Use of the oxygen content-based index, Fshunt, as an indicator of pulmonary venous admixture at various inspired oxygen fractions in anesthetized sheep.

OBJECTIVE: To evaluate the use of the oxygen content-based index, Fshunt, as an indicator of venous admixture (Qs/Qt) at various fractions of inspired oxygen (FIO(2)s) in anesthetized sheep undergoing Flung or 2-lung ventilation. Animals-6 healthy adult female sheep. PROCEDURES: Sheep were anesthetized and administered 5 different FIO(2)s (0.21, 0.40, 0.60, 0.80, and 1.00) in random order during 2-lung mechanical ventilation. Arterial and mixed venous blood samples were obtained at each FIO(2) after a 15-minute stabilization period. Vital capacity alveolar recruitment maneuvers were performed after blood collection. The previously used FIO(2) sequence was reversed for sample collection during Flung ventilation. Blood samples were analyzed for arterial, pulmonary end-capillary, and mixed venous oxygen content and partial pressure and for hemoglobin concentration. Oxygen hemoglobin saturation, Qs/Qt, Fshunt, and oxygen tension-based indices (OTIs; including PAO(2):FIO(2), alveolar-arterial difference in partial pressure of oxygen [PAO(2) - PAO(2)], [PAO(2) - Pao(2)]:FIO(2), [PAO(2) - PAO(2)]:PAO(2), and PAO(2):PAO(2)) were calculated at each FIO(2); associations were evaluated with linear regression analysis, concordance, and correlation tests. Intermethod agreement between Qs/Qt and Fshunt was tested via Bland-Altman analysis. RESULTS: Strong and significant associations and substantial agreement were detected between Fshunt and Qs/Qt. Relationships between OTIs and Qs/Qt varied, but overall correlations were weak. Conclusions and Clinical Relevance-Whereas OTIs were generally poor indicators of Qs/Qt, Fshunt was a good indicator of Qs/Qt at various FIO(2)s, regardless of the magnitude of Qs/Qt, and could be potentially used as a surrogate for Qs/Qt measurements in healthy sheep.

RECOVER evidence and knowledge gap analysis on veterinary CPR. Part 6: Post-cardiac arrest care.

OBJECTIVE: To systematically examine the evidence for interventions after the return of spontaneous circulation (ROSC) on outcomes from veterinary cardiopulmonary resuscitation and to determine important knowledge gaps. DESIGN: Standardized, systematic evaluation of the literature, categorization of relevant articles according to level of evidence and quality, and development of consensus on conclusions for application of the concepts to clinical post-cardiac arrest care. SETTING: Academia, referral practice, and general practice. RESULTS: Fifteen standardized clinical questions important for post-cardiac arrest care were asked and research articles relevant to answering these questions were identified through structured, explicit literature database searches. The majority of these articles report research in species other than dogs or cats or consisted of experimental work in canine cardiac arrest models. Outcome metrics reported in these studies widely varied and ranged from quantification of mechanistic endpoints, such as elaboration of reactive oxygen species, to survival, and functional neurologic outcome. CONCLUSIONS: Despite the near complete absence of clinical veterinary studies, the process allowed the formulation of statements for several postcardiac arrest treatments that were either supportive, such as mild therapeutic hypothermia or controlled reoxygenation, or neutral, such as for mannitol administration or seizure prophylaxis. Evidence grading allowed transparency in regards to the strength of these recommendations. Moreover, numerous knowledge gaps emerged that will allow generation of a road map for progress in veterinary post-cardiac arrest care.

The acid-base effects of free water removal from and addition to oxygenated and deoxygenated whole blood: an in vitro model of contraction alkalosis and dilutional acidosis.

This study was conducted to describe the acid-base effects of hydration and dehydration of oxygenated and deoxygenated whole blood. Whole blood samples from goats were equilibrated in a tonometer to a partial pressure of carbon dioxide of 40 mm Hg and oxygen (PO2) of 100 mm Hg or 30 mm Hg. Contraction alkalosis was achieved by evaporating blood samples to 80% of the original volume. Dilutional acidosis was achieved by increasing the blood sample volume by 20% by addition of sterile water. Acid-base, electrolyte, hemoglobin, lactate, albumin, and phosphorus concentrations were measured at baseline and after dehydration or hydration. A 20% dehydration of whole blood caused a 22% increase in sodium concentration and a significant increase in base excess of +3 mEq/L (P < 0.01); bicarbonate concentration increased only 7% to 9%. A concurrent increase was found in phosphorus, albumin, hemoglobin, and lactate concentrations. A 20% dilution of whole blood caused a 21% decrease in sodium concentration and a significant decrease in base excess of -5 mEq/L (P < 0.01) with an 11% to 15% decrease in bicarbonate concentration. A concurrent decrease was found in phosphorus, albumin, and hemoglobin concentrations. No significant difference was observed between the acid-base effects on oxygenated versus deoxygenated blood in any experiment. Dilutional acidosis and contraction alkalosis of whole blood are complex acid-base disorders resulting from direct changes in bicarbonate concentration in combination with changes in the concentration of weak plasma acids and buffering reactions. Therefore, bicarbonate concentration does not change to the same degree as the magnitude of contraction or dilution.

