Attitudes and concerns of French veterinarians towards pain and analgesia in dogs and cats

Veterinary Anaesthesia and Analgesia, 2004, 31, 154 163 RESEARCH PAPER Attitudes and concerns of French veterinarians towards pain and analgesia in dogs and cats Marine Hugonnard * DV, DEA, Agnès Leblond DV, Thèse, Stéphanie Keroackà DMV, IPSAV, DEA, Jean-Luc Cadoré * DV, Diplomate ECVIM-CA & Eric Troncyà DV, MSc, PhD * Ecole Nationale Vétérinaire de Lyon, Small Animal Department, Unit of Internal Medicine, Marcy L Etoile, France Ecole Nationale Vétérinaire de Lyon, Equine Department, Unit of Epidemiology/Biomathematics, Marcy L Etoile, France àecole Nationale Vétérinaire de Lyon, Small Animal Department, Unit of Surgery/Anaesthesiology, Marcy L Etoile, France Correspondence: Eric Troncy, Department of Veterinary Biomedicine, Faculté de Médecine Vétérinaire, Université de Montréal, C.p. 5000, Saint-Hyacinthe, Québec, Canada J2S 7C6. E-mail: Eric.Troncy@UMontreal.CA Abstract Objective To compare the attitudes of French veterinarians to pain, and their provision of analgesia to animals, with that reported from other countries. Study design Epidemiological study. Methods In June 1999, 379 French veterinarians were surveyed to ascertain their views on pain evaluation and control in dogs and cats, and their use of analgesics in daily practice. Survey results are expressed as a percentage of responses. Results The questionnaire was returned by 189 veterinarians (49.9%). The response rate was influenced by age (younger veterinarians were more likely to respond) but not gender. A majority (99.5%) expressed moderate to extreme concern over pain in their patients. Pain evaluation was based on the animal s attitude (88.3% dogs, 82.5% cats), interaction with the caregiver, response to palpation of the painful area (66.5% dogs, 62.7% cats) and inappetence (29.3% dogs, 46.3% cats, p < 0.001). Only 14.3% of respondents considered their knowledge of pain recognition to be inadequate. Many (58.8%) considered their methods of pain quantification and control (47% dogs, 59% cats) to be inadequate. Difficulties in recognizing pain (58.3%), a lack of knowledge in the appropriate use of analgesics (41.7%) and fear of drug side effects (30%) were used to explain inadequate provision of analgesia. Only 16.1 and 8.1% used opioids indogsandcats, respectively. Thislowlevelof useresulted from the imposition of French narcotic legislation (79.9%) and lack of knowledge of opioid pharmacology (73.7%). Nonsteroidal anti-inflammatory drugs (NSAIDs) and glucocorticoids were the most commonly used analgesics in both species (100% (dogs) and 96.7% (cats)). The most popular NSAID used in France was tolfenamic acid, followed by meloxicam (dogs), ketoprofen, nimesulide (cats) and carprofen (dogs). The type of surgery performed influenced the use of analgesics, from 17.2% for castration to 83.7% for orthopaedic procedures. Nonsurgical conditions believed to warrant analgesia included osteoarthritis (97.8%), trauma (97.3%) and bone neoplasia (93.4%). Female veterinarians were more likely than males to evaluate pain and provide analgesia. Conclusion French practitioners demonstrate a level of interest in analgesia, which appears to be at least equivalent to that reported from Englishspeaking countries. The signs used to indicate the presence of pain do not, in general, appear to differ. Excessive confidence in their ability to recognize pain (despite a general ignorance of the subject), the minor role of animal health technicians in pain management and misconceptions about analgesics (mainly opioids) are French pecularities. Keywords analgesia, cat, dog, epidemiological survey, France, pain. 154

Introduction Although pain is one of the commonest symptoms encountered in human and veterinary medical practice, little attention has been paid to its assessment and alleviation in animals and children until recently, particularly in Latin countries (Bertolini et al. 2002). The reasons for this are complex: ideological obstacles, e.g. belief that pain is natural or that animals do not feel pain, failure to recognize pain, ignorance of its deleterious effects, fear of analgesic side effects and toxicity, and lack of familiarity with current therapies probably contribute to the undertreatment of pain in companion animals. Ethical considerations, the high level of concern expressed by pet owners towards pain control and growing evidence supporting the medical importance of analgesia are valuable reasons for rejecting previous beliefs. The recognition and treatment of pain has been held to represent an important indicator of the quality of care delivered to (human) hospitalized patients (Klopfenstein et al. 2000). Little information is currently available regarding the attitudes of veterinarians towards pain evaluation and relief, suggesting that pain may be frequently ignored, under-diagnosed or mistreated (Hansen & Hardie 1993). Only a few observational studies have specifically investigated the extent and adequacy of analgesic use within the veterinary profession; most have focused on perioperative analgesia (Watson et al. 1996; Dohoo & Dohoo 1996a,b; Capner et al. 1999). To our knowledge, this paper represents the first survey conducted in a Latin country investigating French veterinary attitudes towards pain recognition and relief in dogs and cats undergoing surgery, and