Critically ill patients and those undergoing general anesthesia usually require repeated collection of blood samples to assess their clinical status and response to treatment. Those blood samples can be obtained by direct venipuncture or through an indwelling IV catheter. Repeated venipuncture can cause trauma to blood vessels and pain and anxiety for the patient. To minimize patient stress and decrease vascular trauma, blood samples are often obtained from indwelling IV catheters.1 Although collection of blood samples from indwelling IV catheters is routinely performed in human medicine,2 information regarding the effect of blood collection by an indwelling catheter versus direct venipuncture on hematologic and biochemical variables in veterinary medicine is limited. Results of routine hematologic and biochemical profiles for horses3 and coagulation profiles for dogs4–6 obtained for blood samples collected via an indwelling IV catheter after complete removal of dead space volume were comparable to those obtained for blood samples collected by direct venipuncture. In cats, statistically significant but clinically irrelevant differences in potassium, TP, and albumin concentrations were detected between blood samples collected by direct venipuncture and those obtained from a venous access port.7
A concern when collecting a blood sample from an indwelling IV catheter is the volume of blood that is discarded (presample) prior to collection of the blood sample that is analyzed. Generally, the presample should represent 300% of the catheter dead space3; therefore, in small patients, repeated collection of blood samples from an indwelling IV catheter could result in substantial iatrogenic blood loss and anemia.8,9 One method for addressing that concern is the reinfusion of the presample to the patient after the blood sample for analysis is collected. Possible adverse events associated with reinfusion of the presample include the infusion of blood clots; hemolysis, hemodilution, or contamination of the presample; and accidental submission of the presample instead of the correct blood sample to the laboratory for analysis.8,10,11 To minimize the risk for adverse events associated with reinfusion of the presample, the push-pull, or mixing, technique was devised whereby the presample blood volume (ie, a volume equal to 300% of the catheter dead space) is aspirated into a syringe and then reinfused without disconnecting the syringe from the catheter 3 times, after which the syringe is disconnected from the catheter and a second syringe is used to obtain the blood sample required for analysis. This method limits iatrogenic blood loss and the potential for contamination of the presample.8 The push-pull technique has been validated in adult human patients11,12 and children8 but, to our knowledge, has not been evaluated in veterinary patients.
Aside from evaluation of critically ill patients, serial blood sample collection is frequently necessary for anesthetized patients. Unfortunately, some anesthetics can affect vascular tone, splenic size, and fluid distribution, all of which can alter blood cell counts.13,14 In dogs, administration of ketamine, thiopental, or propofol can decrease Hct,14 which indicates that hematologic results should be interpreted with caution following administration of certain anesthetic drugs. However, further research is needed to quantify the effect of various anesthetic drugs on blood electrolyte concentrations and other biochemical variables. In North America, alfaxalone is gaining popularity as a general anesthetic for both routine and emergency surgical procedures as well as for use in critically ill patients that require mechanical ventilation. To our knowledge, the effects of alfaxalone on hematologic and biochemical variables relative to those of propofol have not been assessed.
The primary objective of the study reported here was to determine the effect of blood collection by the push-pull technique from an indwelling IV catheter versus direct venipuncture on venous blood gas variables as well as PCV and electrolyte, glucose, lactate, and TP concentrations in healthy dogs. The secondary objective of the study was to assess differences in those variables before and after propofol or alfaxalone administration.
Materials and Methods
Animals
All study procedures were reviewed and approved by the Institutional Animal Care and Use Committee of the University of Pennsylvania, and owner consent was obtained for all dogs prior to study enrollment. Thirty-four client-owned dogs that weighed ≥ 10 kg (22 lb) and were undergoing an elective surgical procedure were initially enrolled in the study. All dogs were considered healthy on the basis of results of a physical examination, CBC, and biochemical analysis. Dogs were excluded from the study if the patient struggled substantially during blood sample collection or blood samples could not be obtained because of difficulty with restraint or catheter malfunction.
Anesthetic protocol
A random number generatora was used to assign dogs to 1 of 2 anesthesia induction protocols (alfaxalone or propofol protocol). All dogs were premedicated with methadone (0.5 mg/kg [0.2 mg/lb], IM). Twenty to 30 minutes later, anesthesia was induced with either alfaxaloneb (1 to 3 mg/kg [0.5 to 1.4 mg/lb], IV to effect; alfaxalone protocol; n = 15) or propofolc (2 to 6 mg/kg [0.9 to 2.7 mg/lb], IV to effect; propofol protocol; 15). Thereafter, all dogs were maintained under general anesthesia and monitored in accordance with the standard of care for patients of the University of Pennsylvania Matthew J. Ryan Veterinary Hospital.
