What they don’t teach you in medical school about pain-killers

A little under three years ago I fractured my T12 vertebra while squatting in the gym with a 70Kg barbell on my shoulders.  At the time I didn’t know that the bones had been weakened by myeloma (a form of bone marrow cancer).  It was remarkably painful.  They gave me some paracetamol (personal trainers are allowed to prescribe that) and eventually a paramedic arrived, who had a little blue cylinder containing Entonox.  This is a mixture of oxygen (so that you can breathe it without asphyxiating) and nitrous oxide (laughing gas; not as far as I know one of the nitrogen oxides that you will be taxed on if you drive a diesel car).  It has the effect of numbing the pain, and perception generally, though I didn’t find being lifted from the floor into a wheelchair particularly amusing.  The paramedic wheeled me all the way to the hospital (which wasn’t very far) and I remember thinking that he could have done with a few gym sessions himself.

I was seen fairly quickly by the Casualty Consultant (one of the perks of being staff) and he immediately gave me double doses of morphine and diazepam (Valium).  I was greatly relieved to find myself being treated by somebody who realised what was actually required and I certainly felt a lot better for it.  I remained in hospital for a week on bed-rest (I don’t know why they call it that, as it is difficult to rest when nurses are taking your blood pressure day and night every four hours, and the kitchen staff serve meals in between at very unnatural times; maybe they think we are all navvies working the early shift).

Each time the nurses came they asked me about my pain.  Specifically how bad it was on a scale of 1 to 10.  Obviously 1 was a very mild pain and 10 was quite a bad one but I found myself racking my brains as to exactly what number to report, eventually devising a detailed and unambiguous scale with 1 being pain that you really had to concentrate in order to feel, 3 getting quite intrusive and anything more than 5 indicating that more pain-killers were required.  9 was vomiting with pain and 10 was loss of consciousness.

Level 10+

I have actually witnessed someone pass out with pain, and sad to say I was the one who inflicted it.  They had a pleural effusion, which is fluid in the chest cavity, surrounding and compressing the lung, in this case due to a tumour.  I was the junior doctor draining the fluid out with a hollow needle in the chest, and I was to inject tetracycline (more commonly used as an antibiotic) through the same needle in order to achieve what is usually called pleurodesis – sticking together the parietal (inside of the chest wall) and visceral (surface of the lung) pleura (membrane lining the chest).  I have always felt that it should be pleuradhesis, since it made the pleura adhere; pleurodesis ought to mean writing an ode about the inside of the chest excluding the heart – a sort of anti-love poem, perhaps, but doctors are no longer required to have a proper Classical education before being admitted to medical school and can’t spell.

After removing a pint or two of straw-coloured fluid I injected the tetracycline as instructed, whereupon the patient screamed and then collapsed.  I had previously only used bleomycin (a somewhat odd chemotherapy drug) for this purpose and nobody had warned me that tetracycline could be painful.  I now know to mix it with local anaesthetic, or better still, to refer to a chest surgeon who will fill the cavity with talcum powder, which works much better.

Pain scales

After dutifully reporting my pain level several times a day it occurred to me that they weren’t actually interested in what the numbers were, only whether they were going up or down.  In fact using a scale such as this, or else a visual analogue scale, which involves marking the level of pain on a line, is quite a reliable way to assess how pain is responding.  Much more reliable than just asking the patient how they feel; generally they don’t seem to have a clue.  You might wonder at this, but we are all very bad at remembering pain.  This is probably just as well, otherwise no mother would agree to bear a second child.  A patient will tell you about the pain he has now, but if it was half an hour ago then he might deny having felt any at all.  As for last week, forget it.  Thus the importance of pain diaries.

Another way of assessing the severity of the pain is to ask how much it is interfering with what are referred to as the Activities of Daily Living (i.e. sitting up, standing, talking, eating, getting dressed, going to the loo etc.).  The patient might remember that they needed help getting up, or that they couldn’t walk, and hopefully there might even have been somebody else at home (or on the ward) watching, who can actually tell you what they did, and what they didn’t do because it hurt too much.