with medical conditions. The purposes of our study were descriptive and observational, comparative and sociological. By considering results previously obtained in English-speaking countries, we attempted to determine if concern over pain was universal among veterinarians, or influenced by cultural and educational factors unique to France. The study was also designed to identify areas where ignorance and/or ideological barriers to pain relief in animals may exist. Another objective was to use the information obtained as a guide to direct clinical research and the nature of continuing education in this subject. Materials and methods Questionnaire The French Group of Clinical Research on Veterinary Analgesia (GRCAV) 1 developed a questionnaire designed to assess French attitudes to, and knowledge of, pain recognition and control in dogs and cats. All veterinarians involved in small animal practice (100% of working time involving dogs and/ or cats) in the Rhône-Alpes region 2 were eligible for inclusion (n ¼ 379). Interns and clinicians of the Ecole Nationale Vétérinaire de Lyon tested a prototype questionnaire; their comments on length, clarity and content were incorporated into a revised form, which was validated by a statistician. The questionnaire consisted of five parts. Part 1 collected demographic information about the veterinarians surveyed. Part 2 focused on pain assessment; veterinarians were asked about: (i) their level of concern in the areas of pain evaluation and control (extreme, moderate, limited or none); (ii) self-perception of ability to detect and quantify pain (very good, good, rather deficient, deficient); and (iii) opinions regarding factors affecting pain perception in animals. They were also asked to choose from a list of eight the three pain indicators they considered the most useful. Responses were recorded separately for dogs and cats. In analysing data, veterinarians were classified as self-confident in pain recognition, i.e. respondents who answered good or very good for ability to detect pain, or unconfident in pain recognition (those who declared a deficient or rather deficient ability to detect pain). This terminology was also applied to perceived ability to quantify pain. A veterinarian who was self-confident in pain recognition and in pain quantification was classified as self-confident in pain assessment. Part 3 focused on pain control. Veterinarians were asked whether they thought their use of analgesics was adequate in dogs and cats. Reasons for analgesic use or nonuse were evaluated by respondents indicating a choice from provided options. The extent of use of different analgesic classes, i.e. opioids, nonsteroidal antiinflammatory drugs (NSAIDs), dissociative agents, a 2 -agonists or glucocorticoids, were recorded 1 A copy of the survey is available from MH on request. 2 Data sources: Annuaires Vétérinaires ROY 1999, Editions du Point Vétérinaire, Maisons-Alfort, France. Ó Association of Veterinary Anaesthetists, 2004, 31, 154 163 155

separately for dogs and cats. Patterns of use of selected opioids and NSAIDs were also recorded for both species. For each NSAID, a qualitative indication of frequency of use: common ; rare ; or never and perceived safety: rather safe or rather unsafe was required. The extent of knowledge regarding contra-indications to NSAID use and opioid side effects was also gauged. Part 4 investigated pain assessment and analgesic use in practice. A list of 10 surgical procedures and 7 medical conditions was presented. For each, veterinarians were asked to rank pain on a scale of 1 10 (where 1 corresponded to no pain at all and 10 to the worst pain imaginable). Their score was compared to the mean score attributed for the same condition by three reference veterinarians (M.H., J.-L.C., E.T.). In each case, respondents were asked to indicate if they used analgesics, and if so, which ones. Species were not distinguished. A measurement of analgesic use was required to analyse factors distinguishing analgesic users from nonusers. For this purpose, 9 of the 17 painful situations occurring most commonly in practice were selected. A veterinarian who administered analgesics in at least five of these nine situations was defined as an analgesic user. In part 5, respondents evaluated the quality of veterinary teaching on the topic of pain and were questioned about their potential interest in continuing education on this subject. In June 1999, 379 questionnaires were sent out. A cover letter presented the goals of the study and assured participants that results would be confidential. A stamped pre-addressed envelope was included. A second mailing was sent to all practitioners 1 month after the first. Statistical analysis After alphanumerical coding, responses from completed surveys were entered and stored in a microcomputer database management program (Epi Info 6.04; Centres for Disease Control, World Health Organization, Geneva, Switzerland). Data were validated by manually checking a random sample of records. Descriptive statistics were carried out using Epi Info SAS software (SAS 6.12, SAS Institute, Cary, NC, USA) this was also used to study simple associations between variables. Fisher s exact twotailed tests, McNemar s tests, confidence intervals and Student s t-tests were applied where appropriate. Values of p < 0.05 were considered significant. Not all participants