I was interested to read both an editorial and an article relating to the risk of neuraxial blockade in one issue of the European Journal of Anaesthesiology [1,2]. Whilst I wholeheartedly agree with the arguments and sentiments expressed in Kamming and Davies' editorial, I am afraid I cannot agree with Bogdanov and Loveland. I would suggest that it does not require a double-blind trial to recognize that the likelihood of, especially neural, damage is greater whilst needles are being inserted into an anaesthetized patient rather than one from whom immediate feedback is available.
Using the method described by Ho to estimate the incidence of rare adverse events that have not (yet) occurred , we can estimate with 95% confidence that the maximum risk of damage detectable by a surgeon in outpatients following brachial plexus block under general anaesthesia for shoulder surgery using Bogdanov and Loveland's technique is 3 in 548 or 1: 183. Hopefully, the risk to the patient is in reality, much less than this, but it does go to show how little help a small series such as this is in weighing the ‘before induction or after induction’ argument.
I would argue that it is impossible to state to the patient that it is safer to have the brachial plexus block placed after induction of anaesthesia and that the difference in risk is difficult to quantify; but it must exist.
Since the dismissal by the Lords of the appeal against the Chester vs. Afshar ruling  it is now abundantly clear that we have a duty to explain all the risks to the patient and allow the patient to make the decision as to which technique is used. This does not mean that Bogdanov and Loveland should not continue to do what they do, but more that the patient must choose to have the block put in asleep after the risks, as we know them, have been fully explained.
Department of Anaesthesia, Royal Brompton Hospital, London, UK
1. Kamming D, Davies W. Thoracic epidural analgesia for coronary artery surgery. A bridge too far? Eur J Anaesthesiol
2. Bogdanov A, Loveland R. Is there a place for interscalene block performed after
induction of general anaesthesia? Eur J Anaesthesiol
3. Ho AM, Chung DC, Joynt GM. Neuraxial blockade and haematoma in cardiac surgery: estimating the risk of a rare adverse event that has not (yet) occurred. Chest
4. Judgments - Chester (Respondent) v. Afshar (Appellant) SESSION 2003-04  UKHL 41 on appeal from:  EWCA Civ 724