Effects of acute dilutional hyponatremia on acid-base changes and electrolyte concentrations in rats with bilateral renal pedicle ligation.

OBJECTIVE: To describe the effects of increasing the extracellular fluid (ECF) volume by approximately 20% on acid-base changes and electrolyte concentrations in anesthetized rats. ANIMALS: 18 adult male Sprague-Dawley rats. PROCEDURES: Rats were assigned to a control group (n = 6 rats) and a treatment group (12). All rats were anesthetized, and instrumentation and bilateral renal pedicle ligation were performed. The treatment group was infused IV with sterile water throughout a 30-minute period. Acid-base variables and concentrations of electrolytes, lactate, albumin, phosphorus, and hemoglobin were measured before (baseline) and 30 and 60 minutes after onset of infusion. Anion gap, strong ion difference, strong ion gap, and contributions of sodium, chloride, albumin, phosphorus, and lactate concentrations to base excess were calculated at each time point. RESULTS: Infusion of sterile water led to an increase in ECF volume of approximately 18%. This had no effect on acid-base balance, compared with that in control rats. Infusion of sterile water caused a significant decrease in sodium, chloride, ionized calcium, lactate, and albumin concentrations, compared with concentrations in the control group. Anion gap and calculated effects of sodium, chloride, albumin, and lactate concentrations on base excess at 60 minutes differed significantly between infused and control rats. CONCLUSIONS AND CLINICAL RELEVANCE: Infusion of sterile water did not cause clinically relevant dilutional acidosis. The acidotic impact of water administration was offset by generation of new bicarbonate via carbonic acid equilibration and intracellular buffering in combination with the alkalotic effects of decreases in albumin, phosphorus, and lactate concentrations.

The cardiopulmonary effects of dobutamine and norepinephrine in isoflurane-anesthetized foals.

OBJECTIVE: To evaluate the cardiovascular effects of norepinephrine (NE) and dobutamine (DB) in isoflurane-anesthetized foals. STUDY DESIGN: Prospective laboratory study. METHODS: Norepinephrine (0.05, 0.10, 0.20, and 0.40 microg kg(-1) minute(-1)) and dobutamine (2.5, 5.0, and 10 microg kg(-1) minute(-1)) were alternately administered to seven healthy, 1- to 2-week-old isoflurane-anesthetized foals. Arterial and pulmonary arterial blood pressure, right atrial pressure, pulmonary artery occlusion pressure, heart rate, body temperature, cardiac output, arterial and mixed venous blood pH, partial pressure of carbon dioxide, partial pressure of oxygen [arterial partial pressure of oxygen (PaO(2)) and mixed venous partial pressure of oxygen (PvO(2))], and packed cell volume were measured. Standard base excess, bicarbonate concentration, systemic and pulmonary vascular resistance, cardiac index (CI), stroke volume, left and right stroke work indices, oxygen delivery (DO(2)), consumption, and extraction were calculated. Results Norepinephrine infusion resulted in significant increases in arterial and pulmonary arterial pressure, systemic and pulmonary vascular resistance indices, and PaO(2); heart rate was decreased. Dobutamine infusion resulted in significant increases in heart rate, stroke volume index, CI, and arterial and pulmonary arterial blood pressure. Systemic and pulmonary vascular resistance indices were decreased while the ventricular stroke work indices increased. The PaO(2) decreased while DO(2) and oxygen consumption increased. Oxygen extraction decreased and PvO(2) increased. CONCLUSIONS AND CLINICAL RELEVANCE: Norepinephrine primarily augments arterial blood pressure while decreasing CI. Dobutamine primarily augments CI with only modest increases in arterial blood pressure. Both NE and DB could be useful in the hemodynamic management of anesthetized foals.

