La maladie valvulaire dégénérative mitrale (MVDM) est la cardiopathie la plus fréquente chez le chien. Elle se caractérise par la présence d’un flux de régurgitation mitrale, une diminution potentielle du débit cardiaque et un risque d’insuffisance cardiaque congestive. L’anesthésie d’un chien atteint de MVDM nécessite une compréhension de la physiopathologie et des répercussions hémodynamiques en per- et post-anesthésie. Une évaluation préanesthésique rigoureuse (examen clinique, échocardiographie, radiographie thoracique) permet de caractériser la gravité de l’affection, de dépister d’éventuelles comorbidités, et d’orienter la stabilisation préanesthésique. La planification anesthésique repose sur une sélection raisonnée des molécules dépendant de la gravité de l’affection, une analgésie multimodale, une fluidothérapie raisonnée et un monitorage rapproché. La prévention de la bradycardie, de la vasoconstriction et de la dépression myocardique contribue à limiter les flux de régurgitation et à préserver le débit cardiaque. Une approche individualisée et la collaboration entre cardiologie et anesthésie améliorent significativement la sécurité anesthésique des chiens atteints de MVDM.

L’anesthésie locorégionale est essentielle pour examiner et opérer l’œil du cheval en toute sécurité, notamment lors de procédures debout. Elle diminue la douleur, les réflexes palpébraux et le recours à l’anesthésie générale. Cet article présente les principales techniques : anesthésie topique pour l’analgésie de la cornée, bloc auriculopalpébral pour l’akinésie des paupières, blocs sensitifs des quatre branches du nerf trijumeau pour l’analgésie palpébrale, bloc rétrobulbaire pour l’anesthésie et l’akinésie du globe et la réduction du risque de réflexe oculocardiaque, ainsi que les infiltrations locales pour une couverture plus large. Pour chaque technique, les indications, doses, précautions et risques sont décrits, avec des repères pratiques pour une utilisation sécurisée.

L’anesthésie générale du cheval est associée à une morbidité et une mortalité supérieures à celles observées dans les autres espèces domestiques. Les myopathies et neuropathies post-anesthésiques constituent des complications spécifiques dont l’incidence est estimée à 0,4-3,7 %, et qui restent impliquées dans une proportion des décès péri-anesthésiques. Les myopathies résultent principalement d’une hypoperfusion musculaire secondaire à une diminution de la pression artérielle moyenne et/ou à une augmentation de la pression intracompartimentale, dans un contexte de décubitus prolongé. Les neuropathies sont liées à des phénomènes de compression et/ou d’étirement nerveux. L’identification des facteurs de risque (gabarit, durée d’anesthésie, positionnement, hypotension, hypoxémie, prédispositions génétiques), le maintien d’une pression artérielle moyenne > 70 mmHg, la réduction du temps anesthésique et un positionnement rigoureux avec matelassage adapté constituent les piliers de la prévention. Le diagnostic différentiel repose principalement sur la présence de douleur et l’élévation des enzymes musculaires en cas de myopathie. La prise en charge reste essentiellement symptomatique et de soutien. Le pronostic des myopathies localisées et des neuropathies périphériques dépend de la gravité initiale et de l’évolution précoce. La myélomalacie et la myopathie généralisée sont de pronostic très sombre.

