Personal Anaesthesia Practice Opinions

NOTE: These are my personal opinions. Some are supported by the literature, others by experience, personal assessment of the litrature, etc. Use them as you wish, but don't blame me if they don't work for you. Please contact me if you have some feedback!

Premed with propofol

Many patients are very anxious on arrival in the OR. Some anaesthetists use midazolam to reduce this, but I have found that a small amount of propofol works better. Make sure the propofol is mixed with lignocaine to counteract the high pH (propofol is quite basic), or it will hurt. If the dose is right the person will relax and become talkative. Given too rapidly, patients may feel dizzy. Typically 10-30 mg is all that is needed. Respiratory depression is unlikely but obstruction of the airway may occur if the patient becomes unconscious from overdose.

I've recently competed a paper on this, co-authored by Lorenzo Quario-Rondo. Here's the Pub Med link. We found a little iv propofol relieved anxiety just as well as a little midazolam, but didn't affect memory as much, and it wore off more quickly (so could be titrated easily if needed). Minerva Anestesiol. 2008 May;74(5):173-9.

Giving a little propofol prior to induction also approach gives you an idea of the patient's overall sensitivity to anaesthetic drugs as well as being cheap and simple. If you were to continue the process with repeated small boluses titrated to effect, rather than one arbitrary large dose, most patients go to sleep smoothly and comfortably, forgetting when it was that they actually went to sleep and expressing surprise, at the end, that it was all over.

If you have access to a TCI pump I later describe a simple induction technique that quite effectively identifies subsequent infusion requirements.

Neurosurgical anaesthesia

This is a brief summary of my own idisyncratic ideas. I've done two days neuro for 15 years so I've seen a lot. Every patient is different; take care to choose techniques that are most appropriate for the particular patient!

Good positioning is vital. Pad all pressure areas. Seal the eyes shut carefully. Make sure neck rotation is not so extreme as to obstruct venous drainage from the head. Use a small bite block. Secure the tube carefully and without wedging lips against teeth. Double-check all your breathing circuit and IV connections before the patient is covered by the drapes. If the head is in the pins, the rest of the body must be securely immobilised with bars and sandbags, or else when the table is rotated the neck may be subject to considerable force.

Monitor however you think is best. I put the arterial line in when awake and use peripherally inserted central venous lines to avoid the risk of jugular or subclavian approaches.

Propofol, Remifentanil and Nitrous Oxide - a great combination! 0.3-0.4 MAC sevo in place of nitrous is just as good.

Propfol (by TCI infusion) and remifentanil (by infusion) with either nitrous oxide or very low dose volatile is in my hands the best approach.

If the brain is tight (eg head injuries) nitrous shouldn't be used, but for most tumours nitrous is fine and mannitol is rarely required.

I keep CO2 between 25 and 30mmHg and check that arterial is less than 35 if concerned that the end-tidal may be lower than usual. Because propofol is not a significant cerebral vasodilator, the he brain does not swell if the blood pressure rises, so you can keep mean arterial pressure high enough to ensure good cerebral perfusion, particularly under retractors I aim for a normal arterial pressure. I will deliberately induce hypotension for cerebral aneurysms but only for the shortest possible time (using nitroprusside). I will occasionally run low-normal there is significant arterial bleeding from a vascular tumour., but the higher the better.

You can do OK with volatile agents but I find it much easier and better with propofol - and emergence is definitely nicer with propofol. I always use EEG monitoring and aim to give the absolute minimum amount of propofol that is needed. This can vary from none (just nitrous plus remi) to a target of up to 6, but typically 1-2 is all that's needed. If you don't have EEG monitoring, then read how a slow induction can identify the amount that's needed.


I always administer remifentanil by continuous infusion on a ug/kg/hr basis. Most people use ug/kg/minute, but this results in silly numbers when you set the pump and can easily result in errors by a factor of 10 if the decimal point is not right. If you use ug/kg/hr, then typically 2ug/kg/hr would suppress nocioception and respiration to a modest extent (similar to say 0.1mg/kg morhpine). This is all I find is needed for most phases of a neurosurgical case, but for the pins, opening and closing around 6-7ug/kg/hr may be needed. These are nice easy to remember numbers and are easy to program into the pump; they are kind of similar to your propofol settings also.

Remember that the time to onset of a bolus - in the brain - is 90 seconds. The blood pressure takes longer to fall (3-5 minutes). For intubation, don't wait much longer than 2 minutes; as soon as you see the blood pressure taking a dive, intubate straight away. If there's no blood pressure fall at all at 2 minutes, and the drip is running, perhaps give more and wait, but after 2 minutes the remi level in the brain is falling....

