Abstracted by Paul Flecknell from his book Laboratory Animal Anaesthesia 2nd Edition, 1996, and presented at the CALAS/ACTAL Convention in Prince Edward Isle, Canada
This article originally appeared in the CALAS/ACSAL Newsletter Vol 30 #5 October 1996. It is reprinted with the permission of the Canadian Association for Laboratory Animal Science/L'Association Canadienne pour la Science des Animaux de Laboratoire.
The effective alleviation of post-operative pain in laboratory animals should be considered an important goal in all research establishments. Despite the emphasis given to humane treatment of laboratory animals in the national legislation of many countries, analgesia may still not be administered routinely in the post-operative period. This omission is particularly common when the animals concerned are small rodents. When analgesics are administered, assessment of methods of pain recognition or severity may account in part for the relatively infrequent use of analgesics in animals, in comparison to their use in man. This is not meant to imply that veterinary surgeons and others involved in animal care are incapable of recognizing that an animal is in pain, but preconceptions about animal pain may limit the value of any assessment of its severity (see below).
Although we would wish to alleviate pain because of concerns for animal welfare, a number of counterarguments have been advanced to justify withholding analgesics:
Alleviation of post-operative pain will result in the animal injuring itself. Provided that surgery has been carried out competently, administration of analgesics, which allow resumption of normal activity, rarely results in problems associated with the removal of pain's protective function. Claims that analgesic administration results in skin suture removal are unsubstantiated, and contrary to findings in our laboratory. In certain circumstances, for example after major orthopaedic surgery, additional measures to protect and support the operative site may be required, but this is preferable to allowing the animal to experience unrelieved pain. All that is required in these circumstances is to temporarily reduce the animal's cage or pen size, or to provide additional external fixation or support for the wound. It must be emphasized that these measures are very rarely necessary, and in our institute, administration of analgesics to laboratory animals after a wide variety of surgical procedures has not resulted in any adverse clinical effects.
Analgesic drugs have undesirable side-effects such as respiratory depression. The side-effects of opiates in animals are generally less marked than in humans and should rarely be a significant consideration when planning a post-operative care regimen.
We don't know the appropriate dose rates and dosage regimens. This is primarily a problem of poor dissemination of existing information. Virtually every available analgesic drug has undergone extensive testing in animals. Dose rates are therefore available for a range of drugs in many common laboratory species (17, 36). It is occasionally difficult to extrapolate available dose rates from one species to another and to translate dose rates that are effective in experimental analgesiometry into dose rates which are appropriate for clinical use. Nevertheless, in most instances, a reasonable guide as to a suitable and safe dose rate can be obtained.
Pain relieving drugs might adversely affect the results of an experiment. Although there will be occasions when the use of one or another type of analgesic is contra-indicated, it is extremely unlikely that there will be no suitable analgesic that could be administered. More usually, the reluctance to administer analgesics is based upon the misconceived idea that the use of any additional medication in an experimental animal is undesirable. The influence of analgesic administration in a research protocol should be considered in the context of the overall response of the animal to anaesthesia and surgery. The responses to surgical stress may overshadow any possible adverse interactions associated with analgesic administration. An additional consideration is that many arrangements for intraoperative care fail to control variables such as body temperature, respiratory function and blood pressure. It seems illogical to assume that changes in the function of the cardiovascular or respiratory systems are unimportant, but that administration of an analgesic will be of overriding significance. It should be considered an ethical responsibility of a research worker to provide a reasoned, scientific justification if analgesic drugs are to be withheld. It is also important to realize that the presence of pain can produce a range of undesirable physiological changes, which may radically alter the rate of recovery from surgical procedures (28).
When debating the nature of pain in animals, considerable parallels can be drawn with the situation in human infants. In adult humans, the ability to provide direct verbal communication, complete pain questionnaires or scoring systems, or to directly manage analgesic dosage using patient controlled analgesia systems allows reasonably reliable estimates to be made of the degree of pain and the efficacy of pain control. In young human infants, written and verbal communication is not possible, nevertheless, extrapolation from adult humans, coupled with objective demonstrations of the adverse effects of surgical stress, has led to a huge increase in interest in providing pain relief to these patients (2, 41).
The approaches used in human infants can provide a framework for animal pain assessment. The most widely used techniques have been pain scoring systems based upon criteria such as crying, facial expression, posture and behaviour (42). This type of approach was advocated as a means of assessing pain and distress in animals (43). This paper influenced a large number of other groups, who modified the original hypothesis, but retained the central notion of identifying pain specific behaviours, and rating them in some way (3, 18, 33). Surprisingly, progress in validating this hypothesis has been remarkably slow. An early report (35) indicated that the technique could be applied successfully, but the few subsequent published data are less encouraging. Particular problems noted were considerable between observer variation and the poor predictive value of certain of the parameters scored (5, 6). The between observer variation is not unexpected, and parallels problems recognized in human pain scoring. It appears that if the number of observers is restricted, and the criteria used was carefully selected, reasonable agreement can be achieved (49).