Blood sample collection
Paired blood samples were collected from all dogs immediately prior to premedication and within 1 minute after anesthesia induction. A random number generatora was used to determine whether the sample collected by direct venipuncture or the push-pull technique would be collected first at each sample acquisition time. For each dog prior to blood sample collection, the area over both cephalic veins was clipped and scrubbed with a 2% chlorhexidine solution. All blood samples acquired by direct venipuncture were obtained from the same cephalic vein, and all blood samples acquired by the push-pull technique were obtained from the contralateral cephalic vein. Direct venipuncture was performed with a 20-gauge needle attached to a sterile, empty (ie, the syringe did not contain any heparin) 3-mL plastic syringe and moderate compression of the forelimb proximal to the venipuncture site; approximately 1 mL of blood was obtained at each sample acquisition. For the push-pull technique, a 20-gauge, 30-mm IV catheterd was aseptically placed in the cephalic vein contralateral to the vein used for direct venipuncture. An extension set with a T-connectore was attached to the catheter, and the catheter was flushed with 1 mL of heparinized (1 U/mL) saline (0.9% NaCl) solution through the proximal port. Prior to study initiation, the dead space of the catheter and attached extension set (0.4 mL) was calculated by use of the saline solution displacement technique as described.3 A sterile, empty 3-mL plastic syringe was used to aspirate 1.2 mL of blood (ie, 300% of the calculated catheter and extension set dead space) from the catheter. The blood was immediately pushed back into the catheter, and the process was repeated 3 times. Then, a new sterile, empty 3-mL syringe was used to collect 1 mL of blood for analysis. Finally, the catheter was flushed with 1 mL of heparinized saline solution through the distal port of the T-connector.
All blood samples were immediately analyzed by use of an automated blood gas analyzerf that provided results for the following blood gas and hematologic variables: pH, Pco2, BEecf, and sodium, chloride, potassium, ionized calcium, lactate, glucose, and bicarbonate concentrations. Blood gas variables were not temperature corrected in accordance with the American Association for Respiratory Care guidelines.15 Additionally, PCV and TP concentration were determined in duplicate by microcentrifugation and refractometry, respectively.
The catheter placed in the cephalic vein was used for collection of blood samples by the push-pull technique only. Another IV catheter was aseptically placed in a saphenous vein for administration of drugs until all study blood samples were collected.
Statistical analysis
A sample size of at least 30 dogs was selected for this study on the basis of previous studies8,16 conducted to validate the push-pull technique for blood sample collection in human patients, which included 25 to 30 patients. Data were analyzed for normality by use of the Shapiro-Wilk test. Results were reported as the mean ± SD for variables with normally distributed data and the median (range) for variables with data that were not normally distributed.
For both sample acquisition times, results obtained for blood samples collected by direct venipuncture were compared with those for blood samples collected by the push-pull technique by use of the Student t test for normally distributed variables and Wilcoxon signed rank test for variables that were not normally distributed. The Bonferroni correction was used for multiple comparisons; values of P < 0.004 were considered significant for the primary objective of comparing variables between blood sample collection techniques, and values of P < 0.002 were considered significant for the secondary objective of comparing variables before and after propofol or alfaxalone administration. Bland-Altman analyses were performed when statistically significant differences were identified for a particular variable. Bias was calculated by subtraction of the value for the sample collected by direct venipuncture from the value for the sample collected by the push-pull technique. All analyses were performed with a statistical software program.g
Results
Dogs
Thirty-four dogs with body weights ranging from 10 to 65 kg (22 to 143 lb; mean, 27 kg [59 lb]) were initially enrolled in the study. Three dogs were subsequently excluded from the study because of excessive struggling during catheter placement, and 1 dog was excluded because of blood sample loss secondary to malfunction of the blood gas analyzer. Catheters were successfully placed, and blood samples were obtained and analyzed without complications from the remaining 30 dogs. The sex distribution and mean age and body weight did not differ significantly between dogs assigned to the alfaxalone and propofol protocols. Anesthesia was induced without complications, and recovery from anesthesia was uneventful for all dogs.
Venous blood gas results
For blood samples obtained prior to premedication, mean Pco2, BEecf, and TP, potassium, ionized calcium, and bicarbonate concentrations were significantly (P < 0.004) higher, whereas the mean chloride concentration was significantly lower for samples collected by direct venipuncture, compared with samples collected by the push-pull technique (Table 1). The PCV, pH, and sodium, glucose, and lactate concentrations did not differ significantly between samples collected by direct venipuncture and the push-pull technique before premedication.
Venous blood gas, PCV, and TP and electrolyte concentration results as determined for blood samples obtained by direct venipuncture or the push-pull technique for 30 healthy client-owned dogs immediately before premedication and within 1 minute after anesthesia induction.