responded to all questions. Therefore, results are expressed as the exact number or percentage of occurrences for qualitative data referred to the number of responses for each question. Quantitative data are expressed as mean (±SD). Results The questionnaire was returned by 189 out of 379 veterinarians surveyed (49.9% response rate). Demographic characteristics of respondents are presented in Table 1. The men s response rate (130/ 275, 47.3%) was insignificantly (p ¼ 0.3) lower than that of female veterinarians (56/104, 53.8%). The response rate of veterinarians aged between 23 and 35 years (57/92, 62%) was significantly greater (p ¼ 0.006) than that of veterinarians aged between 35 and 50 years (109/229, 47.6%), or 50 and 65 years (21/58, 36.2%). Table 1 Demographic characteristics of 189 veterinarians responding to a survey of pain recognition and control in small animals Characteristic Distribution Gender (186 responses) 130 males (69.9%) 56 females (30.1%) Age (187 responses) Between 25 and 35 years 57 responses (30.5%) Between 35 and 50 years 109 responses (58.3%) Between 50 and 65 years 21 responses (11.2%) Practice size (188 responses) 107 practices with 1 to 2 veterinarians (56.9%) Animal health technician use (188 responses) 147 employed at least 1 AHT (78.2%) 156 Ó Association of Veterinary Anaesthetists, 2004, 31, 154 163

Pain assessment Most veterinarians felt moderately (103/189, 54.5%) or extremely (85/189, 45%) concerned about recognition and alleviation of pain. Level of concern was not influenced by gender (p ¼ 1), age (p ¼ 0.89) or the presence of an animal health technician (AHT) (p ¼ 0.58). A majority of respondents (Table 1) employed at least one AHT. Most (171/185, 92.4%) considered that pain evaluation was strictly the veterinarian s responsibility. A very small number (14/185, 7.6%) shared this task with their AHT. The existence of pain scales was ignored by 135 of 184 respondents (73.4%). Veterinarians generally considered their skills and knowledge in recognizing pain to be adequate. Most (162/189, 85.7%) were classified as self-confident in pain recognition. In contrast, a majority (111/ 189, 58.8%) had an unfavourable opinion of their ability to quantify pain, and were subsequently classified as not self-confident in pain quantification. Of the 189 respondents, 77 (40.7%) were classified as self-confident in pain assessment. Self-confidence in pain assessment was not influenced by gender (p ¼ 0.15), age (p ¼ 0.93), the presence of an AHT (p ¼ 0.59) or attending specific continuing education programmes within the previous 5 years (p ¼ 0.61). Compared with those classified as not self-confident, self-confident veterinarians were significantly more likely to regard their use of analgesics to be adequate in dogs (p ¼ 0.016) and in cats (p ¼ 0.001). However, they appeared no more likely to be analgesic users: the percentage of analgesic users among veterinarians self-confident in pain assessment (45/77, 58.4%) was not significantly (p ¼ 0.881) different from that of nonconfident veterinarians (67/112, 59.8%). A high majority of veterinarians thought that patient species (136/175, 77.7%), breed (104/170, 61.2%), individual sensitivity (174/182, 95.6%), experience of previous pain (121/168, 72%) and environment (114/173, 65.9%) had an effect on the pain experience of animals. The rank order of pain indicators considered the most useful in dogs and cats in the current study is presented in Table 2. The preferred pain indicators were attitude, interaction with the caregiver and response to handling or palpation of the surgical area in both species. These were followed by medical history in dogs, and inappetence in cats. No discrepancy was recorded with regard to relative value of pain indicators in dogs and cats, except for anorexia, which was considered more predictive in cats than in dogs (p <0.001). Pain control Ethical considerations were the first (184/189, 97.4%) and medical considerations the second (142/189, 75.1%) most important reasons cited to treat pain. The reasons most frequently given to explain restricted analgesic use were: difficulties in recognizing pain (70/120, 58.3%); lack of knowledge about appropriate therapy (50/120, 41.7%); fear of drug side effects (36/120, 30%); insufficient awareness of pain in animals (22/120, 18.3%); and fear of suppressing protective nociception (18/120, 15%). Fear of side effects (89/126, 70.6%) was cited as the principle reason for withholding NSAIDs, whereas restrictive narcotics legislation (143/179, 79.9%) and lack of knowledge (132/179, 73.7%) were the reasons for limiting opioid use. Over half of responding veterinarians (95/181) considered that their analgesic use was adequate in Table 2 Ranking of pain indicators considered most useful in dogs and cats Number citing (%) Pain indicator Dogs (188 responses) Cats (177 responses) Attitude e.g. depression, immobility, abnormal posture or function 166 (88.3) 146 (82.5) Interaction with caregiver and response to handling or palpation 125 (66.5) 111 (62.7) History 56 (29.8) 44 (24.9) Inappetance 55 (29.3) 82 (46.3) Comfort e.g. sleep, restlessness, aggressiveness 50 (26.6) 57 (32.2) Appearance e.g. facial expression, lack of grooming 44 (23.4) 44 (24.9) Physiological variables e.g. heart rate, blood pressure, respiratory rate 34 (18.1) 27 (15.3) Vocalization 23 (12.2) 16 (9) Ó Association of Veterinary Anaesthetists, 2004, 31, 154 163 157

