For some time I felt that I was a man alone in expressing reservations about the current enthusiasm for the use of lipid emulsion in the management of local anaesthetic toxicity, so I read of the concerns of Mathieu and Cranshaw  with some relief! However, some of the comments in the reply from Picard et al.  cannot go unanswered, and the letter which follows, wherein Whiteside describes  the use of lipid in the management of CNS toxicity, requires comment as well.
First though, a word about the likely mode of action of lipid emulsion. Mathieu and Cranshaw refer to the original theory that bupivacaine partitions preferentially into the lipid globules, but there is good evidence that the effect may be biochemical. Lipophilic local anaesthetics interfere with mitochondrial energy-linked processes by the uncoupling of oxidative phosphorylation [4, 5] and inhibition of respiratory chain enzymes [6, 7]. Most specifically there is inhibition of mitochondrial carnitine-acylcarnitine translocase, something which elevated concentrations of triglyceride will overcome and so restore energy production in the myocardium. Such a mechanism is quite fascinating and might explain why lipid has been effective in myocardial toxicity due to other drugs as well.
Picard et al.  claim that ‘…the Association is correcting a grave deficiency by promulgating in these guidelines the first rational, organised approach for managing an intoxication that hitherto was often fatal.’ Well, with due respect, all one has to do is look at any modern text of regional anaesthesia to find a rational, organised approach to local anaesthetic toxicity, and it is the same in each. Prevention comes first (as I have said before [8, 9]), and treatment demands, sequentially, a clear airway, lung ventilation with oxygen, control of convulsions and cardiovascular support. This must be the order of priority because failure to ensure the first three will result in hypoxia, hypercapnia and metabolic acidosis, all of which will make cardiovascular resuscitation more difficult, if not impossible, no matter what specific treatment is used.
The consequences of focusing too much on the use of lipid are well shown in the report from Whiteside . A technique involving a spinal anaesthetic, sciatic nerve and lumbar plexus blocks for an apparently routine knee arthroplasty is one which leaves me anxious on several counts, but in the current context it cannot be denied that it requires a very large dose of bupivacaine (215 mg in all). The details given make it most unlikely that there was any direct intravascular injection and I would infer that toxicity from what was an overdose for this patient is what caused her convulsions. Such things happen, and the initial resuscitatory measures were exemplary, but I am most concerned that recent attention on use of lipid distorted the subsequent approach. We are told that it was possible to obtain a second anaesthetist and administer a bolus of Intralipid® (Frenius Kabi, Warrington, UK) within 2 min ‘…while anticonvulsant medication was being prepared’ yet this patient had already been given some midazolam, and presumably more was available. In any case, whenever large doses of local anaesthetic are in use all the drugs and equipment which might be needed for the management of complications must be immediately to hand. Fortunately, the worst consequence of this event seems to have been that surgery was delayed by a week, although I do wonder how the patient was anaesthetised on that occasion.
Increasingly, it looks as though lipid emulsion is a useful addition to our armamentarium for dealing with cardiovascular collapse due to local anaesthetics (and perhaps other causes). However, its use must be placed in context, and without losing sight of the fundamental rules of regional anaesthesia.