Les anti-inflammatoires non stéroïdiens (AINS) jouent un rôle majeur dans le cadre d’une analgésie périopératoire, notamment grâce à leur action inhibitrice sur les cyclo-oxygénases (COX) impliquées dans la réponse inflammatoire et la douleur associée. Cet article détaille les mécanismes d’action des AINS, leur sélectivité COX-1/COX-2, ainsi que les contre-indications et effets secondaires fréquents, en particulier au niveau rénal et digestif. Piliers de l’analgésie multimodale, leur utilisation est recommandée pour optimiser l’efficacité analgésique. L’administration anticipée d’AINS permet d’améliorer considérablement la douleur périopératoire, mais augmente le risque d’hypoperfusion rénale en cas d’hypotension peranesthésique, justifiant une surveillance hémodynamique rigoureuse pendant l’anesthésie générale. L’administration précoce postopératoire lors du réveil, à l’extubation de l’animal, semble un compromis sûr et efficace pour limiter les risques, mais optimiser la prise en charge de la douleur inflammatoire postopératoire. Les indications, précautions et recommandations pratiques pour une utilisation sécurisée sont exposées. Les AINS disponibles en France, les alternatives thérapeutiques et le cas particulier du grapiprant sont aussi abordés. Cette synthèse vise à optimiser l’efficacité analgésique tout en limitant les complications en contexte périopératoire.

Hemorrhage is a common and significant complication of trauma or surgical procedures. Rapid and accurate assessment of a patient's condition during hemorrhage is crucial for management. The primary objective was to describe pupillometry parameters and parasympathetic tone activity (PTA) during hemorrhage. The secondary objective was to develop a model to predict blood losses. In this preclinical study, 15 pigs were used. After anesthesia, 35 mL/kg of blood was removed over 1 h. Hypovolemia was maintained for 90 min. Indirect assessment of autonomic nervous system balance was based on the monitoring of heart rate (HR), blood pressure (BP), shock index (SI), PTA and pupil diameter (PD). PD and the pupillary light reflex (PLR) were measured using a portable pupillometer. HR and SI significantly increased during hemorrhage, whereas BP decreased. PD increased significantly during hemorrhage (+0.03 mm/min, [0.01; 0.04]; then +0.01 mm/min, [0.01; 0.02]) and the hypovolemia period (+0.004 mm/min, [0.002; 0.005]). PLRs were absent in 251/254 measurements. Large variations in PTA parameters were observed over time, which prevented statistical analysis. The best models predicted blood losses with a mean absolute error ranging from 3.61 to 3.67 mL/kg. In conclusion, hemorrhage was associated with an increased PD potentially influenced by anesthesia. The selected models allowed reasonable prediction of blood losses with a slight tendency to overestimate. The study provides proof of concept for the use of this kind of models to predict blood losses. The results must be confirmed in other contexts: with different species, other anesthetic protocols, or different surgical procedures.

A 2‐year‐old cat was anaesthetised for pelvic fracture repair. Premedication included methadone 0.2 mg/kg and medetomidine 5 µg/kg intravenously (IV). Anaesthesia was induced with ketamine 2 mg/kg and propofol 1 mg/kg IV and maintained with isoflurane. A sacrococcygeal epidural with ropivacaine 0.75% was performed with nerve stimulation: injection of 0.21 mL/kg initially, followed by 0.11 mL/kg after needle repositioning due to clear fluid extravasation pericatheter (total: 2.4 mg/kg). After 10 minutes, increased anaesthetic depth was noted with eye centralisation, no corneal reflex and bradycardia. Isoflurane was stopped. Atipamezole, naloxone and lipid emulsion 2 mL/kg over 1 minute twice followed by 0.25 mL/kg/min for 1 hour did not improve neurological status. An intracranial hypertension was suspected, treated with mannitol (total 2 g/kg). Metabolic acidosis and hyperkalaemia arose. After 7 hours of coma, neurological status improved. Full recovery was achieved within 12 hours. Re‐anaesthesia a week later without regional analgesia proceeded without complication.