canine patients. However, a significantly (p <0.001) lower percentage believed this to be true in cats (75/ 181, 41.4%). However, as indicated previously, veterinarians estimating their analgesic use to be sufficient in both species appeared to be no more likely (p > 0.05) to be analgesic users than nonusers. Analgesics used in dogs and cats Glucocorticoids and NSAIDs were the most commonly used analgesics in both species: 96.7% (178/ 184) and 100% (189/189) of users, respectively. These were followed by ketamine (39.6% (57/144) in dogs, 40.3% (54/134) in cats) and a 2 -agonists (36.7% (51/139) in dogs, 31% (39/126) in cats). The use of opioids was limited to 16.1% (26/161) of users in dogs and 8.1% (13/161) in cats, indicating significantly higher use (p <0.001) in dogs. Pure opioid agonists were more widely used than partial agonists in both species. Local anaesthetics were rarely mentioned. The percentage of opioid users in dogs was significantly (p ¼ 0.008) higher among veterinarians who had attended continuing education programmes devoted to pain within the last 5 years (34.6% (9/26) versus 12% (18/150) for participants and nonparticipants, respectively). Continuing education did not influence opioid drug prescription (p ¼ 0.248) in cats. The most popular NSAIDs used in dogs were tolfenamic acid (163/174, 93.7% of users), ketoprofen (149/176, 84.7%), meloxicam (141/167, 84.4%), aspirin (119/170, 70%) and nimesulide (79/154, 51.3%). Carprofen was least popular (77/ 153, 50.3%). In cats, tolfenamic acid (150/165, 90.9%) and ketoprofen (106/155, 68.4%) were the preferred NSAIDs, followed by nimesulide (64/143, 44.8%) and niflumic acid (55/142, 38.7%). When asked to classify NSAIDs as rather safe or unsafe, veterinarians ranked meloxicam (130/171, 76%), tolfenamic acid (97/174, 55.7%), carprofen (79/163, 48.5%) and ketoprofen (79/174, 45.4%) as rather safe, whereas aspirin (133/183, 72.7%) and phenylbutazone (114/164, 69.5%) were considered to be unsafe. For all NSAIDs evaluated (except noramidopyrine and flunixin), veterinarians who were frequent users of a particular NSAID considered it significantly safer than other NSAIDs. Three of the most popular NSAIDs (tolfenamic acid, ketoprofen and meloxicam) were considered among the safest available. Compared with the overall sample, veterinarians who had attended continuing education programmes dedicated to pain within the last 5 years used significantly (p ¼ 0.028) more carprofen 12/23 (52.2%) than those who had not 31/129 (24%), and significantly (p ¼ 0.028) less ketoprofen 13/25 (52%) versus 93/149 (62.4%). Concerns with analgesic use A large proportion (66/149, 44.3%) of veterinarians declared no knowledge of opioid side effects. For the remaining 55.7%, the greatest concerns were respiratory depression (61/152 (40.1%) in dogs, 46/146 (31.5%) in cats), excitement in cats (38/ 146, 26%) and sedation in dogs (28/152, 18.4%). Concern about the risk of excitement were significantly (p <0.001) higher in cats (38/146, 26%) than in dogs (9/152, 5.9%). For each side effect, the ranking of concerns did not significantly differ between opioid users and nonusers in both species. In spite of the high level of concern associated with NSAID administration, the veterinarians surveyed demonstrated a poor knowledge of contraindications to NSAID use (Table 3). Adequacy of knowledge was independent of age (p > 0.05). Dehydration and congestive heart failure were not considered to be absolute contra-indications by 136/162 (83.9%) and 157/172 (91.3%) of veterinarians, respectively. The importance of certain contra-indications such as the presence of renal failure or concomitant glucocorticoid use were frequently underestimated and considered relative, rather than absolute contraindications. Evaluation and treatment of pain in daily practice Evaluation of pain The most painful medical conditions among those listed in part 4 were bone tumours (8.1 ± 1.5 pain score), pancreatitis (7.9 ± 2) and trauma (7.5 ± 1.7). Among surgical operations, neurosurgery (7.5 ± 1.9) thoracotomy (7 ± 1.9) and orthopaedic surgery (6.6 ± 1.7) were ranked as the most painful (Table 4). Differences between the pain scores described by the respondents and the expert panel were statistically significant for all surgical conditions, except for castration and urogenital surgery. There was greater consensus for medical conditions. For surgical procedures, respondents systematically ranked pain lower than those in the reference panel (Table 4). This was particularly obvious for exploratory laparotomy (difference of 2.6 between both scores) and for oral, dental and ocular surgery (difference of 2.3). This pattern was not evident for the medical conditions. 158 Ó Association of Veterinary Anaesthetists, 2004, 31, 154 163

