A study to evaluate buprenorphine at 40 lg kg )1 compared to 20 lg kg )1 as a post-operative analgesic in the dog

Transcription

1 Veterinary Anaesthesia and Analgesia, 211, 38, doi:1.1111/j x RESEARCH PAPER A study to evaluate buprenorphine at 4 lg kg )1 compared to 2 lg kg )1 as a post-operative analgesic in the dog Louisa S Slingsby*, Polly M Taylor & Joanna C Murrell* *School of Clinical Veterinary Science, University of Bristol, Bristol, UK Gravel Head Farm, Downham Common, Little Downham, Nr Ely, Cambridgeshire, UK Correspondence: Louisa S Slingsby, School of Clinical Veterinary Science, University of Bristol, Langford House, Langford, Bristol BS4 5DU, UK. louisa.slingsby@bristol.ac.uk Abstract Objective Comparison of the analgesic effect of buprenorphine at 2 or 4 lg kg )1. Study design An investigator blinded, randomised study. Animals Twenty six dogs presented for ovariohysterectomy. Methods Dogs were premedicated intramuscularly with acepromazine.3 mg kg )1 and buprenorphine at either 2 (B2, n = 12) or 4 lg kg )1 (B4, n = 14) followed by anaesthetic induction with propofol and maintenance with isoflurane. During anaesthesia non invasive blood pressure, heart rate, respiratory rate, blood oxygen saturation, inspired and expired volatile agent, end-tidal carbon dioxide and ECG were recorded. Pain and sedation were assessed using interactive VAS scores; mechanical nociceptive thresholds were measured at the wound and hindlimb all were assessed before and up to 22 hours after administration. Carprofen was used for rescue analgesia. Results There were no significant differences between the two groups for any of the parameters examined. Rescue analgesia was required around 5 hours after administration of buprenorphine in a significant number of animals. Sedation was good preoperatively and scores decreased over time postoperatively. Hock thresholds did not change over time; wound thresholds decreased significantly compared to the baseline value from 3 hours onwards. Conclusions Administration of buprenorphine at either 2 or 4 lg kg )1 IM with acepromazine provided good pre-operative sedation. Cardiovascular and respiratory values remained within clinically acceptable limits during anaesthesia. There was no evidence that increasing dose increased adverse events that may be associated with opioid administration (e.g. bradycardia and respiratory depression). Clinical relevance Increasing the dose of buprenorphine from 2 to 4 lg kg )1 did not provide any benefits with respect to analgesia after ovariohysterectomy as assessed using the VAS scoring system. Keywords analgesia, antinociception, buprenorphine, dog, ovariohysterectomy, pain score. Introduction Buprenorphine (as Vetergesic, Alstoe Animal Health, UK) has UK marketing approval for use in dogs as a pre- and post- operative analgesic and sedative at a dosage of 1 2 lg kg )1 of buprenorphine. Toxicological studies have indicated there is a good safety margin for the drug and few problems are to be expected from doses above these approved levels, with single doses at 8 lg kg )1 being well tolerated locally and systemically (Alstoe 584

2 Animal Health, UK, data on file). However, clinical studies have not been conducted to examine whether the dose response curve of buprenorphine in dogs continues to rise with dosage above the approved level. This study was intended to establish whether there are additional clinical benefits to be obtained from using a dose of 4 lg kg )1 compared to a dose of 2 lg kg )1 for the prevention of moderate to severe post-operative pain. Ovariohysterectomy is a surgical procedure that is considered to give rise to moderate to severe, abdominal post-operative pain (Capner et al. 1999; Coleman & Slingsby 27). Pain under clinical conditions may be assessed using a variety of methods, including assessment of the patient s behaviour and a variety of nociceptive threshold testing devices. Behavioural assessment methods for pain and sedation commonly use both observation and interaction with the animal and are scored on systems such as the Visual Analogue Scale (VAS). These are generally subjective and can suffer from a lack of consistency between observers, although this can be minimised by training the observers or by using a single observer. Mechanical threshold testing devices have the advantage that they are more likely to produce an objective measurement of the animal s threshold response to a noxious mechanical stimulus and allow assessment of both central and peripheral sensitisation after surgery. Patients generally will become more sensitive to a mechanical stimulus at the wound site after surgery; however different measurement methods and the individual variation in mechanical thresholds (measured by any method) mean that there may not be a direct correlation between the threshold measured and pain. This study was designed to test