| Blood collection technique | Bland-Altman analysis | ||||
|---|---|---|---|---|---|
| Sample acquisition time | Variable | Direct venipuncture | Push-pull | Mean bias | Percentage of paired blood samples within the 95% limits of agreement for the mean |
| Before premedication | PCV (%) | 47 ± 6 | 47 ± 6 | −0.767 | 90.0 |
| TP (g/dL) | 6.5 ± 0.5* | 6.3 ± 0.6 | −0.177 | 93.3 | |
| pH | 7.392 ± 0.032 | 7.398 ± 0.029 | 0.006 | 93.3 | |
| Pco2 (mm Hg) | 36.4 ± 4.5* | 34.5 ± 4.6 | −1.963 | 96.7 | |
| Sodium (mmol/L) | 147.0 ± 2.14 | 146.9 ± 1.86 | −0.183 | 96.7 | |
| Potassium (mmol/L) | 4.22 ± 0.26* | 4.10 ± 0.27 | −0.119 | 96.7 | |
| Chloride (mmol/L) | 115.1 ± 2.1* | 115.9 ± 2.0 | 0.700 | 93.3 | |
| Ionized calcium (mmol/L) | 1.37 ± 0.03* | 1.34 ± 0.03 | −0.035 | 93.3 | |
| Glucose (mg/dL) | 96 ± 7 | 93 ± 7 | −2.833 | 93.3 | |
| Lactate (mmol/L) | 1.4 (0.7–3.8) | 1.4 (0.7–3.8) | −0.060 | 96.7 | |
| Bicarbonate (mmol/L) | 22.4 ± 2.2* | 21.2 ± 2.1 | −1.19 | 100 | |
| BEecf (mmol/L) | −2.7 ± 2.2* | −3.8 ± 2.2 | −1.127 | 100 | |
| After anesthesia induction | PCV (%) | 44 ± 6 | 44 ± 5 | 0.067 | 93.3 |
| TP (g/dL) | 6.25 ± 0.6* | 6.16 ± 0.5 | −0.09 | 96.7 | |
| pH | 7.323 ± 0.047 | 7.321 ± 0.042 | −0.002 | 93.3 | |
| Pco2 (mm Hg) | 44.8 ± 7.7 | 44.5 ± 6.7 | −0.333 | 96.7 | |
| Sodium (mmol/L) | 147.97 ± 1.93 | 148.04 ± 1.997 | 0.070 | 96.7 | |
| Potassium (mmol/L) | 3.86 ± 0.26* | 3.79 ± 0.27 | −0.067 | 93.3 | |
| Chloride (mmol/L) | 115.9 ± 2.2* | 116.2 ± 2.2 | 0.270 | 93.3 | |
| Ionized calcium (mmol/L) | 1.39 ± 0.03 | 1.38 ± 0.03 | −0.007 | 96.7 | |
| Glucose (mg/dL) | 101 ± 11 | 101 ± 10 | −0.267 | 96.7 | |
| Lactate (mmol/L) | 1.1 (0.7–3.0) | 1.2 (0.7–2.9) | −0.037 | 93.3 | |
| Bicarbonate (mmol/L) | 23.3 ± 2.3 | 23.0 ± 2.2 | −0.283 | 93.3 | |
| BEecf (mmol/L) | −2.9 ± 2.1 | −3.3 ± 2.1 | −0.213 | 96.7 | |
Values represent the mean ± SD or median (range) unless otherwise indicated. All dogs weighed ≥ 10 kg (22 lb); were considered healthy on the basis of results of a physical examination, CBC, and biochemical analysis; and were anesthetized for elective surgical procedures. All dogs were premedicated with methadone (0.5 mg/kg [0.2 mg/lb], IM), and 20 to 30 minutes later, anesthesia was induced with either alfaxalone (1 to 3 mg/kg [0.5 to 1.4 mg/lb], IV to effect; alfaxalone protocol; n = 15) or propofol (2 to 6 mg/kg [0.9 to 2.7 mg/lb], IV to effect; propofol protocol; 15). Bias was calculated by subtraction of the value for the sample collected by direct venipuncture from the value for the sample collected by the push-pull technique.
Value differs significantly (P < 0.004) from the corresponding value for the push-pull technique.
For blood samples obtained after anesthesia induction, the mean TP and potassium concentrations were significantly increased, whereas the mean chloride concentration was significantly decreased for samples collected by direct venipuncture, compared with samples collected by the push-pull technique (Table 1). None of the other variables evaluated differed significantly between samples collected by direct venipuncture and the push-pull technique after anesthesia induction. Bland-Altman analysis of PCV, TP concentration, Pco2, and potassium concentration for blood samples collected by direct venipuncture, compared with those collected by the push-pull technique, revealed that there was a high level of agreement between the 2 sample collection methods, with 90% to 100% of all paired samples falling within the 95% limits of agreement (Figure 1).
Bland-Altman plots for comparison of PCV (A and B), TP concentration (C and D), Pco2 (E and F), and potassium concentration (G and H) between blood samples collected by direct venipuncture and those collected by the push-pull technique from 30 healthy client-owned dogs immediately before premedication (A, C, E, and G) and within 1 minute after anesthesia induction (B, D, F, and H). All dogs weighed ≥ 10 kg (22 lb); were considered healthy on the basis of results of a physical examination, CBC, and biochemical analysis; and were anesthetized for elective surgical procedures. All dogs were premedicated with methadone (0.5 mg/kg [0.2 mg/lb], IM), and 20 to 30 minutes later, anesthesia was induced with either alfaxalone (1 to 3 mg/kg [0.5 to 1.4 mg/lb], IV to effect; n = 15) or propofol (2 to 6 mg/kg [0.9 to 2.7 mg/lb], IV to effect; 15). Within each plot, each dot represents the results for 1 dog; the thick, solid, black horizontal line represents the mean bias; and the dashed horizontal lines represent the 95% limits of agreement for the mean bias. K = Potassium. PP = Push-pull sample. V = Venipuncture sample.
