Evaluation of Growing Turkey Blood Biochemistry Panel Measured Using the VetScan VS2

Abstract

Point-of-care analyzers such as the Vetscan VS2 (VS2) that provide fast results are becoming an important tool in a variety of clinical and research settings. Several studies have explored the use of the VS2 to evaluate blood chemistry values in chickens; however, the VS2 analyzer has not yet been tested in turkeys. The first objective of this study was to utilize the VS2 to evaluate possible time-dependent changes in biochemical blood analytes over time, such as glucose, calcium, or sodium. Blood samples from 30 turkeys were analyzed with the VS2 at three time points (0, 15–20, and 40–60 min post collection), and it was found that the blood biochemistry values were stable when stored at 4 °C and analyzed within 60 min of collection. Next, we assessed the agreement between the VS2 portable analyzer and the reference benchtop analyzer (Cobas c501) and found that all of the blood analytes were comparable between both analyzers. Finally, we determined reference intervals using 120 blood samples collected from turkeys in the brooding period (0–7 days) and 118 samples collected from turkeys in the growing period (1–12 weeks). Although the values were compared to the published data, the understanding of any observed differences was challenging because reference intervals for the biochemical values in turkeys are limited, and the origin of these interval values could not be traced. Significant differences in the whole blood analytes were noted between brooding and growing birds. This study offers reassurance of the validity of the use of portable analyzers in the field for turkey veterinarians. We conclude that the VS2 can be used to measure biochemical analytes in turkeys and provide the first reference intervals for brooding and growing turkeys using this instrument.

1. Introduction

Blood biochemistry analysis is commonly used in human and small animal practices to aid in the diagnosis of disease [1]. To date, this technology has been less common in poultry medicine because traditional blood chemistry analysis relies on off-site laboratory benchtop analyzers, which add cost and time to obtaining a diagnosis. Portable analyzers are becoming increasingly popular in veterinary practices because of several advantages—most notably decreased cost—compared to traditional benchtop machines [1]. The Vetscan VS2 (VS2) portable blood analyzer offers 12 blood chemistry profiles. The Avian/Reptilian Profile Plus is the only rotor designed for the avian patient, and it has been used in chickens [2,3]. On the other hand, this device has not been validated in turkeys. More broadly, the reference intervals for blood electrolytes and enzymes are virtually non-existent in turkeys. Published reference values for the domestic turkey (Meleagridis gallopavo) can be found in the Exotic Animal Formulary [4] and the Exotic Companion Medicine Handbook for Veterinarians [5], but the origin of these interval values could not be traced. These limitations hinder the interpretation of these analytes in sick or apparently healthy turkeys.

Blood biochemistry analysis can provide useful information for poultry veterinarians and nutritionists by detecting metabolic conditions often missed using traditional methods of diagnosis such as necropsy. Examples of these metabolic conditions include calcium deficiency, cardiomyopathies, heat stress, and starve outs. Since blood biochemical analysis can be performed antemortem on live animals and is not a terminal procedure, veterinarians and nutritionists can obtain important diagnostic information without decreasing the total number of a producer’s birds.

Our group previously showed that, in chickens, the VS2 results for most blood analytes were comparable to the conventional benchtop used in the study [6]. Yet, some blood analytes measured by the VS2 analyzer in chickens were significantly impacted by the time between the collection and testing. These results prompted us to investigate the suitability of the VS2 portable analyzer for use in turkeys. Thus, the objectives of this study were to determine the effect of time on turkey blood chemistry analytes, compare the blood analyte parameters by the VS2 and the Cobas c501 benchtop laboratory analyzer, and generate reference intervals for blood chemistry parameters in growing turkeys.

2. Methods

2.1. Effect of Time and Comparison of the Results Obtained by the VS2 to Those of the Cobas c501

At the day of hatch, 2000 hen poults were placed in the Poultry Barn of the Teaching Animal Unit (TAU) at the College of Veterinary Medicine, North Carolina State University (CVM-NCSU). This barn has a similar style to a commercial turkey house with side-curtain ventilation and measures 45 by 12 m² (150 by 40 sqft). A standard corn-soybean meal diet, provided by the integrator, was used for the duration of the experiment. The animal handling practices utilized in this experiment were approved by the North Carolina State University Institutional Animal Care and Use Committee (IACUC #19-001).

Ten turkeys were sampled on three different days for a total of thirty birds. The birds were marked to ensure that no bird was tested more than once. Between 1 to 2 mL of blood was collected from the brachial vein with 21-gauge needles pre-flushed with lithium-heparin (Thermo-Fisher Scientific, Fair Lawn, NJ, USA) to prevent clotting. To prevent hemolysis, the needle was removed after collection, and the blood was transferred into a 2 mL lithium heparin tube (BD Biosciences, Franklin Lakes, NJ, USA). The heparin tubes were inverted 5–7 times to ensure mixing and were then stored at 4 °C between analyses.