On the other hand, remi has about a 7 minute time constant after a step increase in infusion rate. Without a small bolus it can be easy to get the yo-yo's (overshoot up and down causing BP oscillation). I have a simple rule: for each step increase of 1ug/kg/hr, give a bolus of 0.1ug/kg. For example, if I was increasing the infusion from 2 to 6 ug/kg/hr, I'd give a bolus of 0.4ug/kg. In practice this works well, with a more rapid response and little overshoot.

If you really want to learm about remifentanil I strongly suggest using it during a spontaneously breathing case of some kind; then you get a good feel for its potency. It doesn't make much sense to use it only during paralysed ventilated cases! I keep my remifentail dilutions relatively weak, typically 20ug/ml, so that pumps run at a decent rate all the time (avoiding unwanted bolusing if the syringe sticks intermittently) and reducing deadspace worries.


Mild hypothermia reduces cerebral metabolism and bloodflow (reducing brain volume), enhances the action of anaesthetic agents while reducing drug elimination (so you don't need to give so much as you would if they were warmer) and may provide some degree of cerebral protection. I run temperatures in the mid 34's and rewarm with a forced air blanket.


Mannitol is usually only needed for deep lesions, aneurysms and the occasional really huge tumour, where the primary role is to reduce retractor ischaemia. Your patient will feel much better afterwards if, once the tumour and retractor is out, you start to administer an appropriate amount of dextrose. Since 500ml of 20% mannitol pulls 1000ml of free water out of the cells, 1000ml of 5% dextrose is needed. If you give this very quickly, you will get a very high BSL, so take at least an hour to do this

Paralysis isn't needed

I am accustomed to providing anaesthesia for acoustic neuroma excistion. Our surgeons require a completely unparalysed patient so that they can monitor the facial nerve. At the same time they have retractors right on the brainstem. This is scary, but having done over 500 now I've found that provided the patient is properly asleep (use an EEG monitor!) and adequately narcotised (run remi continuously!) there is little to fear. So now I use muscle relaxants only to intubate the patient, and rarely paralyse any of my neuro patients during the maintenance phase (except for aneurysm surgery and people with a big tendency to cough. Unparalysed patients don't require reversal, emerge strong and can readily maintain and control their airway as soon as they wake up. Unrecognised awareness is possible but most unlikely, and wake-up functional testing during surgery is easy. (By the way, if you do want to wake someone up during a case for functional testing make sure there is no nitrous around - not even one or two percent - or they take a long time to become co-operative.)

Emergence & recovery

Recovery within 5 minutes of the end of surgery, even after 10 hours or so, using a propofol remi combination and EEG monitoring, is not difficult. The most important issue in relation to rapid emergence is to avoid dependence on slow offset drugs during the maintenance phase. The two fastest offset anaesthetic agents that we currently have are nitrous and remi, so one should try to get as much as possible out of them, particularly towards the end. Monitor the EEG (or use BIS or Entropy etc), and continually strive to give the minimum amount of propofol while really driving the remi hard to control reflexes. Nitrous is not a popular agent but is very fast in offset, sparing the need for a lot of slower-offset agents. At the end of the case, I drop the propofol to about 1/3 of that required at the start and run it to within 3 minutes of the end, while pushing the remi up to whatever is needed to control blood pressure. Once the pins are out, everything goes off at once!

At the end of a remifentanil-dominant anaesthetic, incremental boluses of fentanil (typically 25-50ug boluses to a total of say 100-200ug) is definitely the best way to control the hyperventilation, hypertension and pain associated with withdrawal of remifentanil at the end. This is because the onset time of fentanil is very similar to the offset time of remifentanil, so fentanil can be titrated quite easily to the respiratory rate and to control the patients headache. The onset of morphine is too slow.

The hazards of coughing at the end (ie bleeding into the head) are best minimised if the intubation at the start was gentle, if ICP at the end is slightly positive, there is no air (to compress) inside the skull, and there are enough narcotics in the system to suppress coughing. To achieve this I try to return brain volume to the 'normal' state during dural closureby getting the end-tidal CO2 up to say 40mmHg. Arterial CO2 would then be say 45-50, but the brain is approxiamtely 'normal' because the brain swelling caused by the high CO2 is offset partially by the depressant effects of the anaesthetic agents and hypothermia. Make sure the surgeons fill all airspaces with saline prior to final dural closure. All this is important to ensure that the brain volume will be neutral on recovery; then if the patient coughs, the ICP will rise in parallel with the rise in venous pressure, so that the transmural gradient across intracerebral veins will not change much. Sometimes, eg after releasing a chronic subdural, surgeons may ask you to swell the brain to fill the head by articifially elevating the CO2 to very high levels. But complying is not a good idea because once the patient wakes up and they get rid of the CO2, they will end up with a very low pressure inside the head which increases the likelihood of bleeding, not reducing it. The patient is better off if you give some dextrose and ask the surgeons fill the gap with saline.