Citation: Journal of the American Veterinary Medical Association 251, 10; 10.2460/javma.251.10.1166
Bland-Altman plots for comparison of PCV (A and B), TP concentration (C and D), Pco2 (E and F), and potassium concentration (G and H) between blood samples collected by direct venipuncture and those collected by the push-pull technique from 30 healthy client-owned dogs immediately before premedication (A, C, E, and G) and within 1 minute after anesthesia induction (B, D, F, and H). All dogs weighed ≥ 10 kg (22 lb); were considered healthy on the basis of results of a physical examination, CBC, and biochemical analysis; and were anesthetized for elective surgical procedures. All dogs were premedicated with methadone (0.5 mg/kg [0.2 mg/lb], IM), and 20 to 30 minutes later, anesthesia was induced with either alfaxalone (1 to 3 mg/kg [0.5 to 1.4 mg/lb], IV to effect; n = 15) or propofol (2 to 6 mg/kg [0.9 to 2.7 mg/lb], IV to effect; 15). Within each plot, each dot represents the results for 1 dog; the thick, solid, black horizontal line represents the mean bias; and the dashed horizontal lines represent the 95% limits of agreement for the mean bias. K = Potassium. PP = Push-pull sample. V = Venipuncture sample.
Citation: Journal of the American Veterinary Medical Association 251, 10; 10.2460/javma.251.10.1166
Bland-Altman plots for comparison of PCV (A and B), TP concentration (C and D), Pco2 (E and F), and potassium concentration (G and H) between blood samples collected by direct venipuncture and those collected by the push-pull technique from 30 healthy client-owned dogs immediately before premedication (A, C, E, and G) and within 1 minute after anesthesia induction (B, D, F, and H). All dogs weighed ≥ 10 kg (22 lb); were considered healthy on the basis of results of a physical examination, CBC, and biochemical analysis; and were anesthetized for elective surgical procedures. All dogs were premedicated with methadone (0.5 mg/kg [0.2 mg/lb], IM), and 20 to 30 minutes later, anesthesia was induced with either alfaxalone (1 to 3 mg/kg [0.5 to 1.4 mg/lb], IV to effect; n = 15) or propofol (2 to 6 mg/kg [0.9 to 2.7 mg/lb], IV to effect; 15). Within each plot, each dot represents the results for 1 dog; the thick, solid, black horizontal line represents the mean bias; and the dashed horizontal lines represent the 95% limits of agreement for the mean bias. K = Potassium. PP = Push-pull sample. V = Venipuncture sample.
Citation: Journal of the American Veterinary Medical Association 251, 10; 10.2460/javma.251.10.1166
For dogs that received the alfaxalone protocol, the mean blood pH and potassium concentration were significantly (P < 0.002) decreased following anesthesia induction, compared with those prior to premedication, for samples collected by direct venipuncture (Table 2). When samples were collected by the push-pull technique, the mean blood pH and potassium concentration were significantly decreased, whereas the mean Pco2 and ionized calcium concentration were significantly increased, following anesthesia induction, compared with those prior to premedication.
Venous blood gas, PCV, and TP and electrolyte concentration results for the dogs of Table 1.
| Alfaxalone protocol | Propofol protocol | ||||
|---|---|---|---|---|---|
| Sample collection method | Variable | Before premedication | After anesthesia induction | Before premedication | After anesthesia induction |
| Direct venipuncture | PCV (%) | 49 ± 5 | 46 ± 6 | 46 ± 6 | 43 ± 5 |
| TP (g/dL) | 6.7 ± 0.5 | 6.4 ± 0.6 | 6.3 ± 0.5 | 6.1 ± 0.5 | |
| pH | 7.390 ± 0.038 | 7.338 ± 0.045* | 7.394 ± 0.026 | 7.309 ± 0.045* | |
| Pco2 (mm Hg) | 35.8 ± 4.7 | 41.6 ± 5.5 | 37.0 ± 4.3 | 48.1 ± 8.4* | |
| Sodium (mmol/L) | 147.7 ± 1.7 | 148.3 ± 1.2 | 146.3 ± 2.3 | 147.6 ± 2.4 | |
| Potassium (mmol/L) | 4.22 ± 0.22 | 3.85 ± 0.24* | 4.21 ± 0.30 | 3.87 ± 0.30* | |
| Chloride (mmol/L) | 115.4 ± 1.9 | 116.1 ± 1.9 | 114.8 ± 2.2 | 115.7 ± 2.5* | |
| Ionized calcium (mmol/L) | 1.38 ± 0.03 | 1.39 ± 0.03 | 1.37 ± 0.04 | 1.38 ± 0.03 | |
| Glucose (mg/dL) | 97 ± 6 | 102 ± 11 | 95 ± 8 | 100 ± 11 | |
| Lactate (mmol/L) | 1.7 ± 0.7 | 1.6 ± 0.8 | 1.5 ± 0.8 | 1.2 ± 0.4 | |
| Bicarbonate (mmol/L) | 22.1 ± 2.5 | 22.5 ± 2.3 | 22.8 ± 1.9 | 24.1 ± 2.0 | |
| BEecf (mmol/L) | −3.1 ± 2.6 | −3.5 ± 2.4 | −2.3 ± 1.7 | −2.4 ± 1.6 | |
| Push-pull | PCV (%) | 48 ± 5 | 47 ± 5 | 45 ± 6 | 42 ± 5 |