Table 3 Evaluation of adequate knowledge of NSAIDs contra-indications Percentage responses Issues (number of responses) Absolute CI a Relative CI a Not a CI a Does not know Anaesthesia (163) 0 (0%) 24 (14.7%) 131 (80.4%) 8 (4.9%) Renal failure (186) 56 (30.1%) 126 (67.7%) 4 (2.2%) 0 (0%) Hepatic insufficiency (179) 18 (10.1%) 132 (73.7%) 18 (10.1%) 11 (6.1%) Congestive heart failure (172) 15 (8.7%) b 70 (40.7%) 69 (40.1%) 18 (10.5%) Dehydration (162) 26 (16%) 47 (29%) 41 (25.3%) 48 (29.6%) Concurrent use of a diuretic (172) 18 (10.5%) 79 (45.9%) 56 (32.6%) 19 (11%) Concurrent use of a corticosteroid (176) 65 (36.9%) 85 (48.3%) 23 (13.1%) 3 (1.7%) Thrombocytopaenia, bleeding (175) 74 (42.3%) 82 (46.9%) 13 (7.4%) 6 (3.4%) Impending surgery (166) 6 (3.6%) 61 (36.7%) 95 (57.2%) 4 (2.4%) Gastrointestinal disorder (181) 68 (37.6%) 95 (52.5%) 14 (7.7%) 4 (2.2%) Shock or trauma (162) 10 (6.2%) 50 (30.9%) 95 (58.6%) 7 (4.3%) Question: For each of the proposed conditions, would you consider that the use of an NSAID absolutely, relatively or not contraindicated? Correct response is in bold. Most frequent response is in italics. a Contraindication. b The risk of side effects associated with NSAID administration in dogs with congestive heart failure is indirect, being secondary to impaired cardiac output and tissue/organ perfusion. Table 4 Pain ranking a by veterinarians and percentage of analgesic users for 10 selected surgical procedures and 7 medical conditions Pain score (±SD b ) Reference pain score c Percentage of analgesic users Surgical procedures Neurosurgery 7.5 (±1.9) 9.3 (±0.9) * 69.7 Thoracotomy 7 (±1.9) 8.7 (±1.2) * 65.8 Orthopaedic surgery 6.6 (±1.7) 8 (±0.8) * 83.7 Feline onychectomy 6.2 (±2.1) 7.7 (±1.2) * 58.3 Urogenital surgery 5 (±1.6) 5.7 (±0.5) 51.1 Oral, dental and ocular surgery 4.9 (±1.9) 7.2 (±0.9) * 58.3 Mammary surgery 4.9 (±1.8) 6.7 (±0.5) * 62.6 Spay 3.8 (±1.9) 5.3 (±0.5) * 36.3 Exploratory laparotomy 3.7 (±2.5) 6.3 (±0.5) * 33.5 Castration 2.8 (±1.7) 3.7 (±0.9) 17.2 Medical conditions Bone tumours 8.1 (±1.5) 8.3 (±0.9) 93.4 Pancreatitis 7.9 (±2) 7 (±0.8) 70.5 Trauma 7.5 (±1.7) 6.7 (±0.5) 97.3 Glaucoma 7 (±2) 7.7 (±0.5) 71.1 Osteoarthritis 6.4 (±1.6) 5.7 (±0.9) 97.8 Severe enteritis 6.2 (±2.1) 7 (±0.8) 49.5 Gingivitis, stomatitis 5 (±1.9) 5 (±0.8) 66.7 a Ranking performed on a scale where: 1, no pain at all; 10, the worst pain imaginable. b Standard deviation. c Mean score attributed by three reference veterinarians (M.H., J.-L.C., E.T.). * Difference between pain score and reference pain score statistically significant (p <0.05). Ó Association of Veterinary Anaesthetists, 2004, 31, 154 163 159

Veterinarians classified as analgesic users ranked pain significantly higher for 10 of the 17 painful conditions evaluated, compared with nonusers. Differences between pain scores assigned by men and women were statistically significant for exploratory laparotomy (p ¼ 0.021), feline onychectomy (p ¼ 0.031), spaying (p ¼ 0.036), urogenital surgery (p ¼ 0.048) and thoracic surgery (p ¼ 0.017) for which women ranked pain significantly higher than men. For each condition evaluated, pain scores assigned by veterinarians classified as self-confident in pain assessment did not significantly differ from those of veterinarians classified as not self-confident. Treatment of pain There was considerable variation amongst veterinarians giving analgesics, with the operation itself being the most influential factor. Most analgesics were given for orthopaedic surgery (144/172, 83.7%) and least for castration (31/180, 17.2%) (Table 4). Except neurosurgery, for which glucocorticoids were preferred, NSAIDs were the analgesics primarily chosen for all operations surveyed. Variability in analgesic provision was less pronounced for medical conditions (Table 4). Medical conditions receiving the most aggressive therapy were osteoarthritis (182/ 186, 97.8%), trauma (178/183, 97.3%) and bone neoplasm (169/181, 93.4%). In these cases, analgesia was most often provided with NSAIDs and/or corticosteroids. Conditions for which the highest pain scores were recorded were not necessarily those for which the use of analgesics was greatest. Using previous definitions, 112 of the 189 veterinarians surveyed (59.3%) were classified as analgesic providers. The single significant predictor of analgesic provider status identified by unconditional analysis was gender (p ¼ 0.035). The majority of women 71.4% (40/56) and fewer men 54.6% (71/130) who completed the survey used analgesics. French veterinary education and pain Most veterinarians surveyed felt that undergraduate veterinary school had completely (59/187, 31.6%) or partially (78/187, 41.7%) failed to enable them to assess animal pain. Younger veterinarians (<35 years) were more satisfied with their education than their elders; 38.6% (22/57) felt moderately or completely satisfied with training, compared with only 25.2% (27/107) aged between 35 and 50 years and 4.8% (1/21) of the 50 65 years olds. Similar tendencies were noticed with regard to pain management: most veterinarians believed that undergraduate veterinary school had totally (52/ 183, 28.4%) or partially (86/183, 47%) failed to teach pain management in animals. Veterinarians under 35 years of age were the least critical; 41.8% (23/55) felt moderately or completely satisfied with their training compared with 19.6% (21/107) of the 35 50 years olds and 5.3% (1/19) of those aged between 50 and 65 years. A majority of practitioners (165/182, 90.7%) were willing to attend continuing education courses devoted to small animal pain management. Amongst these, 79 (43.4%) claimed they would support a training course in this subject for their animal health technicians. Discussion The response rate was not influenced by gender. The response rate was highest amongst young veterinarians, but level of concern with pain management was independent of age, gender and the presence of an AHT. This may