Toyota is famous for improving their cars through a process of continuous, small, incremental improvements, a technique known as Kaizen, or the Toyota way. In this way many small improvements, each inconsequential on their own, when added together produce significant results.
I think this is a great model to use when looking at anaesthesia. Anaesthesia and surgery are complicated processes, and most of the “low hanging fruit” in terms of safety improvements have already been made. It is unlikely that any single factor will make a major difference to outcomes. However that doesn’t mean we should stop trying to improve, and using a wide range of small improvements in different areas will collectively improve the patient’s experience.
An example of this is IV cannulation, something we all do every day and which we often forget can be quite painful. In addition, this is often patient’s only way of judging the ability of their anaesthetist. I recall a poster presentation where only 2 things determined a patients satisfaction with their anaesthetist- did the IV hurt, and did they visit the patient more than just in recovery. The patient has no baseline to judge postoperative pain or nausea and can awaken after all sorts of intraoperative near-catastrophes none the wiser, however if you want to make a patient happy, get the drip in first time and make sure it doesn’t hurt!
Two articles on this topic have got my attention and changed my practice. Both involve randomised trials where patients were either warned they were going to feel a “sharp sting” or used more neutral and comforting words, eg “I am going to apply the tourniquet on the arm. As I do this many people find the arm becomes heavy, numb and tingly. This allows the drip to be placed more comfortably”. The patients not only reported lower pain scores but were also less likely to withdraw their hand in the “kind words” group compared to the “nocebo” group. This is contrary to the common practice of warning someone, with the rational that it then won’t be as bad as they expect. In fact all this achieves is heightened anxiety and more pain (read here and here).
This is an example of avoiding the “nocebo” (opposite of placebo) effect of harsh words like sharp, sting, needle and pain. There is increasing evidence that placebo plays a major part in many interventions. Recently i went to an intriguing talk about placebo where the concept of the “open/hidden” trial was discussed. This is the opposite to a placebo controlled trial. Instead of everyone getting told they were getting morphine and half getting a sugar pill, all patients get given morphine but only half are told about it. The rest had it quietly slipped into a bag of fluid without being told. There were significantly greater reductions in pain in the group that were told they were getting the “powerful painkiller”, compared to the group that had it slipped into their fluids. The presenter gave a range of slides for different analgesics showing that for virtually all of them the pain score reductions were double in the open “powerful painkiller” group compared to the hidden ones.
Finally, three further topics for the “continuous improvement” theme, all of which i will talk about more in the future.
The first is transthoracic echo for use by anaesthetists in preoperative assessment. This is something that was big when I was doing my fellowship at the Royal Melbourne Hospital and is spreading around the world rapidly. In this month’s Anaesthesia the RMH team have provided the first (weak) evidence that preoperative echo may improve outcome, instead of simply changing management (which has been shown in previous studies). The study is observational, of poor quality, subject to the Hawthorne effect and shows an implausibly large mortality difference, but for all that makes pleasing reading for transthoracic echo exponents such as myself (reference here).
The second is the use of dexamethasone to prolong peripheral nerve blocks. This is something we have been doing recently in our hospital in Mackay in Australia, and the results can only be described as “spectacular, bordering on scary” – 24-30 hours duration from a single shot interscalene block, including complete motor block at 24 hours. This is consistent with studies showing dexamethasone effectively doubles the duration of most nerve blocks. Just remember that the phrenic nerve is also paralysed for 24 hours!
When I first read about this I had some concerns on potential neurotoxicity, but these were alleviated by 2 things. The first of these were the words of a chronic pain physician colleague who stated that they add dexamethasone to every block they do, have been doing so for years and have had no problems. The second is a study showing that in an animal model dexamethasone was significantly less neurotoxic that ropivicaine, and the ropi/dex combination was less toxic than other common combinations such as ropi / buprenorphine and ropi/ clonidine (reference here and here and here)
The final interesting note is on SCDs- the sequential calf and thigh compressors now ubiquitous on the legs of patients having surgery in our hospitals. Two recent articles showed that they reduce intraoperative hypotension- a bonus that at first seems unexpected until you think about it, then seems quite logical.
