A week may be a long time in politics, but in medicine changes are generally much slower. However, in at least one respect, much has changed in British anaesthesia in less than 3 years.

An editorial published in this journal in February 2006 considered the use of intravenous lipid emulsion to treat overwhelming overdoses of local anaesthetic drugs [1]. The editorialists outlined the persuasive evidence from experiments on animals, and argued that local anaesthetic intoxication is so very rare, so catastrophic and so unpredictable that ethically acceptable, controlled human trials are, and always will be, impossible. Their conclusion was that, if a patient was in refractory cardiac arrest attributable to local anaesthetic intoxication, there would be little to lose from giving intravenous lipid, and potentially everything to gain.

Much has happened since then. First, lipid’s successful use in humans suffering cardiac arrest attributable to local anaesthetic intoxication has been reported in this journal and in other peer-reviewed publications [2–5]. More reports have also appeared on a website dedicated to the use of intravenous lipid to treat drug overdose ( Those patients treated in this way have typically had their cardiac output restored within a few seconds of the administration of the bolus of lipid. If that was not good enough, there have so far been no reports of lipid either failing or being associated with adverse side effects.

Given these successes, some clinicians felt it reasonable to try giving lipid intravenously to patients intoxicated by local anaesthetic but still short of having developed refractory cardiac arrest. The first such report described lipid’s success in reversing an adult patient’s precipitate cardiovascular collapse after injection of levobupivacaine 100 mg [6]. Other authors have described the beneficial effects of lipid on cardiac arrhythmia attributed to local anaesthetic toxicity [7, 8]. Arrhythmias such as ventricular tachycardia and bigeminy provoked by drugs such as lidocaine, ropivacaine, mepivacaine and prilocaine have been successfully converted to sinus rhythm. Lipid has also apparently quickly reversed agitation and loss of consciousness attributable to racemic bupivacaine intoxication [9].

There have been developments in committee rooms too. After the publication of the first few reports, the Council of the Association of Anaesthetists of Great Britain and Ireland established a working party to consider the production of a guidance document on the use of lipid to treat local anaesthetic intoxication. As a result, local anaesthetic overdose joined the fearsome company of Anaphylaxis and Malignant Hyperpyrexia as the subject of a laminated, one-page guidance document [10]. Given the limited information available at the time, its recommendations were inevitably blunted by compromise. For example, it recommended simply that treatment with lipid be ‘considered’. However, it firmly recommended that a litre of 20% lipid emulsion be kept immediately available for all patients receiving potentially cardiotoxic doses of local anaesthetic.

In the ensuing 3 years, treatment with lipid emulsion, or at least knowledge of the potential value of treatment with lipid emulsion, has swept the nation. Writing in this journal, Williamson et al. set out how far the tide had spread through obstetric departments by the first half of 2007 [11]. A study reported in the current issue describes the spread of lipid therapy in more detail [12]. It shows just how extraordinarily quickly lipid has been adopted in 66 NHS hospitals in and around London. In late 2006, just 18% of the hospitals held lipid available. A year later the position was reversed; just 23% of the sampled hospitals did not have lipid or did not provide data. The study underscores some hard truths about guidelines, our colleagues and the adoption of innovation. About half of the sampled hospitals already had lipid available when the laminates arrived. Perhaps it is inevitable that, when national authorities prepare guidance, it takes time, during which their recommendations may be widely implemented. However, the arrival of the guidelines prompted the adoption of lipid in fully a quarter of hospitals. So the guideline’s delicate drafting and expensive distribution were far from futile. Even so, at the end of 2007, one-eighth of hospitals had not yet made lipid available. Posterity may vindicate such conservatism; only time will tell.

Widespread acceptance of lipid as an antidote to local anaesthetic poisoning is not the end of the matter: it is just the end of the beginning. There remain many details to resolve. It is unclear, e.g. how resuscitation with lipid should be combined with other drugs, and with adrenaline in particular. In a rodent model of bupivacaine overdose, lipid treatment combined with cardiopulmonary resuscitation was better than adrenaline with identical cardiopulmonary resuscitation [13], as the adrenaline tended to cause pulmonary oedema and ventricular arrhythmia. Meanwhile, in the more recent protocols, adrenaline is to be given first some minutes after resuscitation has begun. If patients respond quickly to a bolus of lipid, and get the fat as resuscitation begins, they may never receive adrenaline. Would that be for the good? We do not know.

We also still do not know how lipid really works in this context. There is gathering evidence that the metabolism of the cardiac myocyte is relevant. For example, bupivacaine is known to inhibit mitochondrial function [14]. Isolated hearts are more susceptible to bupivacaine toxicity if they are running (as ours normally do) on fatty acids rather than carbohydrate [15]. Does this mean that the local anaesthetic inhibits fatty acid metabolism in a ways that makes resuscitation difficult and that can be reversed by lipid? This theory is not proven; it may still be working simply as a circulating lipid ‘sink’ that draws local anaesthetic unto and into itself, without any direct metabolic influence.