Stairway to Heaven

Pain specialists often talk of the “analgesic ladder”, with mild pain-killers such as paracetamol on the bottom rung (acetaminophen across the pond, tairenoru in Japan), paracetamol / codeine mixtures on the next rung, possibly another rung containing relatively mild opiates such as tramadol, and finally opioids on the top.  The idea is that you ascend the ladder until you find something that works.

In addition there are drugs such as NSAID’s (non-steroidal anti-inflammatory drugs) which can be added to the simple analgesics, and depending on the type of pain they can be very effective.  Finally there are pain-modulating drugs which don’t do much on their own but can significantly increase the analgesic action of strong opioids.  These tend to be either anticonvulsants (carbamazepine, pregabalin, gabapentin) or antidepressants (amitriptyline).  These are especially helpful when there is an element of nerve compression, invasion or irritation which otherwise doesn’t respond well to most analgesics and has a particularly distressing quality.

Sauce for the goose…

When we learn about prescribing, we tend to regard the effect (and the required dosage) as being much the same in everybody.  But just consider the variable effects of alcohol at the family Christmas dinner.  One aunt will be quite tiddly after her annual glass of sherry and another will be putting away the gin and tonics.  Most drugs are tested in young adult males and the effects extrapolated to other groups (paediatricians have a particularly hard time here as they are seldom tested in children.  But even something as familiar as paracetamol can have inconsistent effects. 

Personally I have never found paracetamol to be very helpful for pain.  My wife, on the other hand, finds it works very well, even when the pain is severe.  I read somewhere that women in general find it more effective than men (I’m sorry, I can’t find the reference to check this).  But although it doesn’t do much for me as an analgesic, it is still a brilliant pyrolytic (i.e. bringing down a fever and easing the associated headache and muscle pains).

It is quite clear that not everyone experiences pain in the same way, and some gene variants have been identified which seem to play a part here.

Timing is everything

Another thing that can affect drugs is timing.  We have several internal clocks preparing our bodies (and brains) so that we are in the optimal state of readiness when it is time to eat, sleep, exercise, study etc. and these are reflected in varying hormone and enzyme levels, many of which are drug targets and some of which are involved in pain.  I remember learning at medical school that people have died after taking 12 paracetamol tablets (the standard size is 500mg and an adult dose is 1,000mg – two tablets – up to a maximum of 9 in a day), and that others had survived overdoses of 150.  I can’t help wondering whether part of this variation may have been to do with what time the overdose was taken. Studies looking at paracetamol lethality in rats and excretion in humans show quite marked variation with the time of day.

An awful lot of people die from paracetamol overdoses, not necessarily in suicide attempts.  It isn’t the paracetamol itself that is poisonous; it is metabolised to a toxic intermediate in the liver which then reacts with glutathione before being excreted in the urine, but when the available glutathione runs out then the toxic metabolite accumulates, damaging the liver and kidneys.  It is very easy to replenish the glutathione with oral (or IV) methionine (part of the treatment for a paracetamol overdose) and there was a time when you could buy paracetamol and methionine together in the same tablet, which made an overdose virtually impossible.  However, this never caught on.

Death from paracetamol is a slow and unpleasant process involving liver failure, and once it gets to that stage the only hope is a transplant.  Really, it is altogether much simpler, quicker and less painful to borrow a Japanese tantou ( 短刀 short sword made for the purpose) and disembowel yourself, saving the transplant surgeon the trouble of having to do it for you.


As you ascend the analgesic ladder there is a ceiling for each drug (sorry about the mixed metaphor) where increasing the dose won’t increase the pain-killing effect.  That is, until you get to the top rung where you will find morphine and its ilk.  If a little morphine doesn’t do the trick a lot of morphine probably will.

In my experience very few doctors know how to prescribe morphine and other opioids effectively.  Some worry that they are going to depress respiration and kill the patient, so they don’t give enough if they give it at all.  It is usually possible to get an idea of which patients are at risk of this, however, as they usually have quite bad respiratory problems to begin with.  You can always give a small dose, monitoring the pain and also whether the blood oxygen saturation goes down – if not it is probably safe to increase the dose a bit.