indicate that younger veterinarians are more comfortable responding to questionnaires on this subject. Pain assessment French veterinarians seem willing to improve their skills in evaluating and treating pain, as suggested by the high response rate (49.9%) and the moderate to high level of concern with pain indicated by 99.5% of respondents. These results indicate the presence of a receptive environment for developing and implementing pain management in small animal practice. Pain evaluation was the veterinarian s responsibility according to 92.4% of respondents, with only a minority sharing this task with their AHTs. This contrasts with findings in the UK (Capner et al. 1999) and Canada (Dohoo & Dohoo 1996b, 1998) where AHTs play a major role in the provision of perioperative analgesia i.e. 71% (UK) and 88% (Canada). French veterinarians were unaware of pain scales, but 87.5% of respondents felt confident in their pain recognition skills. Levels of confidence were lower in British (25%) (Capner et al. 1999) and Canadian (23%) (Dohoo & Dohoo 1996a) practitioners. Paradoxically, the main reason that 58.3% of respondents gave for withholding analgesics in the current study was difficulty in recognizing pain. Moreover, veterinarians determined to be self-confident in pain assessment were more likely to consider their use of 160 Ó Association of Veterinary Anaesthetists, 2004, 31, 154 163

analgesics to be sufficient, even though they used less analgesics. This apparent over-confidence in pain recognition was not evident when abilities to quantify pain were assessed, because the majority (58.8%) of respondents felt unable to do this. Anil et al. (2002) recently emphasized the point that the quantification of pain is a prerequisite for choosing appropriate analgesics and judging their effects. Effort has recently been made to propose and validate a multifactorial rating scale for pain in dogs (Conzemius et al. 1997; Holton et al. 1998; Firth & Haldane 1999) and cats (Cambridge et al. 2000). French veterinarians felt that signs of pain did not differ between dogs and cats. They ranked the animal s attitude, its interaction with the caregiver and its response to handling or palpation of the surgical wound as the most important signs of pain in dogs. In cats, the absence of normal behaviour rather than the presence of unusual behaviour was believed to indicate pain. French veterinarians considered that inappetence was more indicative of pain in cats compared with dogs. Practitioners also relied more on behavioural, rather than physiological, signs to indicate pain, which is similar to the results of an Australian survey (Watson et al. 1996) and is corroborated by data from experimental studies (Firth & Haldane 1999; Cambridge et al. 2000; Ferreira et al. 2001). A principle difference between French and other practitioners (Hansen & Hardie 1993; Watson et al. 1996) in evaluating pain is the importance attributed to vocalization, which is regarded as an important indicator of pain in various species, including dogs and cats by non- French practitioners (Anil et al. 2002). That French veterinarians disregard vocalization as a pain indicator may arise from a belief that vocalization is more strongly associated with stress or distress. Pain control The low use of analgesics appears to be related to inadequate education rather than ideological convictions: difficulties in recognizing pain (58.3%), lack of knowledge of appropriate use (41.7%) and fear of side effects (30%) were held to be significantly more important reasons for withholding analgesia, than insufficient awareness of pain (18.3%) or fear of protective reflex suppression (15%). This differs from the relative importance (30.4%) given by British practitioners (Capner et al. 1999) to fears of reflex depression. The belief that some pain is required to limit post-operative activity was not regarded as a good reason for withholding analgesics by respondents in the current study. Unlike local anaesthetics which suppress all sensation, analgesics diminish pain without affecting objective sensations such as touch and temperature. Indeed, human patients have commented that opioids relieve, but do not eliminate pain (Pasternak 2000). Furthermore, it is possible to treat clinical pain with minimal interference of nociceptive paths, thereby eliminating suffering associated with the former while maintaining the protective effects of the latter (Muir & Woolf 2001). This appears to have changed practitioner s opinion over analgesic use, and may encourage future developments in animal pain management. Preemptive multimodal- and mechanism-based (Woolf & Mitchell 2001) therapy should, in theory, help to prevent physiological, phenotypic and behavioural consequences of uncontrolled pain and the nervous system s memory of pain (Muir & Woolf 2001). Opioid drugs, particularly butorphanol, meperidine and buprenorphine, are popular in Great Britain (91% of users; Capner et al. 1999), Australia (82%; Watson et al. 1996) and Canada (Dohoo & Dohoo 1996a), even though their ability to control severe pain is questioned (Dohoo & Dohoo 1996a,b). In France, the percentage of practitioners giving opioids was less. However, French veterinarians who had attended continuing education courses within the previous 5 years were more likely to prescribe them. The French narcotics legislation was cited as