L’espèce équine comporte un risque de mortalité particulièrement élevé lors d’anesthésie générale, pour des raisons propres à l’espèce (anatomie, poids, tempérament), accentuées par les effets des molécules anesthésiques. L’étude internationale CEPEF4, publiée en 2025, portant sur plus de 47 000 anesthésies équines, estime la mortalité globale à 1,2 %, avec 4,2 % pour les coliques et 0,6 % pour les autres interventions. La majorité des décès survient au réveil ou dans les jours suivant l’anesthésie. Les principales causes rapportées sont les fractures, suivies des complications abdominales, puis des complications neurologiques. Plusieurs facteurs de risques intrinsèques et extrinsèques au cheval sont confirmés : âge avancé, gestation, état de santé (score ASA élevé), maigreur, chirurgie de colique, durée d’anesthésie et caractère urgent de l’intervention notamment. La prévention repose sur une évaluation préanesthésique rigoureuse (ASA et grille CHARIOT), une stabilisation préalable, un protocole anesthésique adapté, une réduction du temps d’anesthésie, un monitorage rapproché et des stratégies de réveil individualisées. La formation continue des équipes y contribue également de façon déterminante.

Assessing pupil diameter and pupillary light reflexes (PLR) in a clinical setting can be challenging due to the pupil's sensitivity to light and the rapid nature of these reflexes. Pupillometry is a technique that is helpful in human medicine in anesthesiology and critical care medicine. However, evidence in veterinary medicine is scarce on the application of pupillometry and analgesic response. Therefore, the aim of this pilot observational study was to assess the feasibility and reliability of portable pupillometry in cats, identify individual factors influencing pupillometric parameters, and compare measurements before and after analgesic administration. Twenty-eight cats, including healthy and hospitalized sick cats, underwent pupillometry using a Neurolight® device. Pupil diameter, PLR parameters, and 4 A-VET pain score were measured before and after opioid administration. Test-retest reliability was evaluated, and the influence of factors was analyzed using multivariate models. Pupillometry was feasible in 97 % of cats, with measurements being easier in hospitalized sick cats. Good reliability was observed for pupil diameter measurements, while PLR parameters showed poor reliability. Weight was significantly associated with pupil size, potentially due to the relationship between body weight and eyeball size. No association was found between pain scores and pupil diameter, and no significant changes in pupil size were observed after opioid administration. In conclusion, portable pupillometry using the Neurolight® is feasible in cats and provides reliable pupil diameter measurements. However, the device is not suitable for objective PLR assessment in this species. Further studies are needed to better understand the factors influencing pupillometry measurements in cats.

L’anesthésie d’un animal insuffisant rénal chronique est relativement courante pour tout praticien. Si certains cas sont avérés, d’autres ne sont pas connus. Il est important de détecter les animaux à risque en recherchant la maladie au sein de certaines populations avant l’anesthésie (facteurs de risque en fonction de l’âge, de la race, des anomalies anamnestiques ou cliniques). Quand une maladie rénale chronique a été objectivée, il faut la classer (évaluation de la gravité) et rechercher les comorbidités associées, notamment : déshydratation, anémie, troubles acido-basiques et électrolytiques, hypertension artérielle systémique, etc. Il faudra tenter alors de corriger ces anomalies avant une anesthésie, la phase préanesthésique est ainsi primordiale pour limiter le risque anesthésique et une dégradation de la fonction rénale. La technique anesthésique vise à maintenir au mieux la perfusion rénale et le débit de filtration glomérulaire. Dans cette optique, le choix des molécules d’anesthésie se fait en privilégiant notamment les opioïdes et les benzodiazépines, en cherchant à diminuer les besoins en halogénés via des perfusions continues et des techniques d’anesthésie loco-régionale. La préservation de la fonction rénale passe également par une surveillance en continu (pour éviter des complications, telles que l’hypotension, l’hypoventilation, l’hypoxémie), une fluidothérapie per- et périopératoire, une oxygénation, et une prise en charge en limitant la douleur et le stress.