| TP (g/dL) | 6.5 ± 0.6 | 6.3 ± 0.5 | 6.1 ± 0.5 | 6.0 ± 0.5 | |
| pH | 7.394 ± 0.036 | 7.334 ± 0.047* | 7.401 ± 0.021 | 7.309 ± 0.033* | |
| Pco2 (mm Hg) | 34.2 ± 5.4 | 41.8 ± 6.3* | 34.7 ± 3.7 | 47.2 ± 6.2* | |
| Sodium (mmol/L) | 147.4 ± 1.3 | 148.6 ± 1.7 | 146.3 ± 2.3 | 147.5 ± 2.2* | |
| Potassium (mmol/L) | 4.08 ± 0.23 | 3.75 ± 0.25* | 4.11 ± 0.31 | 3.83 ± 0.30 | |
| Chloride (mmol/L) | 116.1 ± 1.8 | 116.4 ± 1.8 | 115.7 ± 2.3 | 116.0 ± 2.5 | |
| Ionized calcium (mmol/L) | 1.34 ± 0.04 | 1.38 ± 0.03* | 1.34 ± 0.03 | 1.38 ± 0.03 | |
| Glucose (mg/dL) | 93 ± 7 | 102 ± 10 | 93 ± 8 | 100 ± 9 | |
| Lactate (mmol/L) | 1.7 ± 0.8 | 1.5 ± 0.8 | 1.5 ± 0.8 | 1.2 ± 0.4 | |
| Bicarbonate (mmol/L) | 20.7 ± 2.3 | 22.3 ± 2.4 | 21.7 ± 1.8 | 24.8 ± 1.8 | |
| BEecf (mmol/L) | −4.4 ± 2.5 | −3.8 ± 2.4 | −3.3 ± 1.8 | −2.7 ± 1.6 | |
Within an anesthesia induction protocol, value differs significantly (P < 0.002) from the corresponding value before premedication.
See Table 1 for remainder of key.
For dogs that received the propofol protocol, the mean blood pH and potassium concentration were significantly decreased, whereas the mean Pco2 and chloride concentration were significantly increased, following anesthesia induction, compared with those prior to premedication, for samples collected by direct venipuncture (Table 2). When samples were collected by the push-pull technique, the mean blood pH was significantly decreased, whereas the mean Pco2 and sodium concentration were significantly increased, following anesthesia induction, compared with those prior to premedication. Additionally, for blood samples collected after anesthesia induction, none of the variables evaluated differed significantly between the alfaxalone and propofol protocols regardless of the method used for sample collection.
Discussion
Results of the present study indicated that venous blood Pco2, BEecf, and bicarbonate, chloride, ionized calcium, and potassium concentrations varied significantly between blood samples collected by direct venipuncture and those collected by the push-pull technique. However, those differences were not clinically relevant because all values (regardless of blood collection method) were within the respective reference ranges established by the clinical pathology laboratory that processed the samples, and the Bland-Altman plots revealed a high level of agreement between venous blood gas variables determined for blood samples collected by both methods. Additionally, the magnitudes of the biases calculated for the respective variables during the Bland-Altman analyses were fairly small and were within the maximum allowable analytic error for clinical laboratory tests established by the CLIA17 and American Society of Veterinary Clinical Pathologists.18 Thus, the push-pull technique appeared to be an acceptable method for collection of blood samples from dogs for venous blood gas analysis.
The results of the present study were consistent with those of studies12,19 involving human patients in whom there were some significant but clinically irrelevant differences in hematologic and biochemical variables detected between paired blood samples that were collected by the push-pull technique and a discard method. In horses, hematologic and biochemical variables did not differ significantly between blood samples collected by direct venipuncture and those collected from an IV catheter.3 However, in cats, statistically significant but clinically irrelevant differences in serum potassium, TP, and albumin concentrations were detected between blood samples collected by direct venipuncture and those collected from an IV catheter, and those differences were attributed to small amounts of hemolysis.7 Hemolysis can cause an artificial increase in blood or serum potassium concentration. In the present study, the potassium concentration in blood samples collected by direct venipuncture was significantly greater than concentration in those collected by the push-pull technique. Therefore, it is unlikely that the significant differences detected in this study were the result of hemolysis caused by the push-pull technique, but there may have been a small amount of hemolysis in blood samples collected by direct venipuncture owing to excessive negative pressure generated while obtaining samples from moving patients prior to premedication.