In vitro comparison of the effects of two forms of hydroxyethyl starch solutions on platelet function in dogs.

OBJECTIVE: To evaluate the effect of 2 hydroxyethyl starch (HES) preparations (ie, HES solution with a molecular weight of 600 kd and a degree of substitution of 0.7 [HES 600/0.7] and a calcium-containing polyionic HES solution with a molecular weight of 670 kd and a degree of substitution of 0.75 [HES 670/0.75]) on canine platelet function. SAMPLE POPULATION: Blood samples from 10 healthy adult dogs. PROCEDURES: Dilution of citrated whole blood was performed with saline (0.9% NaCl) solution, HES 600/0.7, and HES 670/0.75 at ratios of 1:9 (ie, 1 part saline solution or colloid to 9 parts whole blood) and 1:3. Measurements of time to platelet plug formation in a capillary tube (ie, closure time) were made by use of a bench-top platelet function analyzer with collagen and ADP platelet agonists. RESULTS: Mean baseline closure time was 68.0 +/- 15.3 seconds. A 1:3 dilution of whole blood with saline solution, HES 600/0.7, and HES 670/0.75 resulted in mean closure times of 85.8 +/- 15.7 seconds, 100.6 +/- 18.6 seconds, and 101.6 +/- 16.2 seconds, respectively. Closure time following 1:3 dilution of whole blood with saline solution was significantly different from baseline and from 1:9 dilution with saline solution. Closure time following 1:3 dilution of whole blood with HES 670/0.75 was significantly different from baseline, 1:3 and 1:9 dilutions with saline solution, and 1:9 dilutions with HES 600/0.7 or HES 670/0.75. CONCLUSIONS AND CLINICAL RELEVANCE: Saline solution, HES 600/0.7, and HES 670/0.75 affect canine platelet function by prolonging closure times; HES solutions prolonged closure time to a greater extent than saline solution.

Relative adrenal insufficiency in dogs with sepsis.

BACKGROUND: A syndrome of relative adrenal insufficiency has been identified in septic humans, and is associated with hypotension and death. Relative adrenal insufficiency is generally associated with basal serum cortisol concentration within or above the reference range and a blunted cortisol response to adrenocorticotropic hormone administration. It is unknown whether relative adrenal insufficiency occurs in septic dogs. HYPOTHESIS: That relative adrenal insufficiency occurs in septic dogs, and that relative adrenal insufficiency is associated with hypotension and mortality. ANIMALS: Thirty-three septic dogs admitted to a small animal intensive care unit. METHODS: Dogs were included in the study if they had a known or suspected infectious disease and had systemic inflammatory response syndrome. Dogs were excluded if they had disease or medication history expected to affect the hypothalamic-pituitary-adrenal axis. Serum cortisol and endogenous plasma adrenocorticotropic hormone concentrations were measured before, and serum cortisol concentration measured 1 hour after, intramuscular administration of 250 microg of cosyntropin/dog. The change in cortisol concentration (delta-cortisol) before and after cosyntropin administration was determined in each dog. RESULTS: Hypotension was associated with lower delta-cortisol values (OR 1.3; CI 1.0-1.9; P = .029). delta-Cortisol cutoff of 3.0 microg/dL was most accurate for predicting hypotension, survival to discharge, and 28-day survival. The rate of death in dogs with delta-cortisol < or = 3 microg/dL was 4.1 times that of dogs with delta-cortisol > 3 microg/dL (RR 4.1; CI 1.5-12.3; P = .01). CONCLUSIONS AND CLINICAL RELEVANCE: Delta-cortisol < or = 3 microg/dL after adrenocorticotropic hormone administration is associated with systemic hypotension and decreased survival in septic dogs.

Adverse effect of complete thoracic evacuation during chronic pneumothorax in a dog.

OBJECTIVE: To describe an unusual response to evacuation of a chronic pneumothorax in a dog. CASE: Resolution of a pneumothorax is usually associated with lung expansion and a marked improvement in pulmonary function. In the case presented here, evacuation of a chronic pneumothorax improved oxygenation, but markedly impaired ventilation. CLINICAL RELEVANCE: This is a previously unreported complication associated with the treatment of pneumothorax in dogs. It suggests that in some cases of chronic pneumothorax, total evacuation of the pleural air pocket may improve oxygenation but ventilation deteriorates and increases PCO2, causing further complications.

Indications, management, and outcome of long-term positive-pressure ventilation in dogs and cats: 148 cases (1990-2001).