An 18‐year‐old mare, admitted for colic, underwent an exploratory laparotomy for right dorsal colon displacement. During recovery from anaesthesia, a severe acute onset laryngeal obstruction occurred at standing, requiring orotracheal re‐intubation, which relieved the obstruction. This was followed by secondary pulmonary oedema, treated with furosemide (2 mg/kg intravenously) and oxygen. At re‐extubation, attempted 30 minutes later, the horse re‐obstructed immediately and collapsed, prompting an emergency tracheostomy. Furosemide (2 mg/kg intravenously) and dexamethasone (0.1 mg/kg intravenously) were also administered. The horse stood up thereafter with a patent airway. In the history, to note, nasogastric intubation was particularly difficult at admission, and was identified as a potential cause of laryngeal trauma that may have precipitated laryngeal obstruction due to severe local inflammation. Dynamic laryngeal collapse probably also played a role due to high inspiratory negative pressure generated at the precise timing of standing, similar to exercise.

Coenurosis is a parasitic disease affecting the central nervous system of small ruminants. This case describes the anaesthetic management of a ewe undergoing craniotomy for associated-brain cyst removal. As small ruminants are food-producing animals, there is limited access to veterinary medicinal products, and choosing an appropriate anaesthetic protocol is challenging. After premedication with detomidine (0.03 mg/kg intramuscularly) and butorphanol (0.1 mg/kg intravenously), anaesthesia was induced with lidocaine (1 mg/kg) and ketamine (2 mg/kg intravenously), and maintained with isoflurane in oxygen, using controlled mechanical ventilation. The animal was placed in sternal recumbency, with the head slightly elevated. A lidocaine incisional infiltration was performed, and dexamethasone (0.06 mg/kg intravenously) was administered. Vital parameters remained stable throughout anaesthesia, but moderate agitation was noticed at recovery, treated with butorphanol (0.05 mg/kg intravenously). While the ewe showed clinical improvement initially, its neurological status deteriorated in the following days, leading to euthanasia. Postmortem examination revealed intracranial haematoma.

ObjectivesTo evaluate the safety and feasibility of high flow oxygen therapy (HFOT), and to record SpO2 and desaturation episodes in dogs and cats receiving HFOT or conventional oxygen therapy (COT) during bronchoscopy ± bronchoalveolar lavage (BAL).Materials and methodsDogs and cats undergoing bronchoscopy ± BAL between January and May 2023 were included in the study. Patients were randomly allocated to two groups: HFOT (HFOT group; two cats and four dogs) and COT (COT group; one cat and five dogs). HFOT and COT were started at the beginning of the bronchoscopy. HFOT was delivered with a gas flow rate of 1 L/kg/min at an FiO2 of 100% and a temperature of 34°C (pediatric mode) or 37°C (adult mode). COT was delivered through the working channel of the bronchoscope at a rate of 1.5 L/min. The safety and feasibility of HFOT were assessed, and peripheral oxygen saturation (SpO2) was measured by pulse oximetry every 30 s throughout the procedure.Measurements and main resultsHFOT was feasible and safe in both dogs and cats with no complications reported. While there was no significant difference in the number of desaturation episodes (SpO2 < 94%) between the two groups, none of the patients in the HFOT group experienced severe desaturation (SpO2 < 90%). In contrast, two patients in the COT group had an SpO2 < 90%. Mean SpO2 was significantly higher in the HFOT group compared to the COT group at T0 (98% ± 2% vs. 94 ± 2%), T0.5 (98% ± 2% vs. 94% ± 3%) and T1 (98% ± 2% vs. 94% ± 4%).ConclusionTo the authors’ knowledge, this is the largest study conducted to date using HFOT during bronchoscopy in dogs and cats. Our results suggest that HFOT is feasible and safe during bronchoscopy ± BAL. Furthermore, HFOT may reduce the risk of desaturation episodes in dogs and cats undergoing bronchoscopy and BAL.