the null hypothesis that there is no difference in effect between 4 lg kg )1 compared with 2 lg kg )1 buprenorphine administered at the time of premedication. Material and methods This study was conducted between August 26 and October 27 under an Animal Test Certificate issued to Alstoe Animal Health by the UK Veterinary Medicines Directorate (which performs ethical review as part of this process) and under UK Home Office licence (ethical approval is required internally from the Institutional Ethical Review Body and externally from the UK Home Office Inspectorate). Twenty-six client owned female dogs entering the neutering clinic of the University of Bristol Veterinary School for elective ovariohysterectomy were recruited for the study. Informed owner consent was obtained and all animals were permanently identified with a microchip or ear tattoo. All dogs were examined clinically prior to entering the study to ensure that they were temperamentally suitable to undergo the experimental protocol and also that they had a physical status of 1 or 2 according to the scale developed by the American Society of Anesthesiologists, and that had not received analgesic treatment in the previous 7 days. Pre anaesthetic heart rate (HR), respiratory rate (f R ) and rectal temperature were recorded. Dogs were allocated randomly into two treatment groups (using block randomisation); B2 and B4 that received buprenorphine at 2 lg kg )1 and 4 lg kg )1 respectively with the investigator (LS) unaware to the treatment allocation. The dogs had been fasted overnight, and water was withheld from the time of premedication. Anaesthetic premedication was with acepromazine (ACP, 2 mg ml )1 ; C-Vet, UK).3 mg kg )1 and the relevant dose of buprenorphine (Vetergesic.3 mg ml )1 ; Alstoe, UK) administered intramuscularly (IM) into the quadriceps femoris muscle as two separate injections. Heart rate and f R were recorded immediately prior to induction of anaesthesia. After a minimum period of 3 minutes, an intravenous catheter was placed in the cephalic vein and anaesthesia was induced with propofol given to effect (Rapinovet Schering Plough, UK). A cuffed endotracheal tube was placed and then anaesthesia was maintained via a non- rebreathing circuit [T-piece, Fresh Gas Flow (FGF) 3 minute volume (MV) in small dogs and parallel Lack, FGF at 1 MV (MV calculated as 2 ml kg )1 minute )1 ) for larger dogs] with isoflurane vaporised in oxygen. Normal saline was administered at 1 ml kg )1 hour )1 during surgery. During anaesthesia HR, f R, blood pressure (indirect oscillametric method: systolic, diastolic and mean arterial pressure), haemoglobin oxygen saturation (SpO 2 ) end-tidal CO 2 (PE CO 2 ), and anaesthetic concentration (inspired and expired) were recorded at 5 minute intervals and were reported at 5 and 15 minutes after induction and thereafter at 15 minute intervals; ECG was monitored and any abnormalities recorded. Gas was sampled at the endotracheal tube connector. All monitoring was with a Datascope Passport 2 (Datascope Patient Ó 211 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesiologists, 38,

3 Monitoring) which was calibrated regularly according to the manufacturer s instructions. Ovariohysterectomy via a midline incision was performed by final year veterinary students under direct supervision of a veterinary surgeon. Postoperatively HR and f R were recorded at 2, 3, 4, 5 6, 7, 8 and 1 hours after administration of buprenorphine. Visual Analogue Scale (VAS) scores for pain and sedation were recorded on a 1 mm scale with = no pain/no sedation and 1 = worst possible pain for procedure/full sedation with no response to hand clap made over dog s head. Scores were awarded after the investigator had both observed and handled the patient. VAS scoring occurred at baseline (T =, immediately prior to administration of premedication and buprenorphine), immediately prior to induction of anaesthesia (minimum time T =.5) then postoperatively at T =3,4,5,6,7, 1, and 22 hours. Post surgery sensitisation was investigated using mechanical nociceptive thresholds (MNT). MNT were measured using the Constant Rate of Application of Force (CRAF) device first described by Lascelles et al. (1997). This consists of a blunt ended probe with a 2 mm diameter hemispherical tip mounted in a pen- like handle containing a load cell. When force is applied to the probe the load cell produces a voltage output which is transmitted via a reinforced cable to a data logging unit which transforms the output into a reading in Newtons. The system is linked to a rate measuring device which the investigator uses to maintain a constant rate of application of force (approximately 1.3 N per second equivalent to 4.2 kg-force cm )2 second )1 ). The device was calibrated to 15 N using an accurate load beam. MNT were measured at the wound (=ventral midline at the midpoint of the