Reduced hypotension for caesarian: here
This study used a variation of the normal compressors with higher pressures and longer compression times: here
AnaesthesiaWest 
These are very provocative ideas. I had not thought to add dexamethasone to my interscalene blocks. In the EJA study patients were given 8mg of dex or saline. Dexamethasone prolonged median sensory (1457 vs. 833 min, P < 0.0001) and motor (1374 vs. 827 min, P < 0.0001) blockade compared with the control. At 24 h, the dexamethasone group had lower median verbal analogue scale scores compared with control (3.0 vs. 6.0). At 48 h, the two groups had similar median pain scores (4.0 vs. 5.0, dexamethasone vs. control, respectively). I wonder what the mechanism of action is?
Do you think we should put SCDs on all of our CS patients?
The theories i’ve seen for dex are either that it has a vasoconstrictive effect (like epinephrine), or alternatively that it increases the activity of the (inhibitory) K+ channels in nerve fibres, via glucocorticoid receptors. Increasing K+ channel activity would be similar to inhibiting Na+ channel activity.
Observationally the effect appears “local anaesthetic” like, in that you get a marked prolongation of motor block as well as sensory block. This is qualitatively different from clonidine for example (as well an quantitatively – dex is far far more effective).
My next question is whether it would be possible to combine a low concentration of local anaesthetic with dexamethasone in order to achieve a prolonged sensory block without motor block.
On that note there was an interesting recent study from Auckland where they attempted to find the lowest volume and concentration of local anaesthetic that would result in no pain in recovery in 95% of patients- the result was 20mls or 0.375%.
http://www.ncbi.nlm.nih.gov/pubmed/20460999
Thanks James that is really fascinating. I am something of a steroid sceptic, based on the history of steroids being used for pretty much everything – back in the 50s and 60s, then the list of indications dwindling over time as evidence emerged of their doubtful efficacy in one condition after the next. Indeed, I associate steroid prescriptions with physicians who have run out of ideas and can’t make a diagnosis. It is crucial that the anesthesia community provide high quality data on this and don’t let steroids creep into regional anesthesia based on a few personal observations.
Incidentally, I enjoyed your IV line discussion. I warn patients of a ‘sharp pinch’ before IV lines. This is because I was quite impressed when I needed an IV myself and I felt afterwards that anesthetists undersold the level of pain they can inflict (it wasn’t a ‘little scratch’!). However you could well be right that a warning constitutes a sort of negative placebo. I am going to try ‘positive reinforcement’ for the next few weeks!
Thanks for the letter Leo.
There is absolutely no doubt that the dexamethasone works. When you put a block in at 10am and at 8am the next day the guy has a completely dead arm than he can’t lift or feel it’s a pretty solid endpoint! It’s not a case of differing pain scores or slightly variable durations. I’ve done about 20 blocks with it now and as a rule of thumb it nearly doubles the duration of a block with 0.75% ropivicaine. It’s also a very consistent effect. Something is blocking nerve conduction longer than normal. It’s very different from clonidine for example, where you would see the patients post op and you’d be umming and ahhing about whether it made any difference. The guy can’t move his arm. The block is working. It kind of reminds me of the famous criticism of economists “sure it works in practice, but what about the theory!”. I don’t have a good explanation of why it works but there is no question it does.
On another note i forgot to mention another interesting thing that was discussed at the talk on placebo i went to recently. Naloxone blocks the analgesia provided by a placebo, suggesting an endogenous opioid mechanism for placebo analgesia. Fascinating.
Fascinated James – doing regional fellowship in Sunnybrook – What volume are you using for your interscalene blocks and why 0.75 rather than 0.5? We use ambulatory catheters but will now read literature on dex. Catherine Nix
Good question catherine. I’ve been experimenting with using 20mls of 0.375% + dex for shoulder surgery, with the aim of having less motor block, on the basis of the Auckland study i mentioned above. Previously I’ve used around 30ml of 0.75% +/- dex. In reality i’ve used this strength as this is how our ampoules come in australia! I’ve only done a few blocks so far with this new mix. It seems to work fine, and there may be less motor block (it’s a bit hard to tell with so few patients). It doesn’t appear to last quite as long as the stronger mix, based on the few blocks so far- more around 16-18 hours rather than the 20-24+ i was getting with the 0.75%. To my mind the strength you choose depends on how badly you want to avoid a motor block (+/- phrenic palsy) vs how badly you want to avoid pain. For elderly patients who may have respiratory issues i choose the weaker mix and lower volume, plus use a high supraclavicular/ low interscalene technique aiming to reduce the risk of phrenic palsy.
Let me know how you go!
James
PS a late note with Dexamethasone- there appears to be a dose response effect- 8mg prolongs the block more than 4mg