Sepsis is associated with hypoperfusion and organ failure. The aims of the study were: 1) to assess the effect of pimobendan on macrocirculation and perfusion and 2) to describe a multimodal approach to the assessment of perfusion in sepsis and compare the evolution of the perfusion parameters. Eighteen anaesthetized female piglets were equipped for macrocirculation monitoring. Sepsis was induced by an infusion of Pseudomonas aeruginosa. After the occurrence of hypotension, animals were resuscitated. Nine pigs received pimobendan at the start of resuscitation maneuvers, the others received saline. Tissue perfusion was assessed using temperature gradients measured with infrared thermography (TG = core temperature - tarsus temperature), urethral perfusion index (uPI) derived from photoplethysmography and sublingual microcirculation (Sidestream dark field imaging device): De Backer score (DBs), proportion of perfused vessels (PPV), microvascular flow index (MFI) and heterogeneity index (HI). Arterial lactate and ScvO2 were also measured. Pimobendan did not improve tissue perfusion nor macrocirculation. It did not allow a reduction in the amount of noradrenaline and fluids administered. Sepsis was associated with tissue perfusion disorders: there were a significant decrease in uPI, PPV and ScvO2 and a significant rise in TG. TG could significantly predict an increase in lactate. Resuscitation was associated with a significant increase in uPI, DBs, MFI, lactate and ScvO2. There were fair correlations between the different perfusion parameters. In this model, pimobendan did not show any benefit. The multimodal approach allowed the detection of tissue perfusion alteration but only temperature gradients predicted the increase in lactatemia.

Background:The anesthetic management of adrenalectomies for phaeochromocytoma excision, a catecholamine-secreting tumor, is challenging due to the potential for fatal complications following severe hemodynamic variations, including hypertensive crisis following tumor manipulation or sympathetic stimulation, but also severe hypotension and volume depletion post resection. Case Description:An 11 kg, 15-year-old male neutered Jack Russel Terrier, with mitral valve disease stage B2, was referred for adrenalectomy for phaeochromocytoma resection. The patient was administered per os prazosin 0.11 mg/kg twice a day and amlodipine 0.125 mg/kg once a day for preoperative stabilization. On the day of surgery, the dog received maropitant 1 mg/kg intravenously (IV) and was premedicated with 0.2 mg/kg methadone IV. Anesthesia was induced with alfaxalone 1 mg/kg IV and midazolam 0.2 mg/kg IV and maintained with partial intravenous anesthesia using sevoflurane in 70% oxygen and constant rate infusions of dexmedetomidine 0.5 μg/kg/hour and maropitant 100 μg/kg/hour. After induction of anesthesia, the dog was mechanically ventilated, and a transversus abdominal plane block was performed with ropivacaine 0.2%. The dog remained remarkably stable with a single, self-limiting, hypertension episode recorded intraoperatively. Postoperative rescue analgesia consisted of methadone and ketamine. The dog was discharged 48 hours after surgery, but persistent hypertension was reported at suture removal. Conclusion:The use of a low-dose dexmedetomidine CRI, a maropitant CRI, and a transversus abdominal plane block provided stable perioperative hemodynamic conditions for phaeochromocytoma excision in a dog.

The aim of this prospective clinical study was to evaluate the efficacy of the Surgical Pleth Index (SPI), a validated nociception monitor in human anaesthesia, in dogs. The technology uses a plethysmographic signal from a specific pulse oximetry probe to analyse pulse wave amplitudes and heartbeat intervals. Twenty-six healthy dogs anaesthetised for castration were included. SPI, invasive mean arterial pressure (MAP) and heart rate (HR) were continuously monitored. The occurrence or resolution of a haemodynamic reaction (HDR), defined as a > 20% increase in HR and/or MAP, was assessed at predefined times: cutaneous incision, testicles ' exteriorization, cutaneous suture, and fentanyl administration. Following nociceptive events, the dogs presenting a HDR showed a significant 8% and 10% increase in SPI at 3 and 5 min respectively, whereas after fentanyl administration, a 13% and 16% significant decrease in SPI were noted. Receiver operating characteristic curves analysis indicated a moderate performance for the dynamic variations of SPI over 1 min to predict a HDR (AUC: 0.68, threshold value: +15%) or its resolution after fentanyl administration (AUC of 0.72, threshold value:-15%) within 3 min. The SPI varied according to perioperative nociceptive events and analgesic treatment; however, its performance to anticipate a HDR was limited with high specificity but low sensivity. Refinement of the algorithm to specifically accommodate for the canine species may be warranted. Further studies are required to evaluate the influence of other factors on the performance of this index.