OBJECTIVE: To determine outcome of positive-pressure ventilation (PPV) for 24 hours or longer and identify factors associated with successful weaning from PPV and survival to hospital discharge in dogs and cats. DESIGN: Retrospective case series. ANIMALS: 124 dogs and 24 cats that received PPV for 24 hours or longer. PROCEDURES: Medical records were reviewed for signalment, primary diagnosis, reason for initiating PPV, measures of oxygenation and ventilation before and during PPV, ventilator settings, complications, duration of PPV, and outcome. Animals were categorized into 1 of 3 groups on the basis of the reason for PPV. RESULTS: Group 1 patients received PPV for inadequate oxygenation (67 dogs and 6 cats), group 2 for inadequate ventilation (46 dogs and 16 cats), and group 3 for inadequate oxygenation and ventilation (11 dogs and 2 cats). Of the group 1 animals, 36% (26/73) were weaned from PPV and 22% (16/73) survived to hospital discharge. In group 2, 50% (31/62) were weaned from PPV and 39% (24/62) survived to hospital discharge. In group 3, 3 of 13 were weaned from PPV and 1 of 13 survived to hospital discharge. Likelihood of successful weaning and survival to hospital discharge were significantly higher for group 2 animals, and cats had a significantly lower likelihood of successful weaning from PPV, compared with dogs. Median duration of PPV was 48 hours (range, 24 to 356 hours) and was not associated with outcome. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggested that long-term PPV is practical and successful in dogs and cats.

Evaluation of peripheral and central venous pressure in awake dogs and cats.

OBJECTIVE: To determine whether peripheral venous pressure (PVP) was correlated with central venous pressure (CVP) when measured by use of different catheter sizes, catheterization sites, and body positions in awake dogs and cats. ANIMALS: 36 dogs and 10 cats. PROCEDURES: Dogs and cats with functional jugular and peripheral venous catheters were enrolled in the study. Peripheral venous catheters (18 to 24 gauge) were placed in a cephalic, lateral saphenous, or medial saphenous vein. Central venous catheters (5.5 to 8.5 F) were placed in the jugular vein and advanced into the cranial vena cava. Catheters were connected to pressure transducers and a blood pressure monitor capable of displaying 2 simultaneous pressure tracings. For each animal, the mean of 5 paired measurements of PVP and CVP was calculated. The relationship between PVP and CVP when measured by use of different catheter sizes, catheterization sites, and body positions was determined. RESULTS: Mean +/- SD PVP was 5.7 +/- 5.8 mm Hg higher than CVP in dogs and 6.0 +/- 6.9 mm Hg higher than CVP in cats. However, results of multiple regression analysis did not indicate a significant correlation between PVP and CVP, regardless of catheter size, catheter position, or body position. The relationship was weak in both dogs and cats. CONCLUSIONS AND CLINICAL RELEVANCE: The PVP was poorly correlated with CVP when different catheter sizes, catheterization sites, and patient positions were evaluated. Peripheral venous pressure should not be used to approximate CVP in awake dogs and cats.

The acid-base impact of free water removal from, and addition to, plasma.

Water, compared with plasma at a pH of 7.4, is a weak acid. The addition of free water to a patient should have an acidifying effect (dilutional acidosis) and the removal of it, an alkalinizing effect (concentrational alkalosis). The specific effects of free water loss or gain in a relatively complex fluid such as plasma has, to the authors' knowledge, not been reported. This information would be useful in the interpretation of the effect of changes in free water in patients. Plasma samples from goats were either evaporated in a tonometer to 80% of baseline volume or hydrated by the addition of distilled water to 120% of baseline volume. The pH and partial pressure of carbon dioxide, sodium, potassium, ionized calcium, chloride, lactate, phosphorous, albumin, and total protein concentrations were measured. Actual base excess (ABE), standard bicarbonate, anion gap, strong ion difference, strong ion gap, unmeasured anions, and the effects of sodium, chloride, phosphate, and albumin changes on ABE were calculated. Most parameters changed 20% in proportion to the magnitude of dehydration or hydration. Bicarbonate concentration, however, increased only 11% in the evaporation trial and decreased only -2% in the dehydration trial. The evaporation trial was associated with a mild, but significant, metabolic alkalotic effect (ABE increased 3.2 mM/L), whereas the hydration trial was associated with a slight, insignificant metabolic acidotic effect (ABE decreased only 0.6 mM/L). The calculated free water ABE effect (change in sodium concentration) was offset by opposite changes in calculated chloride, lactate, phosphate, and albumin ABE effects.