the most common (79.9%) reason for withholding opioid drugs from animals in pain. However, the legislation has recently changed, facilitating prescription (Troncy 1999). Inadequate knowledge of opioid use, resulting from deficits in undergraduate education, was a common reason cited (73.7%) and was much higher than a fear of side effects (21.2%). French, English (Capner et al. 1999), Canadian (Dohoo & Dohoo 1996a) and Australian (Watson et al. 1996) practitioners appear to share the same concerns with the alleged side effects of opioids in dogs and cats. Maniacal excitement in cats was cited by 26% of respondents, reflecting the intellectual legacy of Joel & Arndts (1925) who used 20 mg kg )1 of morphine and who have been regularly cited by veterinary pharmacologists ever since. Other opioid side effects causing concern, e.g. respiratory depression, appear to have been transferred from human studies (Dohoo & Dohoo 1996a; Troncy & Langevin 2001). The efficacy and safety of opioid drugs in dogs and cats for surgical, traumatological, burn or neoplastic pain are well recognized (Nolan 2000; Pascoe 2000a,b). Ó Association of Veterinary Anaesthetists, 2004, 31, 154 163 161

Nonsteroidal anti-inflammatory drugs are the most popular analgesics in France, with newer generation drugs like tolfenamic acid, ketoprofen, meloxicam, carprofen and nimesulide being preferred to aspirin, flunixin and phenylbutazone. Tolfenamic acid, ketoprofen and carprofen were also popular in a British study (Capner et al. 1999), while ketoprofen was popular in Australia (Watson et al. 1996). Although phenylbutazone and aspirin were preferred for treating painful or inflammatory musculoskeletal disorders in dogs, flunixin was popular as a general analgesic. In dogs, preferential cyclo-oxygenase (COX-2)-inhibiting NSAIDs were preferred in this species, who considered meloxicam, tolfenamic acid and carprofen to be relatively safe. The recent (February, 1999) introduction of carprofen into France at the time of the current study probably explains its low level of its use and reputation for safety compared to Great Britain (Capner et al. 1999). Carprofen was more popular than ketoprofen amongst practitioners who had attended continuing education courses on pain in the previous 5 years. Tolfenamic acid and nimesulide are the only COX-2-inhibiting NSAIDs licensed for use in cats in France, and along with ketoprofen, are commonly used. The choice of NSAID used appears to depend on their perceived safety. However, many respondents were not aware of the problems associated with NSAID use (Table 3). More than 96% of French practitioners appear to use glucocorticoids for analgesia, which is similar to the rate found in Australia (Watson et al. 1996). Approximately 40% of practitioners also use xylazine, medetomidine or romifidine, and dissociative agents like ketamine and tiletamine, as analgesics, mainly as components of general anaesthesia. This is greater than the rate found in the UK (Capner et al. 1999) but similar to that found in Australia (Watson et al. 1996). Local anaesthetics (lidocaine, bupivacaine) are seldom used in France, in contrast to Great Britain (Capner et al. 1999) and Canada (Dohoo & Dohoo 1996a). Evaluation and treatment of pain in daily practice Comparing the scores given by the respondents and the reference panel to a list of operations indicated that the former underestimated post-surgical pain. This ranking was influenced by the status of analgesic user and by gender. Analgesic users and women tended to accord higher pain scores. These results are similar to those obtained in Canada (Dohoo & Dohoo 1996a) where AHTs have a strong influence on pain quantification (Dohoo & Dohoo 1996a, 1998). However, in France, AHTs appear to participate negligibly in pain evaluation. There was correlation between the ranked level of pain induced by various conditions and the percentage of veterinarians using analgesics to manage them except for osteoarthritis. For this condition, which was classified as moderately painful, a high percentage (97.8%) of respondents used NSAIDs and glucocorticoids. This excessive response occurs because osteoarthritis is a common condition for which clients may have a high expectations of analgesic prescription. Gender was the most accurate determinant of whether French veterinarians prescribed analgesics, with 71.4% of women versus 54.6% of men being prescribers. This is similar to the situation in Canada (Dohoo & Dohoo 1996b) and the UK (Capner et al. 1999). NSAIDs were the analgesics most commonly used for managing pain associated with all surgical procedures investigated, irrespective of the degree of pain believed to be associated with each procedure. A similar proportion of French and British veterinarians (Capner et al. 1999) used analgesics for ovariohysterectomy and castration; this was greater than the use reported in Canada (Dohoo & Dohoo 1996a). Veterinary education and pain French veterinarians considered that undergraduate veterinary schools had failed to provide them with the ability to evaluate and treat pain, although younger practitioners felt significantly more comfortable with these tasks. A majority of veterinarians indicated willingness to participate, along with their AHTs, in future continuing education on the subject of pain. Conclusion Although questionnaires are likely to induce a biased response and limit the interpretation of results (Capner et al. 1999), the results of the current investigation indicate that the provision of analgesia to animals in France is often irrational and sometimes hazardous. Even when analgesia is considered, it appears often to be inadequate, being based mainly on NSAIDs, widely regarded as being suitable only for the control of weak to moderate pain according to the WHO classification (Ventafridda et al. 1985). However, the French veterinary profession s apparent interest in pain and analgesia 162 Ó Association of Veterinary Anaesthetists, 2004, 31, 154 163