surgical wound 1 cm from wound margin) and the hock (on the lateral aspect of the fifth metatarsal just distal to the hock) holding the device perpendicular to the test area. The MNT was taken as at the point when the dog responded to the stimulus e.g. by withdrawing the area or turning to bite. If the dog did not allow testing of the area the MNT was deemed to be. MNTs were measured at the time of the VAS scoring. The calculation of analgesic effect and duration was by VAS score (scores <5 were acceptable and 5 were unacceptable) and requirement of additional (rescue) analgesia. If, in the opinion of the investigator, an animal was in unacceptable pain (corresponding to a VAS pain score of 5 or above), rescue analgesia with carprofen 4 mg kg )1 (Rimadyl, Pfizer Animal Health, UK) was administered subcutaneously. If the animal required additional analgesia after this first rescue intervention it was removed from the study and subsequent treatment was left to the clinical judgement of the investigator. Animals that had been administered rescue analgesia continued to be observed regularly by the investigator at a minimum of those time points outlined in the study protocol. Dogs were returned to their owners the morning after surgery with carprofen tablets to be administered twice daily at 2 mg kg )1 for 3 days. Data were examined using Graph Pad Prism 5.3 (Graph Pad Software, Inc., CA, USA). Distribution free methods (non parametric tests) were used for VAS scores as they are unlikely to be normally distributed. VAS pain scores were compared at the 3 7 hours time points using Mann Whitney tests with a Bonferroni correction for five multiple tests (i.e. p <.1 is significant). No comparisons were made beyond this time as most dogs had then received rescue analgesia. In order to adjust for the use of rescue analgesia, data was analysed as recorded and also using the last observation carried forward (LOCF) method. VAS sedation scores were compared at the 1 hour point (before anaesthesia) and then at 3 7 hours as for pain scores. The dogs started the study pain and sedation free (i.e. VAS scores were zero at baseline) so no comparisons over time within groups were made, as they would not give meaningful results. A difference of 15 in the VAS pain score was deemed to be clinically relevant, which could be demonstrated by a group size of 13. The authors used their own previously published data (Slingsby & Waterman-Pearson 2, 21) to calculate a sample size of 13 based on a difference in VAS pain score of 15 (deemed to be clinically relevant) taking the alpha (a) level (degree of type 1 error) as.5 and the beta (b) level (degree of type 2 error) as.1 (therefore power = 9%) (Stolley & Strom 1986). The combined effects of treatment and time on mechanical thresholds were examined using two way repeated measures analysis of variance (RM ANOVA). Within group comparisons for wound thresholds were made using one way RM ANOVA with post hoc (Dunnett) comparisons to the baseline. Requirement for rescue analgesia was compared using a chi-squared test for small samples (Fisher s exact test). Comparison between single data values obtained for each animal (e.g. age, weight, propofol induction 586 Ó 211 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesiologists, 38,

4 dose) were made using independent samples t-tests. Repeated measures values obtained during anaesthesia and postoperative HR and f R were compared using repeated measures ANOVA followed by post hoc Bonferroni tests. Results Unless otherwise stated values are presented as mean (SD). Twenty six dogs were recruited for the study; no dogs were excluded or failed to complete the study for any reason. On preliminary examination of the data one dog (dog 16) which had been allocated to the B2 group actually was administered B4. All the data from dog 16 were then assigned to the B4 group, giving a total of 12 cases in the B2 group and 14 cases in the B4 group. The two groups were similar for age: 3 (2) years, 3 (3) years; and weight: 22 (11) kg, 16 (1) kg; time from premedication to induction of anaesthesia was 59 (19) minutes, 57 (15) minutes; surgery duration was 67 (16) minutes, 61 (19) minutes; time from premedication to end of surgery was 159 (23) minutes, 145 (29) minutes in the B2 and B4 groups respectively. Pre- anaesthetic check HR was 15 (17) and 112 (22) beats per minute, f R 24 (6) and 23 (1) breaths per minute, rectal temperature was 38.8 (.5) C and 38.6 (.7) C in the B2 and B4 groups respectively. Although there was a trend for a higher dose of propofol to be required for induction of anaesthesia in the B2 [3.3 (1.8) mg kg )1 ] group compared to the B4 group [2.9 (.5) mg kg )1 ], it was neither clinically or statistically significant. Requirement for isoflurane was similar between groups (Table 1). SpO 2 remained within clinically acceptable limits throughout anaesthesia (96 1%). PE CO 2 was at the higher end of the clinically acceptable range (i.e. under 5.3 kpa) for anaesthesia in the B2 group; in the B4 group the mean values were 5.3 kpa at all but the 15, 3 and 6 minute points but there were no significant differences between groups at any time point. There were no significant differences between groups or over time for f R. Hypotension (defined as mean arterial pressure <7 mmhg (McKelvey & Hollingshead 23) was present for the first 3 minutes in both groups; although blood pressure varied over time within groups there were Table 1 Physiological intra-operative variables and number of dogs remaining within each treatment group over time ( 15 minutes) after administration of buprenorphine at time of premedication at 2 lg kg )1 (B2) or 4 lg kg )1 (B4) Dose Number of dogs remaining B B Resp rate (breaths minute )1 ) B2 15 ± 7 14 ± 7 15 ± 8 12 ± 5 13 ± 5 1 ± 5 15 ± ± 8 B4 21 ± ± 2 24 ± 2 23 ± 2 12 ± 6 17 ± 15 9 ± 5 11 ± 9 Heart rate (beats minute )1 ) B2 88 ± ± ± 28* 96 ± ± ± ± ± 22 B4 72 ± ± ± 16 9 ± ± ± ± ± 7 PE CO 2 (kpa) B2 5.3 ± ± ± ± ± ± ± ± 1. B4 4.5 ± ± ± ± ± ± ± FE Iso (%) B2.9 ± ± ± ± ± ± ± ±.3 B4 1. ± ± ± ± ± ± ± FI ISO (%) B2 1.5 ± ± ±.5 2. ±.4 2. ± ± ± ±.3 B4 1.5 ± ± ± ± ± ± ± SPO 2 (%) B2 98 ± 2 99 ± 1 99 ± 2 99 ± 1 99 ± 1 99 ± 1 99 ± 1 99 ± 1 B4 98 ± 2 98 ± 3 99 ± 1 99 ± 1 99 ± 1 99 ± 1 99 ± 1 1 M NIBP (mmhg) B2 58 ± 8 64 ± 7 67 ± 12 8 ± ± ± ± 17 8 ± 16 B4 64 ± ± ± ± ± ± 15 8 ± D NIBP (mmhg) B2 35 ± 8 37 ± 8 4 ± ± 21 6 ± ± 9 49 ± 13 5 ± 1 B4 39 ± 16 4 ± ± ± ± ± 1 54 ± S NIBP (mmhg) B2 15 ± 1 19 ± ± ± ± ± ± ± 26 B4 112 ± ± ± ± ± ± ± Data presented as mean ± SD. PE CO 2, end-tidal CO 2 ;FE Iso, end-tidal isoflurane concentration; FI ISO, inspired isoflurane concentration; SPO 2, haemoglobin oxygen saturation; NIBP, non invasive blood pressure; M, mean; D, diastolic; S, systolic. *B2 < B4 post hoc Bonferroni test. Ó 211 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesiologists, 38,

5 no significant differences between groups. Heart rate during anaesthesia was the only physiological variable that varied significantly both with time and treatment; it was higher in the B2 group for the first 3 minutes of anaesthesia (significantly so at 3 minutes only). Immediately prior to induction and postoperatively both time (p <.1) and treatment (p =.2) had significant effects on HR but there were no post hoc significant differences between groups (Table 2). Treatment (p =.13) had significant effect on f R immediately prior to induction and postoperatively but time was not significant, again there were no post hoc differences between groups. Heart rates and f R were clinically acceptable at all time points in both groups (Haskins 1999). During anaesthesia, ECG recordings were normal in 1 of 12 dogs in the B2 group and 12 of 14 dogs in the B4 group. In the B2 group one dog initially had an irregular rhythm that was not associated with ventilations, but no abnormal complexes thereafter which progressed to sinus arrhythmia at 2 minutes, and one dog had a marked sinus arrhythmia from 1 to 2 minutes. In the B4 group one dog had marked sinus arrhythmia throughout the procedure with a couple of episodes of second degree AV block at 1 minutes which resolved without treatment and one dog had intermittent missing P waves at 1 to 2 minutes, followed by normal sinus arrhythmia from 25 to 35 minutes; thereafter the ECG was unremarkable. Immediately prior to induction of anaesthesia median (range) VAS sedation scores were 31 (7 64) and 35 (19 66) in the B2 and B4 groups respectively. Postoperatively sedation scores were high initially and then fell over time (Fig. 1) with no significant differences between groups at any time. VAS pain scores were zero at baseline and before induction of anaesthesia. In the postoperative period the overall average pain score (when uncorrected for rescue analgesia) remained stable and below 4 in both groups (Fig. 2). The main effect of using the last observation carried forward (LOCF) transformed VAS Pain score (Fig. 3) was to show a steady rise in pain scores up to 6 or 7 hours; at which point they reached a plateau (as most dogs had received rescue analgesia by this point). There were no significant differences in pain scores as recorded or those adjusted using LOCF between groups at the comparison points at 3 7 hours. MNT at the hock (Fig. 4) and at the wound (Fig. 5) at baseline, immediately prior to test drug administration, and immediately prior to induction of anaesthesia did not differ significantly between groups. Equally there were no significant differences when comparing within groups for any of these three time points for the B2 or B4 groups. For hock and wound MNT time was a significant factor (both p <.1) but treatment was not significant (both p =.7). The effect of time was examined further within each group using one way Table 2 Heart rate and respiratory rate immediately before induction of anaesthesia and postoperatively (P = time of premedication) and number of dogs in each treatment group Immed pre op P+3 P+4 P+5 P+6 P+7 P+8 P+1 Number of dogs B2 1* B4 13* * * 14 Resp rate (breaths minute )1 ) B2 27±14 19±6 22±9 26±7 22±7 24±8 26±9 27±11 B4 27 ± ± ± 8 26 ± 9 33 ± ± ± ± 18 Number of dogs B B Heart rate (beats minute )1 ) B2 86 ± ± ± ± ± ± ± ± 18 B4 79 ± ± 34 9 ± ± 12 8 ± ± ± ± 14 *Dogs were panting so no respiratory rate recorded. Dogs were still under anaesthesia. 588 Ó 211 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesiologists, 38,

