Comparison of intraperitoneal ropivacaine and ropivacaine–dexmedetomidine for postoperative analgesia in cats undergoing ovariohysterectomy

ObjectiveTo investigate the intraperitoneal (IP) administration of ropivacaine or ropivacaine–dexmedetomidine for postoperative analgesia in cats undergoing ovariohysterectomy.Study designProspective, randomized, blinded, positively controlled clinical study.AnimalsA total of 45 client-owned cats were enrolled.MethodsThe cats were administered intramuscular (IM) meperidine (6 mg kg⁻¹) and acepromazine (0.05 mg kg⁻¹). Anesthesia was induced with propofol and maintained with isoflurane. Meloxicam (0.2 mg kg⁻¹) was administered subcutaneously in all cats after intubation. After the abdominal incision, the cats were administered one of three treatments (15 cats in each treatment): IP instillation of 0.9% saline solution (group Control), 0.25% ropivacaine (1 mg kg⁻¹, group ROP) or ropivacaine and dexmedetomidine (4 μg kg⁻¹, group ROP–DEX). During anesthesia, heart rate (HR), electrocardiography, noninvasive systolic arterial pressure (SAP) and respiratory variables were monitored. Sedation and pain were assessed preoperatively and at various time points up to 24 hours after extubation using sedation scoring, an interactive visual analog scale, the UNESP-Botucatu multidimensional composite pain scale (MCPS) and mechanical nociceptive thresholds (MNT; von Frey anesthesiometer). Rescue analgesia (morphine, 0.1 mg kg⁻¹) IM was administered if the MCPS ≥6. Data were analyzed using the chi-square test, Tukey test, Kruskal–Wallis test and Friedman test (p < 0.05).ResultsHR was significantly lower in ROP–DEX compared with Control (p = 0.002). The pain scores, MNT, sedation scores and the postoperative rescue analgesia did not differ statistically among groups.Conclusions and clinical relevanceAs part of a multimodal pain therapy, IP ropivacaine–dexmedetomidine was associated with decreased HR intraoperatively; however, SAP remained within normal limits. Using the stated anesthetic protocol, neither IP ropivacaine nor ropivacaine–dexmedetomidine significantly improved analgesia compared with IP saline in cats undergoing ovariohysterectomy.     

Preperitoneal ropivacaine infusion versus epidural ropivacaine–morphine for postoperative analgesia in dogs undergoing ovariohysterectomy: a randomized clinical trial

ObjectiveTo assess the effect of continuous wound infusion (CWI) with preperitoneal ropivacaine on postoperative analgesia and compare it with the epidural administration of ropivacaine and morphine in bitches undergoing ovariohysterectomy.Study designA parallel, randomized, clinical, prospective and nonblinded study.AnimalsA group of 38 Greyhound bitches.MethodsIn the catheter group (CathG), CWI with ropivacaine 1% (1 mg kg^–1 + 0.8 mg kg^–1 hour^–1) was applied to the preperitoneal space over the surgical incision. In the epidural group (EpiG), ropivacaine 0.5% (1.3 mg kg^–1) and morphine (0.1 mg kg^–1) were epidurally administered. Occipital-coccygeal length was used to calculate the volume for the epidural. Pain was scored using a dynamic interactive visual analogue scale (DIVAS) and Glasgow composite measure pain scale–short form (CMPS-SF) before anaesthesia and at 2, 4, 6, 18, 21 and 24 hours after extubation. Incisional sensitivity using a dynamometer (MWTs-incision) was evaluated simultaneously. Plasma ropivacaine and cortisol concentrations, degree of sedation, motor blockade and response to interdigital clamping were measured or assessed. A two-way mixed analysis of variance and a Mann–Whitney U test were used to analyse data; p < 0.05.ResultsNo differences were detected in the DIVAS (p = 0.301), CMPS-SF (p = 0.600) scores, MWTs-incision measurements (p = 0.257) and cortisol values (p = 0.878) between the groups. Rescue analgesia was required in two dogs, one in each group, at 2 hours. Sedation, motor blockade and negative response to interdigital clamping were detected in EpiG at 2, 4 and 6 hours. Mean plasma ropivacaine values were higher in CathG (0.475 ± 0.164 ng mL^–1) than in EpiG (0.184 ± 0.213 ng mL^–1; p = 0.001).Conclusion and clinical relevanceCompared with epidural ropivacaine and morphine, CWI with preperitoneal ropivacaine is an effective analgesic technique for postoperative pain management in bitches undergoing ovariohysterectomy without motor blockade.     