Then there are those that worry they are going to mask important signs of an as-yet undiagnosed problem, for instance in the case of an acute abdomen where you might want to know which spots are tender.  I am not convinced by this argument, particularly as the morphine has generally had plenty of time to have its effects and wear off again by the time the surgeon arrives to examine them.

And those that worry about creating addicts.  I once worked for a nephrologist who had a patient in her late teens with some horrible kind of chronic nephritis, and every so often she would get a flare-up and come into casualty in terrible pain.  His policy was to limit her to a few injections of pethidine.  Now pethidine has its place, particularly when the pain is going to be of short duration such as during an invasive procedure.  However, it is short-acting, and first it makes you high, then the pain eases, but quite soon after that you come down with a bump as the pain returns with a vengeance.  After that is the long wait for the next dose.  We call this clock-watching.

What could be more addictive than that?  Surely you need a pain-killer that isn’t going to intoxicate the patient and isn’t going to wear off until the pain is better.  Nowadays surgeons often use patient-controlled analgesia (PCA) where the morphine is slowly injected via a pump, and the patient can press a button to get a little bit more whenever they need it, which is surely a much better system.

Some are more equal than others

One thing to know about opioids is that, while they are broadly-speaking interchangeable (though you need to know equivalent doses), they do differ in how long they take to wear off (pethidine is the shortest), how much they depress respiration and how long for (methadone and buprenorphine can be troublesome here), and what sort of pain they are best for (opioids aren’t very good at neuropathic pain, involving irritation or compression of a nerve, but oxycodone is better for this than most of them).

Pain is bad, right?

Pain is an essential warning system to prevent us from damaging ourselves, and when you can’t feel it the result is disastrous.  A common problem in diabetes is loss of sensation in the feet due to peripheral neuropathy (damage to small nerves).  This can lead to stones in the shoes or ingrowing toenails going unnoticed until serious infection has set in, and the final stage is often amputation.  I remember a young man who had had poorly-controlled diabetes from childhood, and he came to the hospital after buying a new pair of shoes which were far too tight, but he had worn them all day anyway.  His feet were so squeezed that their blood supply had been cut off and they were starting to go gangrenous.  He had a dotted line tattooed across his forehead saying “cut here”, which wasn’t really very helpful as the surgeon needed to cut across the middle of his feet.  As my Registrar pointed out “intelligence is written in his face”.  Or tattoed.

The French neurologist Charcot described the total destruction of joints that followed when they lost their nerve supply (usually from syphilis, which we don’t see very often any more), and similarly the nerve damage from leprosy can lead to loss of extremities.  So don’t knock pain.

Cancer pain – Where to begin?

My experience is mainly with cancer pain.  Cancer pain really is pointless, and you are not protecting anything by ignoring it.  It is just there, all the time (at least until the cancer is treated) and unlike most other pains it is usually worse at night, interfering with sleep.  Because the level of pain is more-or-less constant, the level of pain-killers should be too.  There is no point in waiting for one dose to wear off before giving the next one because you know that the pain will come back if you do, and it is always harder to treat once it has.

You must start with a short-acting opiate, and morphine elixir (Oramorph) is the easiest one to use.  This comes in little bottles of revolting, sticky raspberry-flavoured stuff, 10mg morphine per 5ml teaspoonful, and if you give it every four hours then that keeps it topped up and the blood levels reasonably constant; not enough to get you too high, but enough to manage the pain.  For most people a sensible starting dose is 10mg every four hours, with instructions to take extra doses in-between as necessary.  The important thing is to make sure that every dose is written down.

After 24 hours you will know how much morphine was required.  Add up all the doses, divide by 6, and this becomes the new dose to take every four hours, extra doses still being allowed but hopefully only needed occasionally.  E.g. if they took 120mg in a 24 hour period the dose is 20mg every 4 hours.