If the detail of the use of lipid in local anaesthetic intoxication is unclear, the overall story has advanced far in just three years. However, it seems that an extraordinary new chapter in the use of lipid is beginning. Lipid has been reported to be useful in animal models of some other forms of intoxication [16–18]. It has now been used in human patients threatened by other poisons. First, it was given in a desperate attempt to restore stable cardiac output [19] in a 17-year-old woman who had been admitted to an intensive care unit after having taken enormous doses of lamotrigine and buproprion. She suffered a cardiovascular collapse, with ventricular tachycardia, ventricular fibrillation and eventually pulseless electrical activity. A protracted resuscitation followed; in a period of 90 min she had an effective cardiac output for just 17 min. Direct current shocks, catecholamines, amiodarone, calcium, magnesium and sodium bicarbonate were all tried but none of them seemed to work at the time, although sceptics might argue that any of them may have had a delayed effect. Fifty-two minutes after she last had a pulse, the team decided to try intravenous lipid. One minute later, a sustained palpable pulse returned. The patient was eventually discharged home lucid and talkative, with only very mild motor and memory deficits. This is arguably a study in which = 1, but it is without doubt an impressive ‘1’. The era of ether anaesthesia was ushered in by a study in which n = 1 and perhaps, like ether, the use of lipid in this context ‘is no humbug’.

In this issue of Anaesthesia, Finn et al. describe using lipid with a different therapeutic intent [20]. They faced a challenge that is painfully familiar to clinicians throughout Britain and beyond. A comatose patient was brought to the hospital. No intensive care beds were available. A transfer to another hospital seemed inevitable, endangering the patient and using up scarce, skilled resources. Finn’s team gave intravenous lipid, and the patient’s level of consciousness quickly increased sufficiently such that the patient had no need of tracheal intubation, intensive care or transfer.

These newly reported uses of lipid present clinicians with challenging ethical questions. The use of lipid in local anaesthetic intoxication was first recommended because controlled trials were impossible, so it was futile to wait for them, and because if a patient were in refractory cardiac arrest, little further harm could be done. Some clinicians may feel that the same logic applies after protracted efforts to restore cardiac output in patients overwhelmed by other poisons. Certainly, if the alternative is to abandon the resuscitation, the use of intravenous lipid could arguably be mandated. Would a controlled study of this clinical scenario be possible? In principle perhaps; however, in practice it would or will be difficult. While out-of-hospital cardiac arrests are relatively common, it is unusual for clinicians to know during resuscitation if poisons have been taken, and rarer still to know their identity. So recruitment to any trial would inevitably include many patients who had not ingested poisons. This confounding variable would mean that a large and protracted trial would be necessary to detect any difference in outcome in patients’ survival.

Given that lipid has been reported to be used in this context only once, some clinicians may prefer to consult their local Clinical Ethics Committees, and consider giving a dose of lipid to all patients in refractory cardiac arrest potentially attributable to intoxication. It would then be important to report all such uses of lipid to a national reporting system that could collate the reports and produce a balanced view on the utility of the treatment. In the absence of such a reporting system, we would suggest that such reports are submitted to

The argument for giving a bolus of lipid to a patient admitted to hospital in coma is weaker. Here there is a theoretical possibility of the potential benefit being outweighed by the possible harm – the patient in this situation would have a viable cardiac rhythm and would therefore be still very much alive and salvageable by other, more proven methods. That said, experience of giving boluses of lipid is gathering, and no side effects have been reported so far. In time, lipid may become a conventional ingredient of the ‘coma cocktail’: naloxone, glucose and thiamine are there because they may do good, and have not yet done harm [21]. Again, meaningful studies of the use of lipid in this context may be nigh on impossible to perform, and we would advise clinicians to engage with their Ethics Committees before adding lipid to their local ‘coma cocktails’.

Whatever the future of lipid therapy, its progress so far reveals a particular characteristic of British anaesthesia. During the last century, it was common to lament the fate of British inventions: we might have the bright ideas, but they would be brought to market by other, far-off countries. Here the reverse has happened. Much of the relevant scientific work has been done in the US and also in New Zealand. However, in promoting lipid therapy for intoxication, the UK is ahead. Lipid is now widely available to patients being given potentially cardiotoxic doses of local anaesthetics. The Association of Anaesthetists’ guidelines were published in August 2007, and similar guidelines based on these are now being developed in other countries. Could this reflect the flexibility and foresight of Anaesthesia’s editorial strategy and the agility of the Association of Anaesthetists? If so, we should all be proud.

Declaration of interests

  1. Top of page
  2. Declaration of interests
  3. References

Dr Picard contributed to the initial development of the website. It has no revenue and he has no financial interest in its past, present or future.


  1. Top of page
  2. Declaration of interests
  3. References