What analgesic dose should I use and why?


Providing effective pain relief requires administration of an appropriate analgesic regimen for an appropriate time. This seems straightforward, but experience in man and animals has shown that effective pain management requires careful and frequent assessment of pain intensity. When dealing with laboratory animals, an additional concern is the potential interaction between analgesic agents and the specific goals of a research project.

This seminar covers advances in pain assessment methods in laboratory species and discusses options for analgesic treatment, the difficult problem of determining the duration of treatment and frequency of dosing and approaches to dealing with interactions between analgesic agents, pain, and research protocols. Particular emphasis is given to developing practical approaches to assessing and managing pain, and integrating this within an overall scheme of optimising animal welfare.


  • Introduction (00:00)
  • Where do published dose rates come from? (3.34)
  • Analgesics don’t always work (14:37)
  • Analgesics and interactions with research projects (19:06)
  • What drug, what dose and when? (24:34)
  • What route of administration should I use? (32:39)
  • Resource issues and summary (37:27)

Questions from the live session

Which analgesic drug are you recommending for liver operations in rodents such as liver resection, portal vein ligation or liver transplantation? Which doses, frequency, and duration of application, and what is the rationale for your recommendation?
The answer to this question sums up what the webinar was all about. We should start with our estimate of the likely pain post-surgery, and our “standard” analgesic doses. In this case, I’d be thinking of using an opioid peri-op (probably buprenorphine) and an NSAID, and local anaesthetic infiltration into the abdominal wound.

I’d do pain assessment post-op to decide whether I needed more than a single dose of buprenorphine, or whether the NSAID was sufficient.

I’d anticipate giving a second dose of buprenorphine but would expect the degree of pain to be reduced greatly after 12-16h.

The impairment of liver function might extend the duration of action of some drugs, but if there is any activity in the transplant, the very small quantities of drugs given should be metabolised. In transplants in larger species, we saw no issues in relation to this, nor in liver resection models in rodents.
What is the practicality of the MGS in cage-side use, given that I thought the original work was done with high-speed cameras?
We have used “real time” assessments and Matt Leach runs through this with some video examples in our Pain workshop – It’s not as accurate as scoring images but much better than no assessment at all. Alternatively, use a camera in burst mode to get a series of photos and scan them, or better still, let the animal poke its head out of a handling tube and get some nice full – face images. Cheap compact cameras, consumer grade digital video and any good smartphone all give sufficient quality.

Do check that your animal facility allows you to use cameras, and if they don’t, start a discussion as to why this should be reviewed to facilitate pain assessment.
You’ve mentioned grimace scales in rodents and horses. What about in pigs?
As far as I am aware, the only published scale is for piglets, but there is a thesis online with adult pigs – I am not sure why there are no publications from it, perhaps due to small group sizes, but it looks as though a pig grimace scale could be developed.
What is your opinion on the relative safety of common NSAIDs (carprofen, ketoprofen, meloxicam) in rats? Which would be your NSAID of choice?
They are all relatively safe and effective. There are rare reports of adverse effects, mainly with ketoprofen but also with carprofen. The only meloxicam adverse reaction I have heard of was shown to be caused by a relative overdose. My drug of choice is meloxicam as it can be orally dosed (to individual animals) easily with the liquid formulation, and we did demonstrate that it was effective in rats post-laparotomy. The reason not to use ketoprofen is that of the three agents, its effects on platelets, increasing clotting times, seems to have the potential to be of clinical significance (at least in rabbits, the species studied in the relevant paper). Carprofen is also safe and effective, but the oral dosing option is not so good for rats.
You have touched on differences in metabolising analgesics. What are your thoughts on creating tailor-made analgesia based on assessment of high vs low metabolisers? Have we reached that stage yet?
Some of the differences in analgesic efficacy are almost certainly related to metabolism, but we can’t really predict this, other than by adjusting our doses depending upon pain assessments. Of course, once you have determined that “your” rats seem to consistently need more or less analgesic, you can adjust the starting dose and have fewer needs for intervention analgesia for animals showing signs of pain – you may also be able to adjust your redosing/reassessment intervals.
Sometimes scientists do not want to use opiods and nsaids due to protocol interactions. Paracetamol is the only possibility left. How do we use that?
Don’t forget the option of local anaesthetics, as an adjunct to analgesia for all surgical procedures. Paracetamol is available as an injectable formulation, and there are a couple of publications (maybe more, but I at least found 2!) which describe pK in rat (and I think mouse) with s/c or i/p dosing. Of course, you also have the more familiar oral route. We’ve been very disappointed with the degree of analgesia, but perhaps it needs to be given with a weak opioid or an nsaid for post-surgical pain, and in the situation you describe, that may not be an option.