Within a few days it should be clear what the correct dose is for that patient and their pain.  It is a nuisance having to take it six times a day (including at night) so this is the time to change to a long-acting version (i.e. slow-release).  For morphine there are various preparations such as MST and Zomorph that are designed to be taken every twelve hours.  Just add up the total amount of morphine elixir required in 24 hours, and this time divide by 2 to get the 12-hourly dose.  E.g. if they took 120mg in a 24-hour period then they need 60mg of slow-release morphine every 12 hours.

There still may be times when a little extra is needed, so they should keep the (short-acting) elixir handy for breakthrough pain, remembering that the breakthrough dose is 1/3 of the 12-hourly dose, and also remembering to write everything down.  Morphine is also available as short-acting tablets (e.g. Sevredol) which are interchangeable with the elixir but easier to keep in your pocket.

I was always taught that Oramorph and Sevredol take 20 minutes to act once you have swallowed them, but my own experience is that it is more like 75 minutes.  For really rapid relief nothing beats an injection, but man drugs are also absorbed rapidly through the lining of the mouth (such as nicotine in cigar smoke), and there are several preparations of fentanyl which you can put under the tongue (sublingual) or between the lip and the upper gum (buccal) which work in minutes.  Personally I have found them to be very good.  100mcg of sub-lingual fentanyl is equivalent to 10mg of oral morphine.

Where there is nerve pain, it might be better to use oxycodone rather than morphine.  5mg oxycodone is equivalent to 10mg morphine, so you start with 5mg every four hours and otherwise continue the same way.  The short-acting version is a tablet called Oxynorm, and the long-acting one is Oxycontin.

Now Oxycontin, MST, Zomorph and all the other long-acting opiates, including those available as trans-dermal patches (i.e. sticking-plasters) are absorbed very slowly.  This means that it takes four or five days for the blood concentration to level off after increasing or decreasing the dose.  I tell my patients that THE TABLETS THEY TAKE TODAY ARE FOR THEIR PAIN TOMORROW, and it is no good taking an extra one if the pain gets worse because it won’t work, it will just accumulate until it causes a problem.  It is absolutely essential to wait 4 – 5 days before assessing whether the dose is correct, and making any changes in the meantime will result in total confusion for all concerned, leading to uncontrolled pain or else opiate toxicity.  If an extra dose of anything is needed, it MUST BE THE SHORT-ACTING VERSION, and WRITE IT DOWN.

I’m a bit sensitive

People do differ a lot in their response to opiates.  From time to time you will come across someone who is unusually sensitive to them.  Often they will find that the initial 10mg of morphine (5mg of Oxynorm or whatever) is very sedating, and quite likely to make them vomit.  They will need much smaller doses, so try halving or quartering it.  Sometimes a different opiate is better tolerated, e.g. oxycodone or fentanyl rather than morphine, remembering to keep the dose low to start with.  It is unusual for somebody not to be able to tolerate any opiates at all, though if they can’t that can make pain control much more difficult.

For long-acting pain control, trans-dermal patches can work very well.  These are available with either fentanyl or buprenorphine.  Personally I can’t stand buprenorphine.  It is a partial agonist, which means that when it attaches to the opiate receptors in the brain it blocks the effect of other opiates, so if you need a breakthrough dose of morphine it won’t work.  It also takes a very long time to clear, which makes it more dangerous in overdose.  It used to come as sub-lingual tablets called Temgesic when I was a House Officer, and my main recollection is that it wasn’t very effective, it made the patients dizzy, and most of them vomited when they used it.

Fentanyl patches work very well.  They are designed to be changed every 72 hours to maintain steady blood levels.  The come in different sizes, measured in micrograms per hour.  25 mcg/hr is equivalent to 10 – 20mg morphine every four hours.  You can make up the dose with several patches, e.g. 12 + 25 = 37mcg/hr.  There are a lot of different makes.  Some stick better than others, and they use different adhesives, which is useful if the patient is allergic to one of them.