Comparison of sedative and some cardiopulmonary effects of intramuscular medetomidine or medetomidine–tramadol in dogs

ObjectiveTo evaluate the clinical and physiologic effects of intramuscular (IM) administration of medetomidine with and without tramadol in dogs.Study designProspective experimental study.AnimalsA group of eight mixed breed dogs of both sexes, aged 1–2 years, weighing 16.0 ± 0.6 kg.MethodsEach dog was studied twice at ≥1 week interval. Medetomidine (5 μg kg^–1; treatment M) was administered IM alone or with tramadol (4 mg kg^–1; treatment MT). Sedation was scored by a system that included vocalization, posture, appearance, interactive behaviors, resistance to restraint and response to noise. Times from drug administration to ataxia, impaired walking, head drop, sternal and lateral position and standing were recorded. Sedation score, heart rate, respiratory rate, rectal temperature, end-tidal carbon dioxide (Pe′CO₂), hemoglobin oxygen saturation and mean noninvasive blood pressure were recorded and compared 15 minutes before and 15, 30 and 45 minutes after drug administration.ResultsDogs administered MT had higher sedation scores than dogs administered M at 30 and 45 minutes after drug administration (p < 0.05). Times to ataxia, impaired walking, head drop and sternal recumbency were not different between the treatments. Time to lateral recumbency was longer in M than in MT (21.1 ± 1.0 versus 17.6 ± 0.7 minutes, respectively; p < 0.05). Time to standing was longer in MT than in M (67.9 ± 1.4 versus 54.5 ± 1.9 minutes, respectively; p < 0.001). Measured physiological variables did not differ between the treatments, with the exception of Pe′CO₂, which was higher in MT than in M at all post-treatment evaluation times (p < 0.001).Conclusions and clinical relevanceTramadol combined with medetomidine resulted in greater sedation scores (deeper sedation) than medetomidine alone in dogs, and minimal adverse changes in the physiologic variables were measured.     

Effects of perioperative saphenous and sciatic nerve blocks,lumbosacral epidural or morphine–lidocaine–ketamine infusion on postoperative pain and sedation in dogs undergoing tibial plateau leveling osteotomy

ObjectiveTo compare the quality of postoperative analgesia and sedation after preoperative saphenous and sciatic nerve blockade, preoperative lumbosacral epidural injection and perioperative intravenous (IV) morphine, lidocaine and ketamine infusions in dogs undergoing stifle arthroscopy and tibial plateau leveling osteotomy (TPLO) under general anesthesia.Study designProspective, blinded, randomized, clinical comparison study.AnimalsA total of 45 dogs weighing 33.9 (15.9–56.7) kg and aged 5.2 (1.0–12.0) years, mean (range), undergoing elective unilateral TPLO for spontaneous cranial cruciate ligament rupture.MethodsClient-owned dogs were enrolled. Dogs were randomly assigned to one of three groups: group MLK, perioperative IV morphine, lidocaine and ketamine infusion; group EPID, lumbosacral epidural with ropivacaine and morphine; or group SSNB, saphenous and sciatic nerve blockade with ropivacaine. Routine stifle arthroscopy followed by TPLO surgery was performed. Sedation and pain scores were assessed at 0, 2, 4, 8 and 24 hours following extubation. Rescue analgesia was administered as prescribed by Glasgow composite pain score–short form score >5.ResultsSedation scores for MLK were higher than EPID and SSNB. Pain scores for SSNB were lower than those for EPID and MLK. No significant differences were found in anesthesia duration or surgery duration among groups. No dogs required rescue analgesia.Conclusions and clinical relevanceAlthough analgesia was adequate in all groups, the best combination of analgesia without increased sedation was recorded for SSNB.     

Comparison of lidocaine,tramadol, and lidocaine–tramadol for epidural analgesia in lambs

