I would like to commend the Editor for providing in recent issues of Anaesthesia a forum for the lively and compelling interchange of opinions on the use of lipid emulsion for treating severe local anaesthetic toxicity. I was pleased to read the initial letter from Dalgleish and Kathawaroo , which cited our work [2, 3] and sought to raise awareness of the efficacy of lipid infusion in treating bupivacaine-induced asystole in animal models of local anaesthetic toxicity. The letter  and subsequent editorial  from Picard and Meek endorsed clinical translation of this method by suggesting anaesthesiologists begin to stock lipid emulsions where regional anaesthesia is performed and by publishing a refinement of my previous recommendations for dosing. I support this position whole-heartedly for two reasons.
Firstly, I am convinced that severe systemic toxicity from local anaesthetics occurs with far greater frequency than indicated in the medical literature. Data collection in large cohort studies of regional anaesthesia is subject to errors of ascertainment, misclassification or misdiagnosis and other causes of under-reporting. I am aware of a dozen near-miss and fatal cardiac arrests involving bupivacaine, ropivacaine and even mepivacaine over the past 5 years. Only one of these was reported in a study of complications in regional anaesthesia. Many of them occurred in a private practice setting and in half the drug was administered by a non-anaesthesiologist. This suggests that the patterns and context of local anaesthetic toxicity are changing and further implies that epidemiologic studies of anaesthesia practices are unlikely to capture many of these cases.
Second is the fact of having watched a large number of rodents and dogs with large and otherwise fatal overdoses of bupivacaine rapidly regain normal haemodynamics after a bolus of lipid emulsion. Nothing provides a more convincing testimony than seeing the return of cardiac function shortly after the lipid washes through the subject's coronary circulatory bed.
The counter-point to this optimistic position was expressed in Professor Wildsmith's cautionary reply to Picard and Meek which warned of ‘their somewhat over enthusiastic advocacy of Weinberg’s untested proposals' . I join Picard and Meek in their support of Wildsmith's suggestion to remain mindful of the need to prevent anaesthetic complications . However, I am hopeful that this viewpoint will not prevent physicians from infusing lipid emulsion in a patient unresponsive to standard resuscitative measures who, for all intents and purposes, is dead.
Moore et al.  provided a balanced position by reporting that they have begun stocking Intralipid in their operating suites, while also pointing out potential complications of large dose lipid infusion. This letter underscores the need for further studies to define an optimal and safe method for lipid infusion in local anaesthetic toxicity.
Even conservative estimates link peripheral nerve block to a rate of severe, systemic toxicity that is nearly an order of magnitude greater than that for malignant hyperthermia . My hope is that lipid infusion could play the same role for local anaesthetic toxicity that dantrolene does for malignant hyperthermia. Such an ‘over-enthusiastic’ position will be far easier to defend if there is a published case report of the successful application of lipid infusion in saving a patient from severe local anaesthetic toxicity. All of us await that report.*