The caprine oxyhemoglobin dissociation curve.

The caprine oxyhemoglobin dissociation curve has not been previously defined. Blood from 10 healthy goats was equilibrated in a tonometer with calibrated gas mixtures of oxygen at concentrations of 95%, 21%, 13%, 12%, 10%, 9%, 8%, 5%, 4%, and 2.5% , 5% carbon dioxide, balance nitrogen. The pH, partial pressure of oxygen (PO2), partial pressure of carbon dioxide (PCO2), total hemoglobin, oxyhemoglobin saturation, carboxyhemoglobin, methemoglobin, and oxygen content were measured. The PO2/oxyhemoglobin and the PO2/oxygen content relationships were graphed with curve-fitting software and a formula for calculating oxyhemoglobin from PO2 was generated. The maximum oxygen content per gram of hemoglobin was 1.29 ml of oxygen per gram of hemoglobin. The PO2 at which hemoglobin was 50% saturated (P50) from the PO2/oxyhemoglobin relationship was 28.6+/-1.5 mmHg and that from the PO2/oxygen content relationships was 29.1+/-1.6 mmHg. The Hill coefficient for the PO2/oxyhemoglobin data was 3.0+/-0.4.

Effects of oral administration of a commercial activated charcoal suspension on serum osmolality and lactate concentration in the dog.

The purpose of this investigation was to determine the effects of an activated charcoal (AC) suspension containing propylene glycol and glycerol on serum osmolality, osmolal gap, and lactate concentration in dogs. Six healthy adult dogs were administered 4 g/kg AC in a commercially available suspension that contained propylene glycol and glycerol as vehicles. Blood samples were taken before and 1, 4, 6, 8, 12, and 24 hours after the administration of the test suspension. Samples were analyzed for osmolality, blood gases, and concentrations of lactate, sodium, potassium, serum urea nitrogen, and glucose. Osmolal gaps were calculated for each time point. Mean serum osmolality, osmolal gap, and lactate concentration were significantly increased after suspension administration compared to baseline. Serum osmolality increased from 311 mOsm/kg at baseline to 353 mOsm/kg, osmolal gap increased from 5 to 52 mOsm/kg, and lactate concentration increased from 1.9 to 4.5 mmol/L after suspension administration (all P < .01). Three of the 6 dogs vomited between 1 and 3 hours after the administration of the test suspension, and 4 of 6 dogs were lethargic. All dogs drank frequently after AC administration. Commercial AC suspension administered at a clinically relevant dose increases serum osmolality, osmolal gap, and lactate concentration in dogs. These laboratory measures and the clinical signs of vomiting, lethargy, and increased frequency of drinking might complicate the diagnosis or monitoring of some intoxications (such as ethylene glycol) in dogs that have previously received AC suspension containing propylene glycol, glycerol, or both as vehicles.

Comparison of measured oxyhemoglobin saturation and oxygen content with analyzer-calculated values and hand-calculated values obtained in unsedated healthy dogs.

OBJECTIVE: To compare direct measurements of canine oxyhemoglobin (HbO2) saturation and blood oxygen content (ContO2) in healthy dogs with analyzer-calculated values derived by use of a human HbO2 relationship and with hand-calculated values derived by use of a canine HbO2 relationship. ANIMALS: 17 healthy dogs. PROCEDURE: 3-mL samples of heparinized arterial and jugular venous blood were collected from each dog. The pH, Pco2, Po2, hemoglobin, HbO2, carboxyhemoglobin, methemoglobin, and ContO2 were measured; HbO2 and ContO2 were calculated automatically by analyzers and also hand-calculated. Blood gas analyzer-calculated and hand-calculated HbO2 values were compared with co-oximeter-measured HbO2 values. Analyzer-calculated and hand-calculated ContO2 values were compared with oxygen content analyzer-measured values. RESULTS: Hand-calculated HbO2 values for arterial and jugular venous samples were slightly but significantly lower than those calculated by a blood gas analyzer or obtained from a co-oximeter. Hand-calculated and analyzer-calculated arterial and venous ContO2 were similar to measured values. CONCLUSIONS AND CLINICAL RELEVANCE: Although certain HbO2 and ContO2 values generated by use of the different methods were significantly different, these differences are unlikely to be clinically important in healthy dogs.

Assessment of the effect of dilution of blood samples with sodium heparin on blood gas, electrolyte, and lactate measurements in dogs.