Epidural anesthesia is commonly utilized in veterinary medicine to allow diagnostic, obstetrical, and surgical intervention, in the perineal region of domestic animal. The following study was carried out to directly compare the time of onset and duration of anesthesia produced by a tramadol and lidocaine–tramadol combination with that produced by lidocaine administration in the epidural space of lamb. Seven healthy female lambs of undefined breed weighing 15–20 kg were selected for this study. Epidural anesthesia was produced in all lambs by 2% lidocaine and with 2 weeks intervals repeated by combination of lidocaine–tramadol and tramadol alone. Analgesia was defined as lack of a response to pin prick test and pressure from hemostat clamp (closed to the first ratchet) applied first in the perineal area and then moved cranially toward the thoracic region until a response (movement associated with pin prick test or hemostat pressure) was observed. Time to onset, duration and cranial spread of analgesia were recorded. Heart rate (HR), respiratory rate (RR), and rectal temperature were recorded before (baseline, 0) and at 15, 30, 45, 60, 75, 90, 105 and 120 min after epidural administration of the solution. The results were expressed as mean ± SD and were analyzed by a one-way analysis of variance and Duncan’s test as a post hoc for heart rate, respiratory rate and body temperature and also, for time of onset and duration of analgesia. Graphpad Prism version 5 software program was used for all analyses. A value of P < 0.05 was considered significant. The tramadol produced a significant (P < 0.05) longer duration of analgesia than lidocaine alone and lidocaine–tramadol combination. Also, lidocaine–tramadol combination produced a significant (P < 0.05) longer duration of analgesia than lidocaine alone. Complete analgesia began more delayed in the tramadol treatment than lidocaine–tramadol and lidocaine alone. The combination of lidocaine–tramadol produced analgesia of longer duration than lidocaine and onset time was approximately same as lidocaine group.     

Comparison of lidocaine and lidocaine–epinephrine for the paravertebral brachial plexus block in dogs

Objective

To compare the motor and sensory block efficacy and duration of a modified paravertebral brachial plexus block (PBPB) after administration of lidocaine alone (LI) or combined with epinephrine (LE).

Concentrations of medetomidine enantiomers and vatinoxan,an α2-adrenoceptor antagonist,in plasma and central nervous tissue after intravenous coadministration in dogs

ObjectiveTo quantify the peripheral selectivity of vatinoxan (L-659,066, MK-467) in dogs by comparing the concentrations of vatinoxan, dexmedetomidine and levomedetomidine in plasma and central nervous system (CNS) tissue after intravenous (IV) coadministration of vatinoxan and medetomidine.Study designExperimental, observational study.AnimalsA group of six healthy, purpose-bred Beagle dogs (four females and two males) aged 6.5 ± 0.1 years (mean ± standard deviation).MethodsAll dogs were administered a combination of medetomidine (40 μg kg⁻¹) and vatinoxan (800 μg kg⁻¹) as IV bolus. After 20 minutes, the dogs were euthanized with an IV overdose of pentobarbital (140 mg kg⁻¹) and both venous plasma and CNS tissues (brain, cervical and lumbar spinal cord) were harvested. Concentrations of dexmedetomidine, levomedetomidine and vatinoxan in all samples were quantified by liquid chromatography–tandem mass spectrometry and data were analyzed with nonparametric tests with post hoc corrections where appropriate.ResultsAll dogs became deeply sedated after the treatment. The CNS-to-plasma ratio of vatinoxan concentration was approximately 1:50, whereas the concentrations of dexmedetomidine and levomedetomidine in the CNS were three- to seven-fold of those in plasma.Conclusions and clinical relevanceWith the doses studied, these results confirm the peripheral selectivity of vatinoxan in dogs, when coadministered IV with medetomidine. Thus, it is likely that vatinoxan preferentially antagonizes α₂-adrenoceptors outside the CNS.     

Evaluation of cardiovascular effects of intramuscular medetomidine and a medetomidine–vatinoxan combination in Beagle dogs: A randomized blinded crossover laboratory study

ObjectiveTo compare the cardiovascular effects of a combination of medetomidine and vatinoxan (MVX) versus medetomidine (MED) alone administered intramuscularly (IM) and to determine whether heart rate (HR) can be used as a surrogate for cardiac output (CO) after the use of medetomidine with or without vatinoxan.Study designA randomized, blinded, experimental, crossover study.AnimalsA group of eight healthy Beagle dogs aged 4.6 (2.3–9.4) years and weighing 12.9 (9–14.7) kg, median (range).MethodsEach dog was injected with 1 mg m^–2 medetomidine with or without 20 mg m^–2 vatinoxan IM with a washout period of 7 days. Cardiovascular data and arterial and mixed venous blood gas samples were collected at baseline, 5, 10, 15, 20, 35, 45, 60, 90 and 120 minutes after treatment administration. CO was measured at all time points via thermodilution. Differences between treatments, period and sequence were evaluated with repeated measures analysis of covariance and the relationship between HR and CO was assessed with a repeated measures analysis of variance; p values < 0.05 were deemed significant.ResultsThe CO was 47–96% lower after MED than after MVX (p < 0.0001). Increases in systemic, pulmonary arterial and right atrial pressures and oxygen extraction ratio were significantly higher after MED than after MVX (all p < 0.0001). HR was significantly lower after MED and the linear relationship to CO was significant (p < 0.0001).Conclusions and clinical relevanceOverall, MED affected the cardiovascular system more negatively than MVX, and the difference in cardiovascular function between the treatments can be considered clinically relevant. HR was linearly related to CO, and decreases in HR reflected cardiac performance for dogs sedated with medetomidine with or without vatinoxan.     