6 VAS sedation score B2 B4 VAS pain score B2 B Time (hours) Time (hours) Figure 1 Median (range) VAS sedation scores after administration of buprenorphine at time of premedication at 2 lg kg )1 (B2, n = 12, range indicated by larger T) or 4 lg kg )1 (B4, n = 14, range indicated by smaller T). NB anaesthesia and surgery occurred between approximately hour. VAS pain score B2 B Time (hours) Figure 2 Median (range) VAS pain scores after B2 (n = 12) or B4 (n = 14). NB anaesthesia and surgery occurred between approximately hours. RM ANOVA. Wound MNT were significantly lower than baseline from 4 to 3 hours to end of testing at 22 hours for B2 and B4 respectively (most post hoc tests p <.1). Hock MNT were significantly lower than baseline at 22 hours for B2 (p <.5) but not significantly lower for B4. Time for first rescue analgesia is shown in Fig. 6. Eleven of the 12 dogs (92%) in the B2 group and 11 of the 14 dogs in the B4 group required rescue analgesia (79%) Of these, one dog in the B2 group required two rescues (at 6 and 7 hours), and one dog in the B4 group required rescue at 5 and 7 hours; in dogs that required a second rescue this was achieved with buprenorphine. Of the dogs that required rescue analgesia, the mean (SD) time of rescue was 5 (2) hours in the B2 group and 5 Figure 3 Median (range) VAS pain using LOCF after rescue analgesia after B2 (n = 12) or B4 (n = 14). NB anaesthesia and surgery occurred between approximately hours. Hock MNT (N) (2) hours in the B4 group. Using Fisher s Exact Test there were no significant differences between groups for overall number of dogs requiring rescue analgesia. Discussion B2 B Time (hours) Figure 4 Mean (SD) Hock MNT after B2 (n = 12) or B4 (n = 14). NB these data are not corrected for rescue analgesia. NB anaesthesia and surgery occurred between approximately hours. There is no consensus opinion in which methods are the best and most reliable for scoring pain in either human or veterinary analgesia. In this study analgesia and sedation were assessed using 1 mm VAS based on behavioural and dynamic observations. This method was chosen as there was a single observer who was experienced with VAS scoring. Alternative methods would have been a simple descriptive scale (SDS) assigning descriptors to a numerical category or a composite pain scale. Ó 211 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesiologists, 38,

7 Wound MNT (N) B2 B Time (hours) Figure 5 Mean (SD)Wound MNT after B2 (n = 12) or B4 (n = 14) NB these data are not corrected for rescue analgesia. NB anaesthesia and surgery occurred between approximately hours. All dogs were well sedated prior to induction of anaesthesia and sedation was very similar to that described when alpha 2 agonists such as medetomidine are used for premedication (Bufalari et al. 1996; Ko et al. 2); this degree of sedation was ideal for easy IV catheterization and anaesthetic induction. In our study dogs, the good sedation (with some ataxia) facilitated catheter placement and subsequent induction of anaesthesia. The dose of propofol required for induction of anaesthesia was similar to induction doses required after premedication with IV medetomidine or dexmedetomidine (Kuusela et al. 21) and similar to that of a comparable study using acepromazine.5 lg kg )1 and buprenorphine 2 lg kg )1 premedication (Shih et al. 28). The higher dose of buprenorphine did not significantly reduce the induction dose; this lack of additional anaesthetic sparing effect was also seen with thiopentone using buprenorphine at 1 or 2 lg kg )1 (Slingsby et al. 26). Inspired and expired isoflurane concentrations were not different between groups. However sampling at the endotracheal connector, particularly using high FGF may have affected the accuracy of the measurements as the sample could have been contaminated with fresh gas. The mean end-tidal isoflurane concentrations (1.4%) were slightly higher than the 1.1% seen in the similar study by Shih et al. (28) which may be due to anaesthetist practice or differences in sampling and measurement techniques. Increasing the dose from 2 to 4 lg kg )1 appeared to cause little or no effect on all measured cardiovascular and respiratory variables. Heart rate increased during anaesthesia compared to just before