The basic methodology selecting clinical signs which might be due to pain has been used to provide pain-scoring systems in veterinary clinical patients. Attempts at scoring have either used descriptive ratings converted to numerical scores to allow statistical analysis, or have used visual analogue scoring systems (VAS) (45, 46, 48, 49, 53). A problem with many of these studies is the difficulty associated with scoring of animal behaviour in a relatively brief period. If it is believed that behavioural responses can indicate pain, and hence the efficacy of analgesia, then more detailed assessments are likely to be required. Support for the value of behavioural observations is provided by studies of the effects of tail docking and castration in lambs (54) and castration in piglets (40).
In laboratory animals, a number of different approaches have been used to assess pain or distress. The most extensive studies have been undertaken to investigate chronic pain, for example, those by Colpaert et al (8, 9, 10, 11), using an adjuvant arthritis model in the rat. Body weight, minute volume of respiration, mobility, vocalizations, specific behaviours and self-administration of analgesics were all considered as indices of pain. When discussing the results of these investigations, the authors concluded that all of the parameters responded to the same stimulus, and that the most reasonable explanation was that they were influenced by the presence of pain (9). Motor behaviour changes have been suggested as indices of pain (7, 55) and loss of appetite and reduction in body weight have been noted in rodents post-operatively (23, 24, 55). Recently, these variables have been studied in rats as potential means of assessing the degree of post-operative pain, and comparing the efficacy of different analgesic regimens (20, 21, 37, 38). As with other pain assessment techniques in animals, these assumed that if a change to a variable occurred after a procedure that would cause pain in man, then the change may be related to pain in the animal. If administration of an analgesic reverses the changes associated with the procedure, this supports the hypothesis that the changes were, at least in part, pain related. Clearly it is important to establish that the analgesic did not have non-specific effects in normal animals that would influence the variable studied. This is a somewhat circular argument, since it is simply stating that indices of pain are those indices that are normalized by administration of analgesic drugs. Although efficacy of these analgesics in reducing peripheral input in animals is well-established, (1, 13, 31, 52), their effects on clinical pain are only validated in humans.
The uncertainty surrounding pain scoring could be circumvented if some independent validation method were available. In man, a series of objective criteria have been proposed to assess pain. These have included pulse rate, skin conductance and resistance, blood pressure and skin temperature. In addition, biochemical and endocrine parameters, such as blood corticosterone or cortisone concentrations or catecholamine concentrations, have been proposed as indicating pain. A major problem in interpreting the significance of these changes is the influence of surgery and anaesthesia, which markedly alter many of these variables, even in patients which are pain free (29). The surgical stress response occurs in all patients, and although it can be reduced by intra-operative use of opioids, it occurs even in patients who receive a high level of post-operative pain control. In man, catecholamine and cortisol responses have shown to be poorly correlated with post-operative pain scores. Use of these variables in animals has the same constraints. Catecholamine rises have been demonstrated in cats (4) and dogs (48), and cortisol response is less following thoracotomy when epidural morphine rather than intravenous morphine is administered (48). However, lack of appropriate controls and influence of surgical stress limit the significance that can be attached to these studies. Despite these reservations, studies such as those of Popilskis et al (48), which correlate both subjective pain scores and endocrine responses, advance a persuasive case of the validity of pain scoring. Nevertheless, the difficulties highlighted by studies in man suggest that biochemical indices are unlikely to provide a reliable objective method of pain assessment in animals.
An alternative approach is to adopt the well-established human clinical technique of administering analgesics as a continuous infusion. Infusions of analgesics have the advantage of maintaining effective plasma levels of the analgesic, thus providing continuous pain relief. This is in contrast to intermittent injections, where pain may return before the next dose of analgesic is administered. This technique obviously poses some methodological difficulties in animals, but if an indwelling catheter and harness and swivel apparatus are available, this can be arranged quite simply. In larger species (3-4 kg body weight), a light weight infusion pump can be bandaged directly to the animal and continuous infusion made simply by means of a butterfly type needle anchored subcutaneously or intramuscularly. When analgesics are to be administered by continuous infusion, the infusion rate can be calculated from a knowledge of the pharmacokinetics of the analgesia to be used. If these data are not readily available, an approximation that appears successful in clinical use is as follows: calculate the total dose required over the period of infusion, reduce this by half and set the pump infusion rate accordingly; administer a single, normal dose of the drug as an initial loading dose and start the infusion. The rate can then be adjusted depending upon the animal's responses.