Intratesticular and incisional line infiltration with ropivacaine for castration in medetomidine–butorphanol–midazolam sedated dogs

Objective

To evaluate whether intratesticular and incisional ropivacaine infiltration produces sufficient intra- and postoperative analgesia for castrating dogs under sedation.

Intervertebral disc disease in dogs – Part 2: Comparison of clinical,magnetic resonance imaging,and histological findings in 74 surgically treated dogs

The relationship between intervertebral disc (IVD) disease and IVD degeneration remains unclear. The aim of the present study was to compare the clinical severity of IVD herniation (IVDH), determined with a neurological grading system, with findings of magnetic resonance imaging (MRI) and histology using grading systems for IVD degeneration in chondrodystrophic (CD; n = 37) and non-chondrodystrophic (NCD; n = 37) dogs. This study is the second part of a two-part investigation, where the first part involved the development and validation of a histological grading scheme for classification of canine IVD degeneration.IVD degeneration graded on MRI correlated significantly with IVD degeneration graded on histology, but not with pre-operative clinical signs. Hansen type 1 hernias were more common in the cervical and thoracolumbar segments and Hansen type 2 hernias were more common in the lumbosacral segment. Type 1 hernias occurred more often in CD dogs than in NCD dogs, and CD dogs were clinically more severely affected than NCD dogs. The grade of IVD degeneration on MRI was higher in CD dogs than in NCD dogs, but there was no difference between dogs with type 1 and type 2 hernias. No significant differences in histological grade were found between CD and NCD dogs or between type 1 and type 2 hernias.It was possible to conclude that IVD degeneration did not correlate with the neurological severity of IVDH. The extent of degeneration identified on MRI correlated with degeneration seen histologically. Although the MRI grading system reflected the severity of IVD degenerative changes as confirmed by histopathology, it appeared less useful in predicting the clinical implications.     

Comparison of the effects of butorphanol–midazolam–medetomidine and butorphanol–azaperone–medetomidine in wild common palm civets (Paradoxurus musangus)

ObjectiveTo assess the efficacy of butorphanol–azaperone–medetomidine (BAM) and butorphanol–midazolam–medetomidine (BMM) protocols for immobilization of wild common palm civets (Paradoxurus musangus) with subsequent antagonization with atipamezole.Study designProspective, randomized, blinded clinical trial.AnimalsA total of 40 adult wild common palm civets, 24 female and 16 male, weighing 1.5–3.4 kg.MethodsThe civets were randomly assigned for anesthesia with butorphanol, azaperone and medetomidine (0.6, 0.6 and 0.2 mg kg^–1, respectively; group BAM) or with butorphanol, midazolam and medetomidine (0.3, 0.4 and 0.1 mg kg^–1, respectively; group BMM) intramuscularly (IM) in a squeeze cage. When adequately relaxed, the trachea was intubated for oxygen administration. Physiological variables were recorded every 5 minutes after intubation. Following morphometric measurements, sampling, microchipping and parasite treatment, medetomidine was reversed with atipamezole at 1.0 or 0.5 mg kg^–1 IM to groups BAM and BMM, respectively. Physiological variables and times to reach the different stages of anesthesia were compared between groups.ResultsOnset time of sedation and recumbency was similar in both groups; time to achieve complete relaxation and tracheal intubation was longer in group BAM. Supplementation with isoflurane was required to enable intubation in five civets in group BAM and one civet in group BMM. All civets in group BAM required topical lidocaine to facilitate intubation. End-tidal carbon dioxide partial pressure was lower in group BAM, but heart rate, respiratory rate, rectal temperature, peripheral hemoglobin oxygen saturation and mean arterial blood pressure were not different. All civets in both groups recovered well following administration of atipamezole.Conclusions and clinical relevanceBoth BAM and BMM combinations were effective for immobilizing wild common palm civets. The BMM combination had the advantage of producing complete relaxation that allowed intubation more rapidly.     

