Materials and Methods

Animals—Fifty client-owned dogs > 16 weeks of age admitted for elective procedures requiring general anesthesia at the University of Missouri Veterinary Medical Teaching Hospital were enrolled in the study. In accordance with the requirements of the University of Missouri Animal Care and Use Committee protocol for use of client-owned animals in research, all dog owners were asked to sign a consent form prior to enrollment of their dog in the study. Procedures included orthopedic and soft tissue surgical procedures, dental procedures, ophthalmic procedures, and computed tomographic imaging. Each dog was reported by its owner to have been eating and drinking normally at home prior to admission, and food had been withheld beginning between 10:00 PM and midnight the night before the procedure. Dogs had not received any fluids IV prior to anesthesia.

Sample collection—Approximately 1 mL of blood was collected from a peripheral or central vein of each dog between 2 and 24 hours prior to anesthesia, and a second 1-mL sample was collected within 1 hour after extubation. All blood samples were stored at 4°C in tubes containing lithium heparin for up to 2 hours prior to testing for COP, PCV, and TS concentration. Colloid oncotic pressure was measured with a colloid osmometer^b used in accordance with the manufacturer's instructions. Packed cell volume was determined by centrifugation of a microhematocrit tube and determining the proportion of the sample that was composed of concentrated RBCs. Total solids concentration was determined by use of a refractometer.^c

Each patient underwent general anesthesia by use of a combination of injectable and inhalant drugs under the guidance of a board-certified veterinary anesthesiologist. Anesthetic protocols were not standardized, but each patient was premedicated with an opioid drug, α₂-adrenoceptor agonist, or phenothiazine sedative (administered alone or in combination); in some instances, an anticholinergic drug was also administered. General anesthesia was induced by IV administration of thiopental sodium or propofol and was maintained by inhalation of isoflurane in oxygen after endotracheal intubation. The target rate for IV administration of crystalloid fluid for all patients was 10 mL/kg/h (4.5 mL/lb/h) during the first hour of general anesthesia and 5 mL/kg/h (2.3 mL/lb/h) for the remainder of the anesthetic period. Fluids were administered by means of gravity flow via a standard IV fluid administration set. Any fluid boluses administered were included in the total volume of fluid administered, and total anesthetic time was recorded for each patient. Blood pressure monitoring was conducted in all patients during anesthesia. The age and weight of each patient and the procedure performed were also recorded. Patients that received vasopressor drugs or synthetic colloid fluids and patients with an estimated blood loss > 5% of total estimated blood volume during anesthesia were excluded from the study.

Statistical analysis—Mean and SD were determined for the following variables: age; weight; IV fluid dose (mL/kg); duration of anesthesia; COP (before and after anesthesia); PCV (before and after anesthesia); TS concentration (before and after anesthesia); and, when preanesthetic and postanesthetic values were compared, the change in COP, PCV, and TS concentration. The Kolmogorov-Smirnov test was used to determine whether values for COP, PCV, and TS concentration were normally distributed. A paired t test was used to determine whether preanesthetic and postanesthetic values for COP and PCV were significantly different, and the Wilcoxon signed rank test was used to determine whether preanesthetic and postanesthetic values for TS concentration were significantly different. Linear regression analysis was used to determine whether anesthesia time, total fluid volume administered IV, or change in TS concentration was associated with the change in COP. Similarly, multiple linear regression analysis was used to determine whether any combination of these 3 variables was associated with the change in COP. A value of P < 0.05 was considered significant. For R², a value of 1.00 was considered perfect correlation between variables; 0.64 was considered a strong correlation; 0.25 was a moderate correlation; 0.04 was a weak correlation; and 0.00 was no correlation. All analyses were performed with statistical software.^d,e

Results

A total of 50 dogs were included in the study. Mean ± SD age was 4.4 ± 3.5 years (range, 0.3 to 12.3 years), and mean ± SD weight was 24.1 ± 14.0 kg (53.0 ± 31.0 lb). The procedures were as follows: elective ovariohysterectomy (n = 9, including 1 dog that underwent concurrent repair of a prolapsed nictitating membrane gland), elective orchiectomy (1), orthopedic procedures (stifle joint surgery, 11; arthroscopy, 3; magnetic resonance imaging followed by arthroscopy, 2; triple pelvic osteotomy, 1; and implant removal, 2), diagnostic imaging (computed tomography, 2; magnetic resonance imaging, 1; and radiography, 1), ophthalmic procedures (enucleation, 2; removal of an anteriorly luxated lens, 1; and repair of a prolapsed nictitating membrane gland, 1), dental procedures (prophylactic cleaning, 9; and root canal, 1), soft tissue surgery (mass removal, 1; and total ear canal ablation, 1), and intradermal skin testing with deep ear lavage (1). Mean ± SD duration of anesthesia was 130.4 ± 55.8 minutes, and mean ± SD total crystalloid fluid volume administered IV was 19.5 ± 10.7 mL/kg (8.9 ± 4.7 mL/lb). Mean calculated crystalloid fluid administration rate was 9.4 ± 4.6 mL/kg/h (4.3 ± 2.0 mL/lb/h). All patients maintained a mean arterial blood pressure (measured by means of an oscillometric technique) ≥ 60 mm Hg throughout anesthesia, and none of the patients had hypertension (defined as a systolic blood pressure > 180 mm Hg). None of the patients received inotropic drugs, vasopressor drugs, or synthetic colloids, and none had an estimated blood loss > 5% of total blood volume.