My aim would be: 1) Discuss with the investigators if other agents really are contraindicated at the doses needed for post-op pain relief, 2) Ensure effective pain assessment is carried out, 3) Remind them that a surgical procedure without analgesia is likely to be categorised as severe, and greater ethical justification is going to be required.
I study skeletal muscle, and a very much used method is synergist ablation, removing a synergist in the lower leg. Any comment on best practice for analgesia, in rats and mice? Pain assessment gives the impression that pain is low even a short time after operation. E.g. mice hang from the cage ceiling after 10–15 minutes.
It’s not a procedure I am familiar with, but I believe it requires surgical removal of the gastrocnemius and soleus muscle. I assume a skilled surgeon can do this and cause minimal tissue trauma. I was pleased to see a methods paper (Kirby TJ, McCarthy JJ, Peterson CA, Fry CS. Synergist Ablation as a Rodent Model to Study Satellite Cell Dynamics in Adult Skeletal Muscle. Methods Mol Biol. 2016;1460:43-52. doi: 10.1007/978-1-4939-3810-0_4. PMID: 27492164.) recommended analgesia – carprofen or meloxicam for 2–3 days. I would check efficacy with grimace scoring post-op, but would anticipate it being sufficient, but would also consider local anaesthetic infiltration, but only after discussion with the investigators as local anaesthetic injected directly into muscle is used to remove muscle fibres (in a different model).

Finally, your comment about the mice and rats being apparently normally active within 15 minutes of recovery is important. One of the problems we had with pain in animals was that we often start with the assumption that “of course animals feel pain, they are just like us”, but then when they don’t behave like us, we can start to think “well maybe that’s because they feel less pain, or maybe they just aren’t in any pain…”. Animals in pain behave in species-specific ways to express that pain – but that’s a topic for another webinar!
Do you have any comments regarding using buprenorphine to reverse Fentanyl anaesthesia, how does it effect the time of emergence? Is there a gold standard on what dose to use? (thinking more specifically about rabbits here if there is a difference)
We published a comparison on this years ago (and it has some very bad stats in the paper which were my fault!) but all the agents worked (nalbuphine, butorphanol and buprenorphine). With buprenorphine, we’ve found prolonged sedation in mice, but not so in rabbits, and have moved to recommending butorphanol for the immediate reversal (as it is easy to obtain, and has a more rapid onset of action). We then follow up with buprenorphine about 1-2 hours later. We give “enough” to produce emergence – as the dose varies depending upon how much fentanyl has been given, and when. When doing this after an i/v infusion in rabbits (which often stops spontaneous respiration), then the dose can be adjusted depending upon the return of normal depth of respiration – we give the butorphanol slowly, i/v and repeat as needed (leaving the i/v catheter in place until the rabbit is fully recovered).
How do you feel about non-invasive analgesic delivery in rodents? We want to reduce the handling of our mice after thoracic surgery. We give analgesics in Nutella and, if not consumed within 20 mins, we give them injections. Do you think this is alright? We will now add checking grimace scores. When is the best time and how often a day should we do the checks? Our mice have myocardial infarction surgery, so it is a severe model and pain control is critical for the mice.
If the mice are group housed, it is going to be tricky to be sure all of them had an appropriate quantity of Nutella – but if it’s a mouse on its own immediately post-op then that should be fine (and I assume they had a dose by injection peri-operatively). I’d wait for full recovery from anaesthesia, then check them – and repeat after about 2h or so. If you anticipate needing to give more analgesic at, say, 6h, (following buprenorphine at the time of surgery), check them at 6h, dose, and check again an hour later. If the scores were high at 6 h, check earlier with the next mice and maybe dose earlier. If the MGS seemed normal, you may want to go to perhaps 8h and assess and dose. This way, you start to adjust the timing to suit your mice. Of course, you might also need to adjust the dose rate, as well as timing – if the MGS is high an hour after re-dosing, or high at your first assessment, try increasing the dose. As I said in the webinar, this is very demanding, but we should do what we can – the more we do, the more confident we become in our pain management.
Could you use local anaesthetics when taking several blood samples for insulin measurements? Could it also impact blood glucose?
It depends over what time course – I am guessing you would be doing tail vein or saphenous bleeding in mice or rat – you could put EMLA cream or similar on the skin overlying the venepuncture site, but it needs to stay in place for 20–30 mins (or perhaps longer on an adult rat tail) and achieving this is a problem as the animals just remove the cream and/or the bandage. It works well in rabbits, but in rodents I think it’s not practicable. Spraying on local anaesthetic, or using sprays to produce local cooling are (In my experience) of very little or no value. So I’d say, habituate the animal to restraint (see the RISE (Previous SWETOX) videos) and then you can perform venepuncture with minimal restraint, and minimal stress and that will keep blood glucose (and presumably insulin) stable. In a very stress sensitive animal (the newborn piglet) handling restraint produced a 3-5 fold increase in blood glucose (I think that is buried somewhere in my PhD thesis and never made it into a publication….).

About the Author

Webinar presented by Professor (Emeritus)  Paul A Flecknell, MA, VetMB, PhD, DECLAM, DLAS, DECVA, (Hon) DACLAM, FRSB, (Hon) FIAT, (Hon) FRCVS Newcastle University, and CEO of FLAIRE Consultants, U.K. Paul has over 40 years of experience in the care and welfare of animals. He has authored numerous scientific publications, books and book chapters and has an international reputation in this field. He worked as Director of a multi-species research facility for over 30 years and has extensive experience in the development and delivery of training and education in animal care and welfare. He has wide experience in advising on facility management and ethical issues. As head of the Pain and Animal Welfare Science (PAWS) group at Newcastle, he published over 150 scientific papers, reviews and book chapters in the field of analgesia and anaesthesia of laboratory animals.

Related Articles