Determination of the effective dosage of tiletamine–zolazepam–ketamine–xylazine,with or without methadone,in dogs

ObjectiveTo determine the effective dosage of the combination tiletamine–zolazepam–ketamine–xylazine (TKX), with or without methadone, in dogs.Study designProspective, randomized, experimental study.AnimalsA total of 29 dogs.MethodsDogs were randomly administered TKX (group TKX, n = 13) or combined with 0.3 mg kg^–1 of methadone (group TKXM, n = 16) intramuscularly. The TKX solution contained tiletamine (50 mg mL^–1), zolazepam (50 mg mL^–1), ketamine (80 mg mL^–1) and xylazine (20 mg mL^–1). The effective dosages for immobility in 50% and 95% of the population (ED₅₀ and ED₉₅) were estimated using the up-and-down method. Approximately 20 minutes after drug administration, a skin incision was performed and the response was judged as positive or negative if the dogs moved or did not move, respectively. The TKX volume for the subsequent dog in the same group was increased or decreased by 0.005 mL kg^–1 if the response of the previous dog was positive or negative, respectively. Heart and respiratory rates, and sedation/anesthesia scores (range 0–21) were recorded before and 15 minutes after drug administration.ResultsEstimated ED₅₀ and ED₉₅ (95% confidence intervals) were: TKX, 0.025 (0.020–0.029) and 0.026 (0.010–0.042) mL kg^–1; TKXM, 0.022 (0.018–0.025) and 0.033 (0.017–0.049) mL kg^–1. Median (interquartile range) scores for sedation/anesthesia were 17 (16–18) and 17 (15–20), and times until lateral recumbency were 5 (4–6) and 6 (4–10) minutes in TKX and TKXM, respectively (p > 0.05). In both groups heart and respiratory rates decreased, but values remained acceptable for anesthetized dogs.Conclusions and clinical relevanceThe results provide a guide for volumes of TKX and TKXM in dogs requiring restraint for minimally invasive procedures. Inclusion of methadone in the TKX combination did not influence ED₅₀.     

Comparison of three continuous positive airway pressure (CPAP) interfaces in healthy Beagle dogs during medetomidine–propofol constant rate infusions

Objective

To compare the efficacy of three continuous positive airway pressure (CPAP) interfaces in dogs on gas exchange, lung volumes, amount of leak during CPAP and rebreathing in case of equipment failure or disconnection.

Effects of fentanyl–lidocaine–ketamine versus sufentanil–lidocaine–ketamine on the isoflurane requirements in dogs undergoing total ear canal ablation and lateral bulla osteotomy

ObjectiveTo compare the isoflurane-sparing effects of sufentanil–lidocaine–ketamine (SLK) and fentanyl–lidocaine–ketamine (FLK) infusions in dogs undergoing total ear canal ablation and lateral bulla osteotomy (TECA–LBO).Study designRandomized blinded clinical study.AnimalsA group of 20 client-owned dogs undergoing TECA–LBO.MethodsIntravenous (IV) administration of lidocaine (3 mg kg^–1) and ketamine (0.6 mg kg^–1) with fentanyl (5.4 μg kg^–1; n = 10; FLK group) or sufentanil (0.72 μg kg^–1; n = 10; SLK group) was immediately followed by the corresponding constant rate infusion (CRI) (lidocaine 3 mg kg^–1 hour^–1; ketamine 0.6 mg kg^–1 hour^–1; either fentanyl 5.4 μg kg^–1 hour^–1 or sufentanil 0.72 μg kg^–1 hour^–1). Anaesthesia was induced with propofol 3–5 mg kg^–1 IV and was maintained with isoflurane. End-tidal isoflurane concentration (Fe′Iso) was decreased in 0.2% steps every 15 minutes until spontaneous movements were observed (treated with propofol 1 mg kg^–1 IV) or an increase of > 30% in heart rate or mean arterial pressure from baseline occurred (treated with rescue fentanyl or sufentanil). Quality of recovery and pain were assessed at extubation using the short-form Glasgow Composite Pain Scale (SF-GCPS), Colorado State University Canine Acute Pain scale (CSU-CAP), and visual analogue scale (VAS). Data were analysed with analysis of variance, t tests, Fisher test and Spearman coefficient (p < 0.05).ResultsFe′Iso decreased significantly in SLK group (45%; p = 0.0006) but not in FLK (15%; p = 0.1135) (p = 0.0136). SLK group had lower scores for recovery quality (p = 0.0204), SF-GCPS (p = 0.0071) and CSU-CAP (p = 0.0273) than FLK at extubation. Intraoperative rescue analgesia and VAS were not significantly different between groups.Conclusions and clinical relevanceCompared with FLK infusion, CRI of SLK at these doses decreased isoflurane requirements, decreased pain scores and improved recovery quality at extubation in dogs undergoing TECA–LBO.     