induction of anaesthesia whereas f R decreased during anaesthesia compared to immediately before induction. Blood pressure, blood haemoglobin oxygen saturation, end-tidal CO 2, HR, respiratory rate, and blood pressure remained at acceptable levels during anaesthesia. The values for HR during anaesthesia in B2 were initially lower than that described by Shih et al. (28) for their buprenorphine groups but may reflect our higher isoflurane use. Respiratory rates and systolic blood pressure were very similar between the two studies. Although buprenorphine (at 16 lg kg )1 IV) has been demonstrated to change various cardiovascular parameters during isoflurane anaesthesia, these changes, while statistically significant, were not clinically important in healthy dogs (Martinez et al. 1997). Postoperatively HR and f R were not significantly different between the two groups and were within normal limits (Haskins 1999). The ECG was normal in 22 of the 26 dogs in the study. In the four dogs that showed ECG changes, these were mostly in the early portion of the 6 5 B2 B4 Number of dogs Figure 6 Time to first rescue analgesia Time of first rescue analgesia from administration of buprenorphine 59 Ó 211 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesiologists, 38,

8 anaesthetic period. Although the study defined anything deviating from a standard ECG trace at regular rhythm as abnormal, sinus arrhythmia is a physiologically normal finding in many dogs and occasional second degree AV block (while commonly associated with opioid administration) can also be seen in physiologically normal dogs with sinus arrhythmia (Newton et al. 197). The incidence of ECG abnormalities in the general healthy dog population and in anaesthetised dogs in particular is unknown. None of the ECG abnormalities appeared to affect the cardiovascular system functionally, and, in the opinion of the anaesthetist, none required treatment. Analysis of VAS pain data can be difficult as it is unlikely to be normally distributed, is a repeated measure over time and will be influenced by use of additional (rescue) analgesia. We chose to compare the two groups at specific time points with and without correction for rescue analgesia and did not demonstrate significant differences between groups. A type 2 error may have occurred (no significant differences were demonstrated in the study although in reality they might exist) possibly due to insensitivity of the scoring system. The combination of physiological and emotional components in pain make it is a difficult thing to assess. There will always be problems when a subjective scoring system is used. We used a VAS scoring system performed by a single experienced observer which involved visual observation and interaction with the animal, encouraging it to move and palpation of the wound area; Lascelles et al. (1998) have called this approach DIVAS (dynamic interactive visual analogue scale). This observer anchors in her mind the appearance of animals at the two end points ( no pain which is easy to envisage as all dogs present as this prior to the start of the study and 1 worst possible pain based on the clinician s clinical experience) and the rescue point (at which the animal first appears to be clinically uncomfortable and requiring rescue analgesia); she awards her score on the VAS line based on the visual and dynamic interactions. The disadvantage of using a single observer is that the score is based on a single opinion, it is impossible to tell if the observer is more or less sensitive to the signs of pain exhibited by an animal than a general population of clinicians. Shih et al. (28) used two independent observers who were experienced in the interpretation of signs of pain in dogs and separated their VAS scores into three phases (observation only, walk and move, palpation of the wound); their results confirmed that all three of their phases were important in assessment of pain and that there was significant inter observer variation even among these experienced people. Additionally they used the Glasgow Composite Pain Score (GCMPS) but based their decision for rescue analgesia on clinical signs and found that the GCMPS score for rescued animal ranged between 6 and 17 and discuss in their paper the problem if the scoring systems does not indicate that rescue analgesia is needed but the clinician wants to give additional analgesia. In the current study rescue analgesia was administered to 92% (11 of 12) dogs in the B2 group of which 67% (eight of 12) required this rescue at or before 6 hours compared to 25% (five of 2) of buprenorphine treated dogs (25%) requiring rescue at or before 6 hours in the Shih et al. (28) study. This difference may be due to the observer in our study being more sensitive to signs of pain and also possibly may relate to nociceptive stimulus during surgery (in our study surgery was