion.

The Bromage scale was graded as follows: 0=fully able to flex knees and feet; 1=just able to move knees, able to move feet; 2=unable to move knees, able to move feet only; 3=unable to move knees and feet.

The hip motor function scale was graded as follows: 0=complete ability to raise straight legs (>30°); 1=partial ability to raise straight legs (<30°); 2=inability to raise straight legs.

There were two possible outcomes. (i) Effective: this required a motor block score >0 in any of the two scales in any leg within 5 min, and directed a 1 mg decrement of the study drug for the next parturient assigned to that group. (ii)Ineffective: this required a motor block score=0 in both scales within 5 min, and directed a 1 mg increment of the study drug for the next parturient assigned to that group.

After the completion of the study we also evaluated the occurrence of any motor block at 10 min. After the 10 min motor block assessment, all parturients received incremental epidural boluses of 2% w/v lidocaine with epinephrine 1:200 000, and sodium bicarbonate 1 mEq to achieve a satisfactory anaesthetic level for Caesarean section.

Statistical analysis
Patient and obstetric data were collected and are presented as mean (SD). Analyses included Student’s t-test, the Mann–Whitney U-test and Fisher’s exact test as appropriate. The up–down sequences were analysed using the independent pairs method and by Wilcoxon and Litchfield probit regression of the proportional responses at each dose level as a backup analysis to determine the median effective doses (ED50) for motor block with 95% confidence intervals (CI). Analyses were carried out using the following software: Microsoft Excel 2000 (Microsoft, Redmond, VA, USA), Number Crunching Statistical System (NCSS) 2000 (NCSS, Kaysville, UT, USA) and GraphPad Instat 3.05 (GraphPad Software, San Diego, CA, USA). Statistical significance was defined for an overall {alpha} error at the 0.05 level. All P-values were two-sided.


    Results
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Personal and obstetric data were similar in the two groups (Table 1). The sequences of effective and ineffective outcomes are shown in Figure 1.


View this table:
[in this window]
[in a new window]
 
Table 1 Patient and obstetric characteristics. Data are mean (range) for age, or mean (SD)
 


View larger version (16K):
[in this window]
[in a new window]
 
Fig 1 Sequences of effective and ineffective analgesia.

 
The ED50 motor block at 5 min was 5.9 mg (95% CI, 4.82, 6.98) for ropivacaine and 4.8 mg for levobupivacaine (95% CI, 4.49, 5.28) using up–down independent pairs. The ED50 of ropivacaine was significantly greater than that of levobupivacaine (95% CI difference, 0.052, 1.98) (P=0.04). The results of probit regression, performed as backup, are shown in Table 2. The relative motor blocking potency ratio of ropivacaine:levobupivacaine was 0.83 (95% CI, 0.69, 0.99).


View this table:
[in this window]
[in a new window]
 
Table 2 ED50 motor block at 5 min (mg). Student’s t-test, P=0.04; 95% CI difference, 0.052, 1.98
 
The ED95 motor block estimates were 8.3 mg (95% CI, 6.30, 10.44) for ropivacaine and 5.9 mg (95% CI, 5.19, 6.71) for levobupivacaine. No parturient with an ineffective outcome (no motor block) at 5 min developed any sign of motor block at the 10-min evaluation.


    Discussion
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
We showed that the median effective intrathecal doses (ED50) that produce clinically detectable motor block were 4.8 and 5.9 mg for levobupivacaine and ropivacaine respectively. From these figures and the standard deviation, the ED95 motor block was estimated to be 5.9 mg for levobupivacaine and 8.3 mg for ropivacaine. In clinical practice, test doses are rounded up to a convenient volume, generally between 1 and 3 ml. With regard to ropivacaine 0.5% w/v, a volume of 2 ml, or 10 mg, approximates to the upper 95% confidence interval for the ED95 and for levobupivacaine 0.5% w/v, 2 ml or 10 mg, exceeds it by a wide margin. Less than this might be inadvisable.

Doses are determined at the 50th centile because this is the part of the dose–response curve that shows the greatest response for small variations in dose. Consequently, estimates made at this point on the slope have a fair degree of precision. Doses estimated higher up the curve are likely to be imprecise for the same reason, and confidence intervals can widen considerably. One advantage of sequential allocation is that the precise nature of the median dose determination allows a satisfactory result with a relatively small sample size. Most sequential allocation dose-finding studies conducted on obstetric patients have used epidural analgesic drugs, and these studies have required 30–40 subjects in each arm.8 Where variability is very small it is possible to achieve satisfactory results with as few as 10 subjects in each arm.9 The sample size of this study was in keeping with expectations of significant variability that did not materialize, and as a result the power is given as >1. Consequently, the 95% CI for the ED95 is sufficiently small to make an estimate worthwhile. The derivation of the ED95 is less certain than the measurement of the ED50, and this should be taken into account when dosage recommendations are made for clinical practice.

Motor block was chosen because it is easily demonstrable and less subjective than a purely sensory end-point. The presence of motor block has been reported to be an easy and reliable clinical sign, producing clear evidence of subarachnoid injection within a few minutes.5 In our study, in addition to the Bromage scale, we graduated the ability to raise the extended leg on a three-point ordinal scale because this has been reported to be a more sensitive means of testing for motor block than the sole use of the Bromage scale.10 According to the up-down method, we used a binary response for the outcome (block–no block) by using two different scales in order to detect subtle evidence for motor block.