Plasma histamine concentrations in horses administered sodium penicillin,guaifenesin–xylazine–ketamine and isoflurane with morphine or butorphanol

ObjectiveVarious drugs administered to horses undergoing surgical procedures can release histamine. Histamine concentrations were evaluated in horses prepared for surgery and administered butorphanol or morphine intraoperative infusions.Study designProspective studies with one randomized.AnimalsA total of 44 client-owned horses.MethodsIn one study, anesthesia was induced with xylazine followed by ketamine–diazepam. Anesthesia was maintained with guaifenesin–xylazine–ketamine (GXK) during surgical preparation. For surgery, isoflurane was administered with intravenous (IV) morphine (group M: 0.15 mg kg^–1 and 0.1 mg kg^–1 hour^–1; 15 horses) or butorphanol (group B: 0.05 mg kg^–1 and 0.01 mg kg^–1 hour^–1; 15 horses). Histamine and morphine concentrations were measured using enzyme-linked immunoassay before opioid injection (time 0), and after 1, 2, 5, 30, 60 and 90 minutes. In a subsequent study, plasma histamine concentrations were measured in 14 horses before drug administration (baseline), 15 minutes after IV sodium penicillin and 15 minutes after starting GXK IV infusion. Statistical comparison was performed using anova for repeated measures. Pearson correlation compared morphine and histamine concentrations. Data are presented as mean ± standard deviation. Significance was assumed when p ≤ 0.05.ResultsWith histamine, differences occurred between baseline (3.2 ± 2.4 ng mL^–1) and GXK (5.2 ± 7.1 ng mL^–1) and between baseline and time 0 in group B (11.9 ± 13.4 ng mL^–1) and group M (11.1 ± 12.4 ng mL^–1). No differences occurred between baseline and after penicillin or between groups M and B. Morphine concentrations were higher at 1 minute following injection (8.1 ± 5.1 ng mL^–1) than at 30 minutes (4.9 ± 3.1 ng mL^–1) and 60 minutes (4.0 ± 2.5 ng mL^–1). Histamine correlated with morphine at 2, 30 and 60 minutes.Conclusions and clinical relevanceGXK increased histamine concentration, but concentrations were similar with morphine and butorphanol.     

Clinical evaluation of intranasal medetomidine–ketamine and medetomidine–S(+)-ketamine for induction of anaesthesia in rabbits in two centres with two different administration techniques

Objective

The aim was to compare efficacy and side effects of induction with medetomidine–ketamine or medetomidine–S(+)-ketamine by intranasal (IN) instillation in rabbits and to evaluate both protocols during subsequent isoflurane anaesthesia.

Survey of bacterial microorganisms in the conjunctival sac of clinically normal dogs and dogs with ulcerative keratitis in Fortaleza, Ceará, Brazil

OBJECTIVE: The ocular microflora in dogs has not been established in north-east Brazil. Thus, the main aim of this research was to determine the bacterial microorganisms in the conjunctival sac of clinically normal dogs and dogs with ulcerative keratitis in Fortaleza, Ceará, Brazil. ANIMALS STUDIED: This study included 60 healthy dogs, 15 dogs with unilateral corneal ulcer, and three dogs with bilateral corneal ulcers. Procedure Samples were taken by a calibrated platinum loop (1 microL) placed directly onto the conjunctival sac and on sterile blood agar. The clinical specimens were incubated at 37 degrees C in an atmosphere of 5% CO2 for 48 h. RESULTS: Of the 120 samples from healthy dogs, only 47 (39%) had positive culture for bacteria, while all of the specimens from eyes with corneal ulcer were positive for bacterial growth. The group of dogs with corneal ulcer had a higher (P < 0.05) number of colony-forming units (CFU) per plate than the group of healthy animals. Of the 59 isolates from healthy eyes, only nine (15.3%) had more than 50 CFU per plate, while in the group of dogs with corneal ulcer, 23 (62.2%) of the 37 isolates presented more than 50 CFU per plate. In both groups Gram-positive bacteria (86.5%) predominated over Gram-negative (13.5%). Staphylococcus spp. was the most frequently isolated genus and S. intermedius predominated in both groups. CONCLUSION: The results of our study are directly applicable to initiate rational, preventive and therapeutic measures with greater accuracy in dogs with corneal ulcer.