ObjectivesThis study was designed to prospectively evaluate the feasibility of an opioid-free anesthesia protocol and describe the quality of recovery and management of postoperative analgesia in dogs after a tibial plateau leveling osteotomy (TPLO).MethodsIn total, 20 dogs presented for TPLO were included. After premedication with intravenous (IV) medetomidine (0.005–0.007 mg/kg) and midazolam (0.2 mg/kg), the dogs were anesthetized using ketamine (2 mg/kg) and propofol and maintained with isoflurane and ketamine CRI (0.6 mg/kg/h). Sciatic and femoral nerve blocks were performed with bupivacaine 0.5% (0.087 +/− 0.01 and 0.09 +/− 0.02 mL/kg, respectively). Meloxicam (0.2 mg/kg IV) was administered intraoperatively, after osteotomy. Fentanyl (0.002 mg/kg IV) was administered intraoperatively, as rescue analgesia in the case of sustained increase in cardiorespiratory variables. Two pain scores (French 4A-VET and Glasgow short form) were performed at conscious sternal recumbency and 2, 4, 6, 8, 12, and 20 h after extubation and compared to baseline using a Friedman test followed by a Nemenyi post-hoc test. The time taken for the first food intake and urination was reported.ResultsIntraoperative opioid-free anesthesia was feasible in 11 dogs, whereas 9 dogs received fentanyl once during arthrotomy. No opioid postoperative rescue analgesia was required. Food intake occurred within 6 h, and all dogs were discharged after 24 h without any complication.ConclusionTotal opioid-free postoperative analgesia was achieved in all dogs, with adequate recoveries. Although opioid-free anesthesia was feasible in 55% of the population, a single dose of fentanyl was necessary in 45% of the dogs during arthrotomy.

Systemic inflammation and hemodynamic or microvascular alterations are a hallmark of sepsis and play a role in organs hypoperfusion and dysfunction. Pimobendan, an inodilator agent, could be an interesting option for inotropic support and microcirculation preservation during shock. The objectives of this study were to evaluate effect of pimobendan on cytokine and nitric oxide (NO) release and investigate whether changes of macro and microcirculation parameters are associated with the release of cytokines and NO in pigs sepsis model. After circulatory failure, induced by intravenous inoculation of live Pseudomonas aeruginosa, eight animals were treated with pimobendan and eight with placebo. Pimobendan did not affect cytokines secretion (TNF-alpha, IL-6 and IL-10), but decreased time-dependently NO release. Data of macro and microcirculation parameters, NO and TNF- alpha recorded at the time of circulatory failure (T hypotension ) and the time maximum of production cytokines was used for analyses. A positive correlation was observed between TNF-alpha and cardiac index (r r = 0.55, p = 0.03) and a negative with systemic vascular resistance (r r =-0.52, p = 0.04). Positive correlations were seen both between IL-10, 30 min after resuscitation (T 30min ), and systolic arterial pressure (r r = 0.57, p = 0.03) and cardiac index (r r = 0.67, p = 0.01), and also between IL-6, taken 2 h after resuscitation and systolic arterial pressure (r r = 0.53, p = 0.04). Negative correlations were found between IL-10 and lactate, measured resuscitation time (r r =-0.58, p = 0.03). Regarding microcirculation parameters, we observed a positive correlation between IL-6 and IL-10 with the microvascular flow index (r r = 0.52, p = 0.05; r = 0.84, p = 0.0003) and a negative correlation with the heterogeneity index with TNF-alpha and IL-10 (r r =-0.51, p = 0.05; r =-0.74, p = 0.003) respectively. NO derivatives showed a positive correlation with temperature gradient (r r = 0.54, p = 0.04). Pimobendan did not show anti-inflammatory effects in cytokines release. Our results also, suggest changes of macro- and microcirculation are associated mainly with low levels of IL-10 in sepsis.