In another study20 involving dogs, dilution of blood samples with sodium heparin decreased Pco2 and ionized calcium, potassium, and lactate concentrations and increased sodium and chloride concentrations. In the present study, although the IV catheter was flushed with 1 mL of heparinized saline solution after it was initially placed in a cephalic vein and after each blood sample collection, which could have resulted in heparin contamination of the blood samples collected by the push-pull technique, the blood samples were otherwise not diluted with sodium heparin. Heparin contamination of the blood samples collected by the push-pull technique could have accounted for the significant differences in Pco2, BEecf, and chloride, ionized calcium, and potassium concentrations detected between the 2 sample collection methods before premedication, but if that was the cause of the differences, the expected changes in sodium and lactate concentrations were not observed. Moreover, after anesthesia induction, the TP and potassium concentrations were significantly increased and the chloride concentration was significantly decreased in blood samples collected by direct venipuncture, compared with those in blood samples collected by the push-pull technique. If the significant differences in venous blood gas variables observed during this study were the result of heparin contamination of the blood samples collected by the push-pull technique, those differences should have been detected both before and after anesthesia induction, especially the difference in ionized calcium concentration, because it is particularly sensitive to heparin dilution.20
The theoretical validity of the push-pull technique is dependent on an adequate presample blood volume being aspirated into the syringe and then mixed sufficiently to eliminate the effect of hemodilution on the blood sample collected for laboratory analysis. To our knowledge, the minimum amount of blood that must be aspirated during the presampling phase of the push-pull technique has not been quantified for veterinary patients. However, studies21,22 involving human patients have indicated that a presample blood volume of 2 to 3 times the catheter dead space is sufficient for accurate measurement of Hct and blood sodium and glucose concentrations. The push-pull technique was first described in a 1988 study,11 in which it was found to yield blood samples with comparable hematologic and biochemical results to those collected by a reinfusion method in adult human patients. In that study,11 the push-pull sequence was repeated 4 times; however, a subsequent study16 indicated that similarly acceptable results could be obtained when the push-pull sequence was repeated only 3 times. A potential concern regarding the number of times the push-pull sequence is repeated is that it may be difficult to repeatedly aspirate a sufficient amount of blood through a small catheter without causing substantial hemolysis. Consequently, we chose to use only 3 push-pull sequences in the present study. For all dogs, blood was easily aspirated through the catheter during the push-pull technique, and hemolysis was not grossly evident in any of the microhematocrit tubes after they were centrifuged.
In the present study, the pH and potassium concentration in blood samples collected by direct venipuncture following anesthesia induction were significantly lower than the pH and potassium concentration in samples collected by direct venipuncture before premedication, regardless of the induction (alfaxalone or propofol) protocol used. However, those differences were not clinically relevant, and the small decrease in blood pH observed was most likely the result of an increase in Pco2 caused by anesthesia-induced hypoventilation. Following anesthesia induction, the Pco2 was significantly greater than that prior to premedication for dogs assigned to the propofol protocol regardless of the blood collection method used, whereas for dogs assigned to the alfaxalone protocol, the Pco2 was significantly greater than that prior to premedication only for blood samples collected by the push-pull technique. Those results were consistent with findings of other studies involving dogs in which anesthesia induction with either alfaxalone23,24 or propofol25 causes an increase in Pco2. The significant, albeit clinically irrelevant, increase in potassium concentration observed in blood samples obtained by direct venipuncture before premedication, compared with that for blood samples obtained by direct venipuncture after anesthesia induction was most likely the result of mild hemolysis associated with the movement of unanesthetized patients during sample collection.2 However, in addition to an increase in potassium concentration, hemolysis would be expected to cause changes in Hct and the ionized calcium, glucose, sodium, and TP concentrations,7 none of which were observed. Moreover, the increase in potassium concentration was < 0.5 mmol/L, the maximum analytic error allowed by the CLIA, and would not alter the treatment decision for any patient.
Interestingly, none of the variables evaluated after anesthesia induction differed significantly between dogs assigned to the alfaxalone and propofol protocols. Compared with results before premedication, small but statistically significant changes in blood chloride, ionized calcium, and sodium concentrations were detected after anesthesia induction. All those changes were within the respective maximum analytic errors for diagnostic tests allowed by the CLIA and American Society of Veterinary Clinical Pathologists, and all results were within the respective reference ranges established by the clinical laboratory that performed the analyses. In dogs and cats, administration of propofol is associated with a decrease in Hct, which has been attributed to changes in erythrocyte sequestration either directly by the anesthetic or secondary to alterations in catecholamine concentrations.14,25–27 Additionally, in some studies involving dogs,14,25 propofol administration was associated with a decrease in TP concentration. In the present study, although the Hct and TP concentration following anesthesia induction for most dogs were lower than those before premedication, those decreases were not significant. This may have been because the postinduction blood samples were collected immediately after anesthesia induction, whereas in other studies,14,25 blood samples were collected at various times after anesthesia induction. Also, unlike the dogs of the present study, the dogs of those other studies14,25 were premedicated with drugs other than methadone and coadministered IV fluids, which also could have affected the magnitude of the changes in the Hct and TP concentration.