Something that I have never understood about transdermal patches is that patients seem to wear them according to what they are treating.  GTN patches for angina get worn on the chest, HRT on the lower back, over where they think the ovaries are, Scopolamine patches (for travel sickness) go behind the ear as this is where the balance organ is.  Since they work by the drug crossing the skin to get into the bloodstream this is all nonsense.  The main thing is that they shouldn’t go somewhere where they will come off easily, so you need to think of how hairy or sweaty or mobile your skin is.


You may need to know some equivalent doses. Mostly you can look these up in a standard formulary, but it is useful to know that:

10mg diamorphine (Heroin) = 15mg morphine

5mg oxycodone = 10mg morphine

100 mcg (microgram) fentanyl = 10mg morphine

25 mcg/hour fentanyl patch = 10 – 20mg morphine every 4 hours

30mg MST / Zomorph every 12 hours = 10mg Oramorph every 4 hours

Giving any of these by injection is equivalent to doubling the dose:

5mg IV morphine = 10mg oral morphine

5mg IV diamorphine = 15mg oral morphine etc.

Unwanted effects

All opiates have fairly similar side-effects.  They tend to make you feel rather spaced-out for the first few days, but this wears off, which is good as most people would much rather be clear-headed.  They are all constipating, so don’t wait for constipation to get established before you treat it.  An osmotic laxative such as Movicol or Laxido is easy to take and works very well.  I suggest starting with a sachet twice a day in a glass of water – not so much that you can’t drink it down, but dilute enough that it isn’t slimy.  Laxido is orange-flavoured and in my opinion the nicest.

Nausea with opiates is quite common.  Ondansetron and granisetron are very good at preventing nausea but not so effective when you are already feeling sick.  They are also even more constipating than opiates.  Cyclizine is very good for stopping nausea, particularly when given by injection, but it can make you a bit sleepy and is also constipating (along with all antihistamines).  Domperidone is cheap and easy and doesn’t have side-effects but is fairly mild.  Metoclopramide is very commonly prescribed, but a lot of people feel rather non-specifically uneasy after taking it, and I think that they are probably having sub-clinical extrapyramidal toxic effects (for non-medics, this is a bit like Parkinson’s disease, and it is rather hard to pin down exactly what is the matter, just that you don’t feel right).  Since it promotes gastric emptying it and the others don’t it has its place, but mostly I try to avoid it.

If somebody does have respiratory depression with opioids, this can be instantly reversed by an intravenous injection of naloxone.  In the case of someone with cancer pain the result is an immediate return of their pain, worse than ever, which you can’t treat for a few hours.  That is very cruel and hardly ever necessary, so think before you give it.

Don’t forget other approaches

In oncology, the best pain control is very often radiotherapy.  A single dose directed at the right spot generally has very little in the way of side-effects and works very well indeed, though it takes up to six weeks to become maximally effective.  For the benefit of medical oncologists reading this – don’t be afraid to ask your clinical oncologist colleagues (for the benefit of non-medics, a medical oncologist is a half-trained oncologist who doesn’t know a photon from a proton, let alone an electron from a Selectron).

When it is all over

Hopefully the pain won’t last for ever. Perhaps the radiotherapy / chemotherapy has worked, or the tumour-related fracture is healing. The you simply reduce the opiate dose, going down by about a third at a time, and waiting long enough to ensure that the pain really isn’t coming back before reducing further. How quickly you can drop the dose depends on how long they have been taking it for, but I have seldom known anybody report withdrawal effects. Curiously, when I have come off opiates I have had restless legs at night for a few days, which is very unpleasant as they keep me awake, but standard sleeping tablets dealt with that. This is not a withdrawal symptom that I have ever heard or read of, but real patients are seldom like the ones in the textbooks (that’s why everybody gets so excited about “a classic case of…”).

2 thoughts on “What they don’t teach you in medical school about pain-killers

  1. After billions spent on painkillers & pain research I am still in pain after 15 months and feel like the medical profession is still in the dark ages , which is a far cry from the exalted position the medical profession has in the UK & elsewhere. This adulation cannot come from anyone who has been in pain for many years . Am I mistaken but has there been no improvement on Opium which has been used for more than 4000 years?


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.