Evaluation of a short duration behaviour‐based post‐operative pain scoring system in rats

Abstract

We have recently demonstrated dose-related analgesic-induced reductions in the occurrence of 7 behavioural activities following midline laparotomy in rats. For these behaviours to be useful in evaluating pain in laboratory rats they must be shown to occur after different types of surgery, and frequently enough to allow rapid scoring of animals. Here, the relevant behaviours were used to test the analgesic efficacy of meloxicam with a variation of our previous laparotomy model. As part of an unrelated project, 57 male Fischer rats were divided into groups to receive either saline (0.2 ml/100 g s/c), meloxicam (0.5, 1 or 2 mg/kg s/c) or carprofen (2.5, 5, or 10 mg/kg s/c) 1 h before surgery. Behaviour data were collected for 10 min following 25 min of recovery from isoflurane anaesthesia. The cumulative frequencies of back arching, fall/stagger, writhe and poor gait were used to compute a composite behaviour score. Irrespective of whether analyses included only 5 or all 10 min of the observation period, the relevant behaviours occurred significantly more often in rats given saline or low dose meloxicam than in those given 1 or 2 mg/kg of meloxicam, or any dose of carprofen. We conclude that this technique of quantifying post-surgery behaviour is an effective pain scoring method following abdominal surgery in rats, and that 1 mg/kg meloxicam significantly attenuates laparotomy induced pain. Since only a short observation period is required, this approach represents an important practical advance in assessing abdominal pain severity and clinical drug potency. Keywords Rat Behaviour Pain Analgesia Meloxicam Carprofen 1 Introduction It is generally accepted that if animals are to be used in biomedical research, any pain or distress that may result should be prevented or minimised. Post-surgical pain can be controlled using analgesics, however the use of an appropriate treatment regimen requires the development of methods of assessing pain. Pain in man is recognised as being a subjective experience but its nature in animals remains elusive. There is, however, a general acceptance amongst many research workers and veterinarians that alterations in animal behaviour are likely to accompany pain ( Morton and Griffiths, 1985 ; Zimmerman, 1986 ; Roughan and Flecknell, 1996 ; Dobromylsky et al., 2000 ; Roughan and Flecknell, 2000, 2001 ). This has prompted some recent efforts to develop behavioural analysis techniques for quantifying post-surgical pain in rats ( Liles et al., 1998 ; Roughan and Flecknell, 2000, 2001 ) and dogs ( Fox, 1995 ; Firth and Haldane, 1999 ; Fox et al., 2000 ), although most work has focussed on farm animals such as lambs ( Molony et al., 1993 ; Kent et al., 1995 ; Molony and Kent, 1997 ; Graham et al., 1997 ; Kent et al., 1998 ; Thornton and Waterman-Pearson, 1999 ; Kent et al., 2000 ) and calves ( Robertson et al., 1994 ; Molony et al., 1995 ). All of this work has been based on the assumption that pain severity and analgesic efficacy can be scored by quantifying the relative magnitude of the behavioural alterations that occur due to surgery and intervention therapy. To date, none of these behavioural assessment techniques have been developed into a practically useful scoring system. The difficulties inherent in applying pain scoring systems include the time required for assessments, training requirements of observers, the occurrence of large variations between observers, and a failure to validate the system by use of appropriate control groups ( Leese et al., 1988 ; Reid and Nolan, 1991 ; Thompson and Johnson, 1991 ; Nolan and Reid, 1993 ; Holton et al., 1998 ; Lascelles et al., 1998 ; Grisneaux et al., 1999 ; Firth and Haldane, 1999 ). Our inability to score pain effectively in animals has contributed to a lack of relevant clinical efficacy data for many of the analgesics available for animal use, and an under use of analgesics. For example, recent surveys report that analgesics are used relatively infrequently in dogs and cats following procedures such as ovariohysterectomy and castration, and are given to less than 25% of small mammals in UK veterinary clinical practice ( Lascelles et al., 1998 ; Capuer et al., 1999 ). It is difficult to judge the frequency of analgesic use following experimental surgery in animals, but it is at best likely to be similarly low. This situation represents a serious welfare concern. With a view to addressing this problem we have studied post-operative pain related behaviour in rats in considerable detail ( Roughan and Flecknell, 2001 ) and demonstrated significant changes in some specific behaviours following surgery. The activities concerned (arching of the back, loss of balance during grooming or rearing, a transient ‘twitching’ while inactive and abdominal writhing were most prominent in animals that received only saline, but were significantly attenuated for between 4 and 6 h in analgesic treated animals. These key behaviours were unaffected by drug treatment in unoperated control groups, and this further supported our belief that their occurrence could be used to assess pain and the effectiveness of analgesic treatment. The behaviours also occurred sufficiently frequently to enable them to be used to develop a pain scoring scheme. To be of practical value, a pain scoring system must be easy and rapid to undertake, and ideally should be applicable to a range of strains and ages of animal, and to different surgical procedures. These features have not been demonstrated for any of the pain assessment techniques currently available. Our previous study had utilised a single age and strain of rats, and a standardised surgical procedure (laparotomy). In addition, animals were observed for prolonged periods. The present objective was to gauge the limits of the scoring procedure, by assessing whether similar behaviour changes could be used to determine pain severity and analgesic efficacy in a different age and strain of rat...