A limitation of the present study was that it was performed in patients. Thus, it could not be designed as a crossover study, which decreased the power for assessing the effect of induction agent on venous blood variables evaluated. Additionally, variation in the response of individual dogs to the induction agent administered could have contributed to the changes observed, but that effect should have been minimal because dogs were randomly allocated to each induction protocol. Premedication of dogs with methadone may also have had a small effect on the results obtained for blood samples collected after anesthesia induction. However, all dogs were premedicated with methadone, and there is no reason to believe that it would have affected dogs subsequently administered alfaxalone differently than dogs subsequently administered propofol. Because initial blood samples were obtained prior to premedication, it was not possible to collect the direct venipuncture sample simultaneously with the push-pull sample, and there may have been small restraint-induced changes in the breathing pattern and catecholamine concentrations during collection of the 2 samples. To minimize the effect those changes might have on the study results, the order in which the blood samples were collected was randomized at each sample acquisition time for each dog.
To facilitate blood sample collection, only dogs that weighed ≥ 10 kg were enrolled in the present study, and 20-gauge catheters were placed in cephalic veins. Results of other studies indicate that, in dogs, a 20-gauge catheter placed in a peripheral vein can be used to reliably collect serial 4-mL blood samples throughout a 13-hour period,1 and blood flow through a catheter during sample collection increases linearly with catheter size.28 In human patients, the extent of hemolysis in a blood sample is inversely correlated with catheter size when catheters smaller than 16 gauge are used,29,30 and the extent of hemolysis in blood samples collected via 16-gauge catheters does not differ significantly from that in blood samples collected via 20-gauge catheters.31 Additional research is necessary to determine whether the push-pull technique can be used to collect diagnostic blood samples from catheters smaller than 20 gauge and animals < 10 kg over a prolonged duration and from catheters of different types (eg, central lines or arterial catheters) placed at anatomic locations other than the antebrachium. Research is also necessary to determine whether blood samples collected by the push-pull technique can be used for accurate measurement of hematologic, biochemical, and coagulation variables other than variables evaluated in this study and evaluate the effect of IV infusion of dextrose, electrolytes, heparin, or lipid emulsion on blood samples collected by the push-pull technique. Validation of the push-pull technique for blood sample collection from animals of all sizes regardless of concurrently administered drugs, fluids, or supplements will be especially important for the management of critically ill small patients that are at high risk of iatrogenic anemia caused by serial collection of blood samples.9,32
In the present study, although minor significant differences in venous blood gas variables, including electrolyte and TP concentrations and PCV, were detected between blood samples collected via direct venipuncture and those collected by the push-pull technique as well as between blood samples that were collected before premedication and after anesthesia induction with alfaxalone or propofol, those differences were not clinically relevant. Therefore, the push-pull technique appeared to be an acceptable method for collection of blood samples from dogs for venous blood gas analysis and may be especially useful for obtaining blood samples from hospitalized dogs with patent indwelling IV catheters that would be unduly stressed by direct venipuncture.
Acknowledgments
Supported by the Frankie's Friends Resident Clinical Study Grant.
ABBREVIATIONS
| BEecf | Base excess in extracellular fluid |
| CLIA | Clinical Laboratory Improvement Amendments |
| TP | Total protein |
Footnotes
Random.org, Randomness and Integrity Services Ltd, Dublin, Ireland.
Alfaxan, Jurox Inc, Kansas City, Mo.
Diprivan, Fresenius Kabi USA LLC, Lake Zurich, Ill.
BD Insyte, Becton Dickinson Infusion Therapy Systems Inc, Sandy, Utah.
Extension set with T-Connector, Abbot Laboratories, North Chicago, Ill.
Nova Stat Profile Prime, NOVA Biomedical, Waltham, Mass.
Stata 14 for MAC, Stata Corp, College Station, Tex.
References
1. Elliott KF, Fleeman LM, Rand JS. Using 20-gauge percutaneous peripheral catheters to reliably collect serial 4-mL blood samples from conscious dogs. Aust Vet J 2010; 88: 215–221.
2. Braun JP, Bourgès Abella N, Geffré A, et al. The preanalytic phase in veterinary clinical pathology. Vet Clin Pathol 2015; 44: 8–25.
3. May ML, Nolen Walston RD, Utter ME, et al. Comparison of hematologic and biochemical results on blood obtained by jugular venipuncture as compared with intravenous catheter in adult horses. J Vet Intern Med 2010; 24: 1462–1466.
4. Millis DL, Hawkins E, Jager M, et al. Comparison of coagulation test results for blood samples obtained by means of direct venipuncture and through a jugular vein catheter in clinically normal dogs. J Am Vet Med Assoc 1995; 207: 1311–1314.
5. Maeckelbergh VA, Acierno MJ. Comparison of prothrombin time, activated partial thromboplastin time, and fibrinogen concentration in blood samples collected via an intravenous catheter versus direct venipuncture in dogs. Am J Vet Res 2008; 69: 868–873.
6. Bauer NB, Er E, Moritz A. Influence of blood collection technique on platelet function and coagulation variables in dogs. Am J Vet Res 2011; 72: 64–72.
