A reply

Drs Pattnaik and Bodra raise some important points. We addressed all of these in our paper, but I am grateful for the opportunity to repeat them. There is a common misconception that instituting epidural blockade for elective Caesarean section is the same as topping up an existing and functioning epidural for emergency Caesarean section. All three studies to which Pattnaik and Bodra refer studied elective cases; the mean (SD) onset times (time to T6) reported for these studies' lidocaine groups were 24.4 (7.6) min, 22.4 (8.8) min and 29.5 (8.7) min. In our study, median (IQR) time to a higher dermatomal level (T4) with lidocaine was 10 (9–18.5) min. Clearly, a mother who already has an epidural block, even a low-dose one, behaves differently to one with no block established. Whilst perhaps pharmacologically simplistic, the analogy of painting a wooden fence (the second coat is much easier and faster than the first one applied directly to the wood) seems to be appropriate. As Pattnaik and Bodra state, it is common belief that lidocaine is faster than bupivacaine when used for topping up an existing epidural. An anasthetist with such a belief might give lidocaine and, observing a rapid onset of block, assume that the reason for this rapidity is the choice of drug. Conversely, if the block is slow to develop, such an anasthetist might think ‘good thing I used lidocaine – if I’d used bupivacaine it would have been even slower'. We all know that beliefs persist in spite of evidence, and no doubt there will be many who continue to believe that lidocaine is the better drug to use for emergency Caesarean section. There is no evidence that this is so. Ours is the only study of which we are aware that compared these solutions for this purpose, and our study does not support the use of lidocaine, or the protocols to which Pattnaik and Bodra refer. Apart from no significant shortening in onset time, we found an increased incidence of high blocks and requirement for conversion to general anasthesia in the lidocaine group, although admittedly the numbers were small. Regarding power and sample size, we found a difference of 4 min in median time to T4 between bupivacaine and 2% lidocaine with epinephrine, but the range of onset times (biggest in the lidocaine groups, ranging from 6 min to 38 min) was so great that this difference was not significant. In order to have an 80% chance of demonstrating a significant difference of this magnitude and with the variability we found would require over 1000 patients. In addition, even if this were a real difference, we have recently shown that preparing a fresh lidocaine and epinephrine solution adds up to 2 min to the preparation time compared with bupivacaine alone (Lucas DN, Borra PJ, Yentis SM. Epidural top-up solutions for emergency Caesarean section: a comparison of preparation times. British Journal of Anasthesia (in press)). As far as bolus administration of local anasthetic is concerned, Pattnaik and Bodra's concerns are valid – accidental intravenous or intrathecal injection of a large volume of concentrated solution is hazardous. However, we are dealing with emergency cases in which a delay of even 5 min (the usually suggested interval between a test dose and the main dose) could be crucial for the fetus. Both intravenous and intrathecal migration of an epidural catheter have been reported but this is very rare; in Paech's large recent study there were three cases of central nervous system toxicity (none of which had convulsions) associated with top-ups through an epidural catheter out of almost 11 000 obstetric epidurals [1]. There was one case of high block requiring tracheal intubation following a top-up through a previously used catheter. Our cases will have had regular top-ups during the preceding labour, so the catheter is far from untested. The risk to which Pattnaik and Bodra refer is therefore migration of the catheter occurring between the last epidural top-up in labour and the 20 ml top-up for Caesarean section. This has to be considerably rarer than the published figures which refer to the incidence throughout the whole labour. If the catheter is subdural, a large bolus may rupture the subarachnoid and lead to an accidental spinal block [2]; however, the catheters in our unit have all been repeatedly tested with 10–15 ml boluses of low-dose solution throughout labour and so this too is unlikely to have escaped attention. If one is unwilling to inject a large bolus down the epidural catheter for fear of complications and time is short, the only other options are (i) disregard the working epidural and perform a spinal; or (ii) use general anasthesia. Neither option is entirely free of risk.