Lors d’une suspicion de fracture phalangienne chez le cheval, la qualité de la sédation-analgésie est essentielle pour assurer la sécurité des intervenants, le confort de l’animal, et permettre l’évaluation clinique ainsi que la stabilisation du membre avant tout transport. Cet article propose une approche raisonnée du choix des molécules sédatives et analgésiques disponibles chez le cheval, en tenant compte de leur potentiel thérapeutique, de leurs effets indésirables et des considérations légales. Le transport ne doit être envisagé qu’après contrôle de la douleur et une immobilisation adéquate du membre atteint. Les alpha-2-agonistes sont les sédatifs de choix, à associer à des opioïdes pour l’analgésie et une synergie des effets sédatifs. L’administration d’anti-inflammatoires non-stéroïdiens (AINS) est recommandée si le cheval est hémodynamiquement stable. Des adjuvants comme la kétamine ou l’acépromazine peuvent compléter le protocole selon les besoins. À domicile, l’analgésie à long terme doit être individualisée. Elle repose principalement sur les AINS, dont le choix et le suivi visent à éviter les effets indésirables et à optimiser l’analgésie. Des recommandations pratiques de surveillance lors d’administration prolongée d’AINS sont proposées. Certains adjuvants peuvent également être ajoutés à domicile. Cette démarche vise à améliorer la prise en charge globale des chevaux traités de manière conservatrice pour une fracture de phalange.

Un chien Samoyède de 5 ans est référé pour abattement et hyperthermie. Le bilan sanguin préalablement réalisé par le vétérinaire traitant est sans anomalies. À l’admission, un pyothorax est diagnostiqué et des lésions au sein du lobe pulmonaire caudal gauche sont visibles à la tomodensitométrie, avec possible corps étranger migrant dans le pilier diaphragmatique gauche. Une lobectomie partielle par sternotomie et le retrait d’un élément végétal sont réalisés sous anesthésie générale. L’animal est placé sous ventilation contrôlée avec mise en place d’une analgésie peropératoire loco-régionale et systémique. Cinq jours après l’intervention, le chien montre une douleur faible, il est sorti d’hospitalisation, avec une association amoxicilline-acide clavulanique et paracétamol comme traitement. L’analyse histologique pulmonaire montre une pleuropneumonie fibrino-suppurée subaiguë sévère.

A prospective study was performed on hospitalized conscious dogs. The objectives were: 1) to evaluate the feasibility and reliability of portable infrared pupillometry using a measure of photo light reflexes, 2) to identify parameters influencing measures, and 3) to compare parameters before and after the administration of analgesia. Twenty-nine dogs were included. Pupillometry was feasible by a single person. There was an excellent reliability for the evaluation of pupil diameter. There was poor to good reliability for the other parameters. There was an association between weight and mean pupil diameter (estimate = 0.1 mm, CI95 = [0.0; 0.1], P = 0.02), mean diameter variation (estimate = 0.2%, CI95 = [0.0; 0.4], P = 0.01), and mean velocity (estimate = 0.08 ms, CI95 = [0.03; 0.13], P = 0.002). Male dogs had a significantly larger mean pupil diameter than females (estimate = 1.3 mm, CI95 = [0.1; 2.5], P = 0.03). Independently of weight and sex, there was an association between mean pupil diameter and pain score: the higher the pain score, the lower the mean pupil diameter. There was a correlation between mean pupil diameter and morphine dose (rho = -0.78, P = 0.0004). There was no significant difference in pupil diameter, velocity, and latency before and after analgesic administration. In conclusion, results do not support the use of PLR measurement by pupillometry as a pain assessment technique in hospitalized conscious dogs.