Always Reverse Neuromuscular Block SLOWLY...

Always give Neostigmine/Atropine reversal slowly (I mean truly slowly, ie over 4-5 minutes in small divided doses). This avoids acute vomiting at the end of the case, as well as unwanted tachycardia. Usual doses of Neostigmine/Atropine given as a bolus does cause acute postop vomiting (probably by inducing gut smooth musclespasm), as well as increasing the incidence of bowel anastomosis dehiscence. Ask Arthur Lam how it causes abdominal pain in conscious volunteers! However, take care doing this if you use Neostigmine/Glycopyrrholate; small doses of this have a dominant Neostigmine action and acute bradycardias are not uncommon!

Spray the vocal cords with local anaesthetic...

I almost always spray the vocal cords in an incremental fashion with 10% aqueous lignocaine hydrochloride on induction. This seems to reduce the likelyhood of coughing at the end. It can last for hours (unlike the brief duration if you just spray it into your mouth).

If you gradually topicalise the airway then the haemodynamic response to intubation will be attenuated, but to be successful one needs to be gentle, thorough and slow. If you just send the patient off to sleep, do a forceful laryngoscopy, quickly puff a few sprays down the trachea, then intubate immediately after, the haemodynamic response will not be reduced (surprise surprise!).

My suggestion is to spray the pharynx in a series of small steps, waiting between each to assess if there is any haemodynamic response (and to give time for the local to work). This divides the intubation stimulus into a series of smaller, less stimulating events. The degree of response to each small stimulus is a good guide to the adequacy of reflex suppression, ie whether or not deeper anaesthesia is needed.

To describe this in detail, first spray the tongue and whatever comes into view without any force as soon as the patient goes to sleep. Once ventilation is established and given all IV agents given, gently put the scope on the numb bit of the tongue and spray the posterior pharyngeal wall and the back of the tongue. After waiting 30 seconds or so, visualise the epiglottis as gently as possible and then spray both the aryepiglottic folds anterially and put some down the trachea (but only if it is plainly visible). I wait a good minute for this to work and then put the scope down anterior to the epglottis and do a formal laryngoscopy. By now the superior laryngeal nerves should be blocked so the laryngosopcopy itself will be a lot less stimulating. I then spray lots of local down the trachea, withdraw again, wait another minute, and intubate.

You can see that this is a slow and tedious process. I only bother if I am looking after a patient in which there is a real need to attenuate the haemodynamic response to intubation. The most common clinical stuation is a very frail or unwell patient where the administration of large amounts of anaesthetic agents would most likely to cause a lot of hypotension. For example, dehydrated patients with cerebral aneurysms or lots of cerebral oedema or elderly patients with cardiac disease.

Titrate all drugs to effect

The most important and fundamental concept in anaesthesia (next to 'give it time to work')!

Human variability is such that one should in all but extreme cases try a small dose first.

If you have access to a propofol target control pump, try starting at a plasma target value of 1 or 2 ug/ml, then increase the target so that the gradient from blood to brain is approximately 1ug/ml. Typically this involves a step up every minute or so. If no other drugs are given, most people will become unconscious at a brain concentration of 3-4 ug/ml in the brain, so it takes about 3-4 minutes. A gradual induction like this is extremely smooth and the cardiovascular system adjusts easily to the changed circumstances. I then maintain the patient at the estimated brain level at the time of loss of consciousness; this is usually a good starting point; if anything, it is usually more than is required once narcotics and nitrous are added.

Titrating narcotics is most easily performed by titration to respiratory rate during spontaneous breathing anaesthesia. If you keep all other agents constant, remifentanil can easily be titrated to blood pressure.

The Strangle Test

I use a lot of peripherally inserted central lines. If the head is laterally flexed towards the ipsilateral shoulder, and the arm abducted so that it points into the chest, they are more likely to go into the chest rather than up the neck. But they still can go up the neck. To check that they have not gone up there, I put them in 'to the hilt' or at least a long way in, so that if they were to go up the neck, they would be at least half way up. Then, keeping the transducer end sterile, I remove the stilette, hook the line up to the CVP transducer and flush it. Gentle pressure (only enough to elevate venous pressure by occluding the jugulars) transiently applied at the base of the neck should NOT make the trace go up; if it 'steps up' when pressure is applied, then it is up the neck. It is usually possible to withdraw the catheter and reinsert it into the chest. I've reviewed the postop xrays and the incidence of unexpected up the neck positioning is very low using this technique. If you want to aspirate air you should use an ECG guided system to get it into the atrium.