7. Henry CJ, Russell LE, Tyler JW, et al. Comparison of hematologic and biochemical values for blood samples obtained via jugular venipuncture and via vascular access ports in cats. J Am Vet Med Assoc 2002; 220: 482–485.
8. Adlard K. Examining the push-pull method of blood sampling from central venous access devices. J Pediatr Oncol Nurs 2008; 25: 200–207.
9. Balakrishnan A, Drobatz KJ, Reineke EL. Development of anemia, phlebotomy practices, and blood transfusion requirements in 45 critically ill cats (2009–2011). J Vet Emerg Crit Care (San Antonio) 2016; 26: 406–411.
10. Cosca PA, Smith S, Chatfield S, et al. Reinfusion of discard blood from venous access devices. Oncol Nurs Forum 1998; 25: 1073–1076.
11. MacGeorte L, Steeves L, Steeves RH. Comparison of the mixing and reinfusion methods of drawing blood from a Hickman catheter. Oncol Nurs Forum 1988; 15: 335–338.
12. Byrne D. Comparing the push-pull versus discard blood sample method from adult central vascular access devices. J Infus Nurs 2016; 39: 130–135.
13. Dhumeaux MP, Snead EC, Epp TY, et al. Effects of a standardized anesthetic protocol on hematologic variables in healthy cats. J Feline Med Surg 2012; 14: 701–705.
14. Wilson DV, Evans AT, Carpenter RA, et al. The effect of four anesthetic protocols on splenic size in dogs. Vet Anaesth Analg 2004; 31: 102–108.
15. Davis MD, Walsh BK, Sittig SE, et al. AARC clinical practice guideline: blood gas analysis and hemoximetry: 2013. Respir Care 2013; 58: 1694–1703.
16. Holmes KR. Comparison of push-pull versus discard method from central venous catheters for blood testing. J Intraven Nurs 1998; 21: 282–285.
17. Medicare, Medicaid and CLIA programs: regulations implementing the atory Improvement Amendments of 1988 (CLIA). Final rule. Fed Regist 1992; 57: 7002–7186.
18. Harr KE, Flatland B, Nabity M, et al. ASVCP guidelines: allowable total error guidelines for biochemistry. Vet Clin Pathol 2013; 42: 424–436.
19. Barton SJ, Chase T, Latham B, et al. Comparing two methods to obtain blood specimens from pediatric central venous catheters. J Pediatr Oncol Nurs 2004; 21: 320–326.
20. Hopper K, Rezende ML, Haskins SC. Assessment of the effect of dilution of blood samples with sodium heparin on blood gas, electrolyte, and lactate measurements in dogs. Am J Vet Res 2005; 66: 656–660.
21. Yucha CB, DeAngelo E. The minimum discard volume: accurate analysis of peripheral hematocrit. J Intraven Nurs 1996; 19: 141–146.
22. Carlson KK, Snyder ML, LeClair HW, et al. Obtaining reliable plasma sodium and glucose determinations from pulmonary artery catheters. Heart Lung 1990; 19: 613–619.
23. Seo JI, Han SH, Choi R, et al. Cardiopulmonary and anesthetic effects of the combination of butorphanol, midazolam and alfaxalone in Beagle dogs. Vet Anaesth Analg 2015; 42: 304–308.
24. Muir W, Lerche P, Wiese A, et al. Cardiorespiratory and anesthetic effects of clinical and supraclinical doses of alfaxalone in dogs. Vet Anaesth Analg 2008; 35: 451–462.
25. Robertson SA, Johnston S, Beemsterboer J. Cardiopulmonary, anesthetic, and postanesthetic effects of intravenous infusions of propofol in Greyhounds and non-Greyhounds. Am J Vet Res 1992; 53: 1027–1032.
26. Bley CR, Roos M, Price J, et al. Clinical assessment of repeated propofol-associated anesthesia in cats. J Am Vet Med Assoc 2007; 231: 1347–1353.
27. Costa P, Nunes N, Belmonte E, et al. Hematologic changes in propofol-anesthetized dogs with or without tramadol administration. Arq Bras Med Vet Zootec 2013; 65: 1306–1312.
28. Fulton RB Jr, Hauptman JG. In vitro and in vivo rates of fluid flow through catheters in peripheral veins of dogs. J Am Vet Med Assoc 1991; 198: 1622–1624.
29. Kennedy C, Angermuller S, King R, et al. A comparison of hemolysis rates using intravenous catheters versus venipuncture tubes for obtaining blood samples. J Emerg Nurs 1996; 22: 566–569.
30. Dugan L, Leech L, Speroni KG, et al. Factors affecting hemolysis rates in blood samples drawn from newly placed IV sites in the emergency department. J Emerg Nurs 2005; 31: 338–345.
31. Grant MS. The effect of blood drawing techniques and equipment on the hemolysis of ED laboratory blood samples. J Emerg Nurs 2003; 29: 116–121.
32. Lynch AM, Respess M, Boll AE, et al. Hospital-acquired anemia in critically ill dogs and cats: a multi-institutional study. J Vet Intern Med 2016; 30: 141–146.
