I thank Brammar and Sharma for their comments on our recent editorial  about maximising interscalene catheter safety. We are in complete agreement with the interscalene catheter placement technique the authors propose. Interestingly, the described technique is very similar to one I developed in 2005 with outcome data from a formal prospective audit of over 300 patients published a short time later .
I would also add the following. The sonogram supplied is of good quality. Often, all that can be seen is one nerve root (occasionally, roots cannot be seen at all). For this reason, and because accurate catheter (and therefore needle) placement is critical to the technique's success, we recommend concomitant neurostimulation to confirm the target hypo-echoic structure is, in fact, a nerve root. One motor response (preferably deltoid, biceps or triceps) is all that is required to confirm the appropriate root. Single-injection techniques do not require the same needle tip precision.
Because the needle is aligned in the direction of the plexus, if using a large calibre blunt needle (e.g. 18-G Tuohy), nerve root penetration is almost certainly impossible, so we try to place the needle as close as possible to the target roots, but still within the middle scalene muscle as the authors propose .
We agree with threading the catheter several centimetres past the needle tip, but then withdrawing to 3–4 cm beyond the original needle tip position (after needle removal) . Less than 3 cm increases the risk of subcutaneous placement resulting from movement of the patient's head. More than 5 cm increases the risk of the catheter's deviating away from an optimal position.
The catheter's position can indeed be verified by observing local anaesthetic spread, but we do not do this for the following reasons: firstly, observing local anaesthetic spread via the catheter is technically challenging when using the out-of-plane technique. Secondly, as the catheter is blindly left 3–4 cm beyond a needle tip aligned parallel to the plexus, it simply cannot deviate significantly away from the plexus . Finally, it is uncertain what the operator should do if they can't see local anaesthetic spread; removing such catheters will result in the removal of many appropriately placed catheters.
The optimal local anesthetic dose for the initial catheter bolus to maximise postoperative analgesia and minimise motor block is 20 ml ropivacaine 0.375%. Local anaesthetic-induced myotoxicity has yet to be confirmed as a clinically relevant issue.
Finally, we urge caution when comparing success rates between studies, as there can be multiple confounding variables .
A detailed description of the above technique with video is available at http://ultrasoundblock.com.