performed by students under supervision rather than more experienced clinicians). This study shows that the anticipated re-dosing interval should be 4 6 hours if buprenorphine is used alone as an analgesic after ovariohysterectomy in the dog. Almost all dogs required rescue analgesia at a mean time of 5 hours after initial administration of buprenorphine, irrespective of dose. The actual VAS pain scores over time when scores were not corrected for use of rescue analgesia remained at a mean of 4 or below, which indicates good analgesia (if the rescue point is 5). Most dogs did not have a VAS score much above 5 when rescue was required and all but two dogs (one in each group) were adequately rescued with carprofen. The two dogs that required further rescue analgesia both appeared to be comfortable after additional buprenorphine at 2 lg kg )1 was administered. As an additional finding we have confirmed the perception (Capner et al. 1999) that ovariohysterectomy is a more painful procedure than castration. In a previous study we investigated male dogs undergoing castration using buprenorphine at 1 or 2 lg kg )1 at premedication with second dose of buprenorphine at 4 (B1) or 6 (B2) hours after administration of the first dose (Slingsby et al. 26). In that study fewer dogs required rescue analgesia; three of 14 dogs (at 3, 3, and 5 hours) in the B1 group and one of 14 dogs (at 5 hours) in Ó 211 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesiologists, 38,

9 the B2 group. The overall VAS pain scores also tended to be lower (mean around 3). Most dogs in the current study required rescue analgesia at 5 hours and the mean VAS pain score was 4. Single drug and dose analgesic comparison studies may not necessarily reflect clinical practice but are required to elucidate the effects of individual drugs, doses and routes and may be required for regulatory approval. In these studies it is important to consider rescue analgesia (Slingsby 21). The regulatory authority (the VMD) required the first line rescue analgesic agent to be a different agent to the test drug, be a UK licensed analgesic drug for the dog and used in accordance to its licence. The choices of licensed analgesic drug were pethidine (meperidine) at 3.3 mg kg )1 or a variety of NSAIDs (non steroidal anti inflammatory drugs). The decision was made to use the NSAID carprofen as it offers similar or better intensity of analgesia with longer duration of action compared to pethidine (Lascelles et al. 1994; Slingsby & Waterman-Pearson 21). It may be argued that the IV route of administration should have been used if carprofen was required for the rescue analgesia as uptake of drugs may be slower and more variable after SC compared to IV administration. However, there are no publications comparing clinical efficacy of carprofen administered SC versus IV at any time in the peri operative period or any direct pharmacokinetic studies of these two routes. The subcutaneous route was chosen as this was the usual route of administration in the clinic. After rescue analgesia the dogs were observed carefully and if necessary additional rescue analgesia could be administered with the choice of drug left entirely to the clinical judgement of the investigator. Buprenorphine may have prevented central but not peripheral sensitisation. Previous studies have demonstrated reduced postoperative pain when pethidine (Lascelles et al. 1997) and carprofen (Welsh et al. 1997; Lascelles et al. 1998) were administered before surgical stimulation compared to after surgical stimulation. Although higher drug plasma concentrations in the early postoperative period may be a contributing factor to this finding Lascelles et al. (1997), demonstrated that, compared to dogs given pethidine postoperatively or un treated controls, preoperative administration of pethidine significantly reduced pain scores at 8, 12 and 2 hours and significantly prevented postoperative hypersensitivity at a site distant to the wound at 12 and 2 hours. These effects occurred long after the expected clinically analgesic effect of this drug and support the pre-emptive analgesic theory proposed. Hypersensitivity at the operative site (wound) after surgery occurs as a result of peripheral sensitisation (although central changes may also contribute to this process). In addition to causing hypersensitivity at the wound, surgery may also initiate central changes (central sensitisation) that result in areas distant to the operative site (in our case the hock) also becoming hypersensitive to mechanical stimulation (Coderre et al. 1993).). Buprenorphine at either dose did not prevent MNT decreasing at the wound but thresholds at the hock did not decrease after surgery. This suggests that, like pethidine, buprenorphine was able to block the development of central sensitisation by acting as a pre-emptive analgesic. There were no