OBJECTIVE: To evaluate the effect of dilution of blood samples with sodium heparin on blood gas, electrolyte, and lactate measurements in dogs. Sample Population-Venous blood samples collected from 6 adult dogs of various breeds. PROCEDURE: Syringes were prepared with anticoagulant via 1 of 4 techniques, and the residual volume of liquid heparin in each type of prepared syringe was determined. Blood gas values and other selected clinicopathologic variables were measured in whole blood samples after collection (baseline) and after aliquots of the samples were diluted with heparin via 1 of the 4 manual syringe techniques. By use of a tonometer, whole blood samples were adjusted to 1 of 3 oxygen concentrations (40, 100, or 600 mm Hg) and the PO2 values were measured at baseline and subsequent to the 4 heparin dilutions. RESULTS: The 4 syringe techniques resulted in 3.9%, 9.4%, 18.8%, and 34.1% dilutions of a 1-mL blood sample. Compared with baseline values, dilution of blood samples with liquid heparin significantly changed the measured values of PCO2, PO2, and base deficit and concentrations of electrolytes and lactate. Of the variables assessed, measurement of ionized calcium concentration in blood was most affected by heparin dilution. CONCLUSIONS AND CLINICAL RELEVANCE: These findings in dogs indicate that dilution of blood samples with heparin can be a source of preanalytical error in blood gas, electrolyte, and lactate measurements. Limiting dilution of blood samples with heparin to < 4% by volume via an evacuation technique of syringe heparinization is recommended.

Incidence of catheter-associated urinary tract infection among dogs in a small animal intensive care unit.

OBJECTIVE: To determine incidence of and possible risk factors for catheter-associated urinary tract infection (UTI) among dogs hospitalized in an intensive care unit and compare results of bacterial culture of urine samples with results of bacterial culture of catheter tips. DESIGN: Prospective study. ANIMALS: 39 dogs. PROCEDURE: A standard protocol for aseptic catheter placement and maintenance was used. Urine samples were obtained daily and submitted for bacterial culture. When possible, the urinary catheter tip was collected aseptically at the time of catheter removal and submitted for bacterial culture. Bacteria that were obtained were identified and tested for antimicrobial susceptibility. RESULTS: 4 of the 39 (10.3%) dogs developed a UTI. The probability of remaining free from UTI after 1 day in the intensive care unit was 94.9%, and the probability of remaining free from UTI after 4 days was 63.3%. Bacteria isolates were generally common urinary tract pathogens and were susceptible to most antimicrobials. Specific risk factors for catheter-associated UTI, beyond a lack of antimicrobial administration, were not identified. Positive predictive value of bacterial culture of urinary catheter tips was only 25%. CONCLUSIONS AND CLINICAL RELEVANCE: Results suggest that placement of an indwelling urinary catheter in dogs is associated with a low risk of catheter-associated UTI during the first 3 days after catheter placement, provided that adequate precautions are taken for aseptic catheter placement and maintenance. Results of bacterial culture of urinary catheter tips should not be used to predict whether dogs developed catheter-associated UTI.

Ivermectin toxicity in 17 collies.

Ivermectin is widely used in veterinary medicine as an anthelminthic and generally has a wide margin of safety, but Collies are prone to ivermectin toxicity. Two groups of Collies were presented to the University of California Veterinary Medical Teaching Hospital (VMTH) with ivermectin toxicity. The medical records of the 2 groups of Collies were reviewed retrospectively. Group I comprised 5 adult Collies that received at least 400 microg/kg ivermectin p.o. and were presented to the VMTH 3 hours after intoxication. These Collies showed marked clinical signs on presentation. Three of these dogs required mechanical ventilation and were euthanized for financial reasons; the remaining 2 dogs were comatose but recovered in 5-7 days. Group II was comprised of 12 adult Collies presented to the VMTH 2 days (n = 10) and 5 days (n = 2) after subcutaneous injection of 200-250 microg/kg ivermectin. These animals showed greater variation in severity of illness among individuals; 5 animals progressed to stupor or coma, whereas 4 animals remained ambulatory. Most of these dogs' clinical signs deteriorated from the day of intoxication until approximately day 6, from which time they showed gradual but steady improvement. All of the Collies in this group survived, but it took 3 weeks for most of them to recover. Collies suffering from ivermectin toxicity can have a severe and prolonged clinical course requiring intensive nursing care. Respiratory, cardiovascular, and nutritional support may all be required. With appropriate care, however, the prognosis for complete recovery is good.