suggests that in future, effective analgesics, e.g. morphine, might become popular. The introduction of new techniques, e.g. epidural and transdermal analgesia, and drugs butorphanol, new NSAIDs, COX-2 specific and lipoxygenase inhibitors, etc. are likely to diversify and improve the provision of French veterinary analgesia. The French practitioner s attitude to the use of analgesics and the role of AHTs in pain management contrasts with that of English-speaking countries. However, the tools used for pain evaluation and the desire to improve small animal pain management, particularly by female practitioners, are common features. Acknowledgements The authors thank Sophie Cuvelliez for reviewing the manuscript and are indebted to PFIZER Laboratories, particularly Emmanuelle MATHIEU, for financial and statistical support. Eric Troncy is funded by the Fonds du Centenaire of the Faculté de Médecine Vétérinaire Université de Montréal. References Anil S, Anil L, Deen J (2002) Challenges of pain assessment in domestic animals. J Am Vet Med Assoc 220, 313 319. Bertolini G, Minelli C, Latronico N et al. (2002) The use of analgesic drugs in postoperative patients: the neglected problem of pain control in intensive care units. An observational, prospective, multicenter study in 128 Italian intensive care units. Eur J Clin Pharmacol 58, 73 77. Cambridge AJ, Tobias KM, Newberry RC et al. (2000) Subjective and objective measurements of postoperative pain in cats. J Am Vet Med Assoc 217, 685 690. Capner CA, Lascelles BD, Waterman-Pearson AE (1999) Current British veterinary attitudes to perioperative analgesia for dogs. Vet Rec 145, 95 99. Conzemius MG, Hill CM, Sammarco JL et al. (1997) Correlation between subjective and objective measures used to determine severity of postoperative pain in dogs. J Am Vet Med Assoc 210, 1619 1622. Dohoo SE, Dohoo IR (1996a) Postoperative use of analgesics in dogs and cats by Canadian veterinarians. Can Vet J 37, 546 551. Dohoo SE, Dohoo IR (1996b) Factors influencing the postoperative use of analgesics in dogs and cats by Canadian veterinarians. Can Vet J 37, 552 556. Dohoo SE, Dohoo IR (1998) Attitudes and concerns of Canadian animal health technologists toward postoperative pain management in dogs and cats. Can Vet J 39, 491 496. Ferreira X, Lambert L, Leblond A et al. (2001) Comparison of analgesia and perioperative problems associated with the preoperative use of ketoprofen and nimesulide in dogs. Vet Anaesth Analg 28, 207 (Abstract). Firth AM, Haldane SL (1999) Development of a scale to evaluate postoperative pain in dogs. J Am Vet Med Assoc 214, 651 659. Hansen B, Hardie E (1993) Prescription and use of analgesics in dogs and cats in a veterinary teaching hospital: 258 cases (1983 1989). J Am Vet Med Assoc 202, 1485 1494. Holton LL, Scott EM, Nolan AM et al. (1998) Comparison of three methods used for assessment of pain in dogs. J Am Vet Med Assoc 212, 61 66. Joel E, Arndts F (1925) Beitrange zur Pharmakologie der Koperstellung und der Labyrinthreflexe. XIX Mittelung: Morphin Arch Gesund Physio 210, 280 283. Klopfenstein CE, Herrmann FR, Mamie C et al. (2000) Pain intensity and pain relief after surgery. A comparison between patients reported assessments and nurses and physicians observations. Acta Anaesthesiol Scand 44, 58 62. Muir WW III, Woolf CJ (2001) Mechanisms of pain and their therapeutic implications. J Am Vet Med Assoc 219, 1346 1356. Nolan AM (2000) Pharmacology of analgesic drugs. In: Pain Management in Animals (1st edn). Flecknell PA, Waterman-Pearson A (eds). W.B. Saunders, London, UK, pp. 21 52. Pascoe PJ (2000a) Opioid analgesics. Vet Clin North Am: Small Anim Pract 30 (757), 772. Pascoe PJ (2000b) Problems of pain management. In: Pain Management in Animals (1st edn). Flecknell PA, Waterman-Pearson A (eds). W.B. Saunders, London, UK, pp. 161 177. Pasternak GW (2000) Opioids. In: Foundations of Anesthesia Basic and Clinical Sciences (1st edn). Hemmings H, Jr, Hopkins P (eds). Mosby Harcourt Publishers Ltd., London, UK, pp. 275 284. Troncy E (1999) Supply and use of third step analgesics in France (French Short communication). Le Point Vet 30 (203), 77 79. Troncy E, Langevin B (2001) Clinical Handbook: Analgesia for Small Animals (1st edn). Editions du Point Vétérinaire, Maisons-Alfort, France, pp. 27 83 (French). Ventafridda V, Saita L, Ripamonti C et al. (1985) WHO guidelines for the use of analgesics in cancer pain. Int J Tissue React 7, 93 96. Watson ADJ, Nicholson A, Church DB et al. (1996) Use of anti-inflammatory and analgesic drugs in dogs and cats. Aust Vet J 74, 203 210. Woolf CJ, Mitchell BM (2001) Mechanism-based pain diagnosis. Anesthesiology 95, 241 249. Received 18 March 2002; accepted 6 March 2003. Ó Association of Veterinary Anaesthetists, 2004, 31, 154 163 163