Mean ± SD preanesthetic values for COP (24.4 ± 4.2 mm Hg), PCV (48.8 ± 5.9%), and TS concentration (7.2 ± 0.7 g/dL) were significantly (P < 0.001) higher than postanesthetic values (19.4 ± 3.6 mm Hg, 41.8 ± 5.4%, and 6.3 ± 0.8 g/dL, respectively). Linear regression analysis revealed significant associations between the change in COP and anesthetic time (P = 0.043; R² = 0.08), total crystalloid fluid volume administered IV (P = 0.007; R² = 0.14), and the change in TS concentration (P < 0.001; R² = 0.47). However, the coefficient of determination was not high enough to allow the change in COP to be reliably predicted on the basis of any of these individual variables. Multiple linear regression analysis revealed an association between the change in COP, body weight, and preanesthetic COP (P = 0.012; R² = 0.17), but again the coefficient of determination was not high enough to allow the change in COP to be reliably predicted.

Discussion

Results of the present study suggested that in healthy dogs undergoing general anesthesia for elective procedures, COP can be expected to decrease by 5 mm Hg on average, but that the change in COP cannot be reliably predicted on the basis of the volume of crystalloid fluids administered IV during anesthesia, or by the concurrent measured decrease in TS concentration. Preanesthetic COP (mean ± SD, 24.4 ± 4.2 mm Hg) in the present study was higher than the previously published reference range (19.95 ± 2.10 mm Hg).³ This may be explained in part by hemoconcentration among dogs in the present study, as suggested by the high preanesthetic TS concentration (mean ± SD, 7.2 ± 0.7 mg/dL). The time that food and water were withheld prior to surgery was not standardized for dogs in the present study, and COP was measured between 2 and 24 hours prior to anesthetic induction. Differences in equipment used to measure COP may also help to explain the difference in preanesthetic COP values for the present and the previously reported reference range.³ The colloid osmometer used in the present study was calibrated daily throughout the study; however, because of individual variability in the osmometer membrane, it may yield slightly higher values than osmometers used to derive previously published reference ranges.³ One laboratory that used the same osmometer model used in our study reported⁴ a reference range for COP in dogs of 21 to 25 mm Hg; the mean preanesthetic COP in the present study was in this range.

The magnitude of the COP decrease detected in the present study was similar to results of recent studies of dogs¹⁰ and horses.^8,9 In a study¹⁰ of sexually intact female dogs 4 months to 4 years of age that underwent general anesthesia for elective ovariohysterectomy, there was a decrease in the COP of approximately 2 mm Hg in dogs that did not receive fluids IV during anesthesia and a decrease of approximately 3.5 mm Hg in dogs that did receive fluids IV (10 mL/kg/h) during anesthesia. In a study⁸ of healthy horses undergoing elective anesthetic procedures, there was a decrease of approximately 7 mm Hg when preanesthetic and postanesthetic values for COP were compared. Similarly, when a group of horses having surgery for colic were evaluated,⁹ a similar decrease of approximately 7 mm Hg in mean COP was detected. A possible explanation for the discrepancy in mean COP decrease in the present study (5 mm Hg) versus the previous study¹⁰ evaluating COP in dogs (3.5 mm Hg) may be that we evaluated a larger number of dogs in the present study (50 dogs vs 5 dogs).

Previous studies conducted to evaluate the effects of anesthesia on COP in dogs¹⁰ and horses^8,9 compared TS concentration to COP at multiple time points, rather than comparing the magnitude of the change from before to after anesthesia. In the previous study¹⁰ of dogs, preanesthetic values for COP and TS concentration were not significantly correlated (r = 0.30; P = 0.40), but postanesthetic values for COP and TS concentration were significantly correlated (r = 0.81; P = 0.004).¹⁰ In the equine studies,^8,9 there was a significant relationship between COP and plasma TS concentration throughout the anesthetic period. In horses having elective surgery,⁸ the R² was 0.70 (P < 0.05), and for horses having colic surgery,⁹ the R² was 0.78 (P < 0.05).

There were several limitations to the present study. The first is that we did not evaluate dogs that were ill at the time of anesthesia; all of the dogs in the present study were healthy and undergoing elective procedures. A similar study involving ill patients would be important in expanding the current knowledge of the effects of anesthesia on COP in dogs. Additionally, the present study did not address the duration that COP was decreased after the end of anesthesia, and collection of data at multiple times after extubation would have been useful. In addition, we did not standardize the anesthetic protocol, choosing instead to more closely simulate a typical clinic situation. However, the anesthetic protocol used in the present study had minimal variation and was considered typical for healthy dogs. It is possible that alternative anesthetic protocols may affect COP in different ways, and future studies are suggested to evaluate this.