Attempts to provide both longer periods of pain control and more effective analgesia have led to the development of alternative methods of drug delivery. The majority of these techniques have been developed in man and some have been used successfully in companion animals.
Administration of small quantities of medicated food does not avoid the need for repeated attendance overnight, but does remove the need for repeated subcutaneous or intramuscular injections in small rodents. Provision of analgesia with buprenorphine in flavoured gelatin, "Buprenorphine Jello" (47), seems to be an effective means of providing post-operative pain relief. In our laboratory, we have noted that rats are initially cautious of jelly pellets, but once one pellet has been consumed, subsequent pellets are eaten as soon as they are offered. It is therefore advisable to commence administering pellets, which do not contain analgesic 2 to 3 days before surgery. After surgery, analgesic containing jelly can be given. The flavoured gelatin used is domestic fruit-flavoured jelly reconstituted at double the recommended strength.
Techniques for administration of food pellets at intervals to experimental animals are well-established, and it would be a relatively simple procedure to introduce an automated means of delivering pellets at appropriate time intervals. The technique could also be used with larger species and need not be restricted to opioids or, indeed, analgesics. Provided that the animal is eating or drinking, small quantities of highly palatable material could be provided at appropriate intervals. Simple timer devices to achieve this are already marketed for delayed feeding of pet dogs and cats.
As mentioned above, the administration of opioids by any route can be associated with the development of respiratory depression. It must be emphasized that this is rarely of clinical significance in animals, unless high doses of pure mu agonists (for example, fentanyl) are used. If respiratory depression occurs, it can be treated by the administration of the opiate antagonist drug, naloxone. Administration of naloxone will also reverse the analgesic effects of the opioid, and it may be preferable to correct the respiratory depression by the use of doxapram. Alternatively, if a mu opioid agonist such as morphine or fentanyl has been used, the respiratory depression can be reversed using nalbuphine or butorphanol, and some analgesia maintained because of the action of these latter two agents at kappa receptors. Repeated administration of these agents may be required, and the animal should be observed carefully for several hours to ensure that adequate respiratory function is maintained.
The expertise of the surgeon can also greatly influence the degree of post-operative pain. Good surgical technique which minimizes tissue trauma and the prevention of tension on suture lines can considerably reduce post-operative pain. The use of bandages to pad and protect traumatized tissue must not be overlooked and forms an essential adjunct to the use of analgesic drugs.
Aside from measures directed towards alleviating or preventing pain, it is important to consider the overall care of the animal and the prevention of distress. Distress is used in this context to describe conditions which are not in themselves painful, but which are unpleasant and which the animal would normally choose to avoid. For example, recovering from anaesthesia on wet, uncomfortable bedding in a cold, unfamiliar environment would be likely to cause distress to many animals. It is essential to consider the methods described for the control of pain in conjunction with the techniques discussed earlier aimed at providing good post-operative care.
When carrying out any surgical procedure, buprenorphine is administered either pre-operatively or immediately following induction of anaesthesia, if a volatile anaesthesia is used. If neuroleptanalgesic regimens are used, or mu opioids are given as part of a balanced anaesthetic technique, then administration of buprenorphine is delayed until completion of surgery. If the procedure is relatively minor (for example, jugular or carotid cannulation) then only a single dose of analgesic is administered. In some circumstances a potent nonsteroidal anti-inflammatory drug (NSAID), such as flunixin or carprofen, may be used as an alternative to buprenorphine.
Following more invasive surgical procedures, such as laparotomy, orthopaedic surgery or craniotomy, opioid administration is continued for 24-48 hours. When undertaking major surgery, particularly in larger species when the degree of tissue trauma tends to be greater, analgesic administration may continue for 72 hours. Frequently, the regimen chosen consists of opioids (buprenorphine) in combination with an NSAID for 24-36 hours, followed by NSAID alone for a further 24-36 hours. (See tables 1-4 for suggested dose rates.)
Paul Flecknell may be contacted at e-mail: p.a.flecknell@newcastle.ac.uk
Go to:
Contents, Animal Welfare Information Center Newsletter
Top of Document
The Animal Welfare Information Center
U.S. Department of Agriculture
Agricultural Research Service
National Agricultural Library
10301 Baltimore Ave.
Beltsville, MD 20705-2351
Phone: (301) 504-6212
FAX: (301) 504-7125
Contact us: http://www.nal.usda.gov/awic/contact.php