Ngan Kee and colleagues11 confirm this view in their study of 36 women at Caesarean section. Sensory blockade did not discriminate between intrathecal and epidural administration, but they found that 8 min was needed for definitive motor block to develop. Speed of onset of block is related to concentration, which could explain earlier motor block after ropivacaine 1.0% w/v compared with 0.75% w/v, used by Ngan Kee, and 0.5% w/v used by Owen and colleagues in non-pregnant patients.12 Allowing 5 min for a possible motor block to develop has its origins in tradition, and is governed by patience rather than science.

Differences in potency also have a bearing on the choice of dose. For practical purposes, no change in regime should be required when levobupivacaine is used as an alternative to racemic bupivacaine,13 14 but a 40% increase in dose may be required when ropivacaine is used.15 16 The 17% difference in potency in favour of levobupivacaine found here does not give this numerical support, but does support a potency hierarchy of bupivacaine>levobupivacaine>ropivacaine. Dose adjustment for ropivacaine is of particular importance if test doses are to be clinically useful. The expressed formulations of the pharmaceutical preparations in use are not equimolar as there is 7% more free base with levobupivacaine compared with ropivacaine when molar concentrations are considered.

Another reason why new local anaesthetic test dose might fail is the increased separation of A{delta} and A{alpha} block compared with racemic bupivacaine.17 Claims that ropivacaine has greater motor-sparing effect can also be explained by the potency difference. In support of this, Lacassie and colleagues have shown that the difference in motor blocking potency between ropivacaine and racemic bupivacaine accords closely with the difference in analgesic efficacy.16 Our evidence suggests that if doses of ropivacaine are increased to reflect the potency difference, and the concentration is sufficient to provide a 5-min response, then it can be made to work as a test of catheter placement.

In summary, ropivacaine 0.5% w/v at a dose of 10 mg approximates closely to the upper 95% confidence interval for the ED95 required for producing motor block within 5 min, and levobupivacaine 0.5% w/v at a dose of 10 mg exceeds the upper 95% confidence limit for the ED95. While the ED95 might serve as a broad guide to dose, it should be recognized that this is a derived value. The difference between the median doses of the two local anaesthetics reflects their relative potencies. Apparent differences in sensory–motor separation can be explained by the potency difference, and should not prevent ropivacaine or levobupivacaine being used in tests of intrathecal placement. Because partial misplacement of a catheter can give misleading results, a negative test dose cannot guarantee that an unexpected high block will not occur. It is not suggested that ropivacaine or levobupivacaine is suitable for testing accidental venous placement.


    References
 Top
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Gleeson C, Reynolds F. Accidental dural puncture in UK obstetric practice. Int J Obst Anesth 1998; 7: 242–6[CrossRef][ISI]

2 Dawkins CJM. An analysis of the complications of extradural and caudal block. Anaesthesia 1969; 24: 554–63[ISI][Medline]

3 Yentis SM, Dob DP. High regional block—the failed intubation of the new millennium? Int J Obstet Anesth 2001; 10: 159–61[CrossRef][ISI]

4 National Institute for Clinical Excellence. Why Mothers Die. Confidential Enquiries into Maternal Deaths in the United Kingdom 1997–1999. London: RCOG Press, 2001

5 Poblete B, Van Gessel EF, Gaggero G, Gamulin Z. Efficacy of three test doses to detect epidural catheter misplacement. Can J Anesth 1999; 46: 34–9[Abstract]

6 Dixon WJ, Massey FJ. Introduction to Statistical Analysis, 4th edn. New York: McGraw-Hill, 1983; 428–39

7 Armitage P, Berry G. Statistical Methods in Medical Research. Oxford: Blackwell Scientific Publications, 1994; 136–41

8 Lyons G, Columb M O, Hawthorne L, Dresner M. Epidural pain relief in labour: bupivacaine sparing by epidural fentanyl is dose dependent. Br J Anaesth 1997; 78: 493–7[Abstract/Free Full Text]

9 Hodgson PS, Liu SS, Gras TW. Does epidural anaesthesia have general anaesthetic effects? a prospective randomised double blinded placebo controlled trial. Br J Anaesth 1999; 82: A342

10 Graham AC, McClure JH. Quantitative assessment of motor block in labouring women receiving epidural analgesia. Anaesthesia 2001; 56; 447–484[CrossRef][ISI][Medline]

11 Ngan Kee WD, Khaw KS, Lee B. The limitations of ropivacaine with epinephrine as an epidural test dose in parturients. Anesth Analg 2001; 92: 1529–31[Free Full Text]

12 Owen MD, Gautier PE, Hood DD. Ropivacaine is unreliable for use as a spinal test dose. Anesthesiology 2001; 94: A64[CrossRef]

13 Lyons G, Columb MO, Wilson RC, Johnson RV. Epidural pain relief in labour: relative potencies of racemic bupivacaine and levobupivacaine. Br J Anaesth 1998; 81: 899–901[Abstract/Free Full Text]

14 Lacassie HJ, Columb MO. Relative motor blocking potencies of bupivacaine and levobupivacaine in labour. Eur J Anaesth 2002; 19: A552

15 Capogna G, Celleno D, Fusco P, Lyons G, Columb MO. Relative potencies of bupivacaine and ropivacaine for analgesia in labour. Br J Anaesth 1999; 82: 371–3[Abstract/Free Full Text]

16 Lacassie HJ, Columb MO, Lacassie HP, Lantadilla RA. The relative motor blocking of epidural bupivacaine and ropivacaine in labor. Anesth Analg 2002; 95: 204–8[Abstract/Free Full Text]

17 McClure HA. Ropivacaine. Br J Anaesth 1996; 76: 300–7[Free Full Text]