increasing physiological effects, as a result of doubling the dose of buprenorphine from 2 to 4 lg kg )1. Analgesia did not appear to be improved with increased dose but this may relate to insensitivity of scoring small changes in pain between dogs. This suggests that for ovariohysterectomy in bitches there is a ceiling effect on potential adverse events (such as respiratory depression and bradycardia) and possibly also analgesia. However a full dose response curve would need to be constructed using a wider variety of doses to confirm this finding. We would recommended that doses of buprenorphine of 2 lg kg )1 or higher should be administered to bitches undergoing ovariohysterectomy and presence of pain should be carefully monitored as additional dosing may be required earlier than is generally recommended. Rodent models have demonstrated that as the intensity of the noxious stimulus increases more buprenorphine is required to achieve a similar effect (Raffa & Ding 27). Conclusion Both doses of buprenorphine provided good preoperative sedation and postoperative analgesia for approximately 5 hours after administration. The results could indicate a ceiling effect for analgesia after ovariohysterectomy in dogs but this requires further investigation. Funding This study was funded by Alstoe Animal Health, York, UK. 592 Ó 211 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesiologists, 38,

10 References Bufalari A, Short CE, Giannoni C et al. (1996) Comparative responses to propofol anaesthesia alone and with alpha 2-adrenergic medications in a canine model. Acta Vet Scand 37, Capner CA, Lascelles BD, Waterman-Pearson AE (1999) Current british veterinary attitudes to perioperative analgesia for dogs. Vet Rec 145, Coderre TJ, Katz J, Vaccarino AL et al. (1993) Contribution of central neuroplasticity to pathological pain: review of clinical and experimental evidence. Pain 52, Coleman DL, Slingsby LS (27) Attitudes of veterinary nurses to the assessment of pain and the use of pain scales. Vet Rec 16, Haskins S (1999) Perioperative monitoring. In: Manual of Small Animal Anesthesia. Paddleford R (ed.). W.B Saunders Company, Philadelphia. pp Ko JC, Mandsager RE, Lange DN et al. (2) Cardiorespiratory responses and plasma cortisol concentrations in dogs treated with medetomidine before undergoing ovariohysterectomy [see comment]. J Am Vet Med Assoc 217, Kuusela E, Raekallio M, Vaisanen M et al. (21) Comparison of medetomidine and dexmedetomidine as premedicants in dogs undergoing propofol-isoflurane anesthesia. Am J Vet Res 62, Lascelles BD, Butterworth SJ, Waterman AE (1994) Postoperative analgesic and sedative effects of carprofen and pethidine in dogs. Vet Rec 134, Lascelles BD, Cripps PJ, Jones A et al. (1997) Post-operative central hypersensitivity and pain: the pre-emptive value of pethidine for ovariohysterectomy. Pain 73, Lascelles BD, Cripps PJ, Jones A et al. (1998) Efficacy and kinetics of carprofen, administered preoperatively or postoperatively, for the prevention of pain in dogs undergoing ovariohysterectomy. Vet Surg 27, Martinez E, Hartsfield S, Melendez L et al. (1997) Cardiovascular effects of buprenorphine in anesthetized dogs. Am J Vet Res 58, McKelvey D, Hollingshead KW (23) General anesthesia. In: Veterinary Anesthesia and Analgesia. McKelvey D, Hollingshead KW (eds). Mosby, St. Louis, MI, USA. pp Newton JEO, Murphree OD, Dykman RA (197) Sporadic transient atrioventricular block and slow heart rate in nervous pointer dogs. Integr Psychol Behav Sci 5, Raffa RB, Ding Z (27) Examination of the preclinical antinociceptive efficacy of buprenorphine and its designation as full- or partial-agonist. Acute Pain 9, Shih AC, Robertson S, Isaza N et al. (28) Comparison between analgesic effects of buprenorphine, carprofen, and buprenorphine with carprofen for canine ovariohysterectomy. Vet Anaesth Analg 35, Slingsby L (21) Considerations for prospective studies in animal analgesia. Vet Anaesth Analg 37, Slingsby LS, Waterman-Pearson AE (2) The postoperative analgesic effects of ketamine after canine ovariohysterectomy a comparison between preor post-operative administration. Res Vet Sci 69, Slingsby LS, Waterman-Pearson AE (21) Analgesic effects in dogs of carprofen and pethidine together compared with the effects of either drug alone. Vet Rec 148, Slingsby LS, Taylor PM, Waterman-Pearson AE (26) Effects of two doses of buprenorphine four or six hours apart on nociceptive thresholds, pain and sedation in dogs after castration. Vet Rec 159, Stolley P, Strom B (1986) Sample size calculations for clinical pharmacology studies. Clin Pharmacol Ther 39, Welsh EM, Nolan AM, Reid J (1997) Beneficial effects of administering carprofen before surgery in dogs. Vet Rec 141, Received 11 June 21; accepted 26 January 211. Ó 211 Association of Veterinary Anaesthetists and the American College of Veterinary Anesthesiologists, 38,