The workflow for this experiment is similar to that described by Ruiz-Jimenez et al. [6]. Briefly, each blood sample was tested at three different time points. All of the samples were analyzed within 60 min of collection. The samples were analyzed in the poultry house immediately following blood collection (T1) and subsequently at 15–20 min post-collection (T2). Then, the samples were moved to the clinical pathology laboratory for testing at 40–50 min post collection (T3), and the blood was simultaneously analyzed with the VS2 portable analyzer (Zoetis, Parsippany, NJ, USA) and the Cobas c501 benchtop analyzer (Roche, Basel, Switzerland). The T3 results were used to determine the agreement between the VS2 and Cobas c501 benchtop analyzers.

The following whole blood analytes were measured with the VS2 Avian/Reptile Profile Plus cartridge: Aspartate Aminotransferase (AST, U/L), Bile Acids (BA,μmol/L), Creatine Kinase (CK, U/L), Uric Acid (UA, mg/dL), Glucose (Glu, mg/dL), Total Calcium (tCa, mg/dL), Phosphorus (P, mg/dL), Potassium (K, mmol/L), Sodium (Na, mmol/L), Total Protein (TP, g/dL), Albumin (Alb, g/dL), and Globulin (Glob, g/dL). The Cobas 501 examined the same blood analytes as the VS2—with the exception of Glob, which was analyzed by the VS2 but not by the Cobas c501.

2.2. Generation of Blood Biochemistry Reference Intervals

Two hundred and thirty-eight blood samples were collected over a three-year period from five independent, clinically healthy Hybrid Converter turkey flocks. Two flocks were raised at the TAU; the other three were raised on three different commercial farms from North Carolina. All of the flocks received a standard corn-soybean meal diet. Blood was collected up to 10 different days in each flock (average: 3 different days). The ages ranged from 1 day to 12.5 weeks, and the results were categorized by age. On each day of collection, a minimum of 10 samples were collected from each flock. The blood collection and handling was performed as described above, except in turkeys less than 2 weeks of age, for which venipuncture was performed with a 25- to 23-gauge needle from the right jugular vein. All of the blood samples were analyzed within 1 h of collection. The tCa-P product and the Alb/Glob ratio were calculated from the VS2 results.

2.3. Statistical Analysis

Statistical analyses were conducted using MedCalc v 19.3 (version 19.1.5, Ostend, Belgium; https://www.medcalc.org; accessed on 15 July 2020). For each analyte, descriptive statistics including the mean, median, standard deviation, maximum value, and minimum value were generated. For all the statistical analyses, an alpha level of ≤0.05 was used to determine the statistical significance. Friedman’s test was used to determine the correlation of the VS2 analyzer results over time. Two tests were used to examine the comparison between the VS2 and Cobas c501 analyzers. Pearson’s correlation coefficient was used to examine the linear correlation between the two machines, while a Bland–Altman test [7] was used to quantify the agreement between the machines. The Passing–Bablok regression analysis was used to detect constant and/or proportional bias between the analytical methods. The normality was determined for all datasets using the D’Agostino–Pearson normality test (D’agostino, Belanger, & D’Agostino Jr, 1990). The reference intervals were determined by age-group (brooder vs. growing) using the Reference Value Advisor v2.1 free software [8]. Tukey’s test was used to detect outlier values in the dataset for each analyte.

3. Results

3.1. Effect of Time

The descriptive statistics are summarized in

Table 1. Descriptive statistics and difference in the results over time for the biochemical analytes tested with the VetScan2 analyzer.

Analyte	Time 1 (T1)			Time 2 (T2)			Time 3 (T3)			p-Value *	Differences
	Median	Mean	SD	Median	Mean	SD	Median	Mean	SD
AST	218.00	214.28	30.56	218.50	214.90	30.40	217.00	214.00	31.09	0.4055	---
BA	15.50	16.36	11.15	13.00	16.56	11.00	19.00	19.79	10.15	0.0455	T3 vs. T1 and T2
CK	998.50	1185.87	661.32	1008.00	1191.37	651.28	1010.50	1195.50	636.59	0.998	---
UA	2.6	2.59	1.25	2.70	2.64	1.23	2.65	2.63	1.22	0.2335	---
Glu	268.00	269.80	16.96	267.00	269.20	17.45	266.50	269.46	17.44	0.3742	---
tCa	11.50	11.36	0.51	11.5	11.36	0.52	11.5	11.32	0.54	0.06824	---
P	7.85	7.79	0.69	7.90	7.79	0.69	7.85	7.83	0.69	0.3285	---
K	5.00	4.96	0.68	4.90	4.89	0.64	4.75	4.86	0.63	0.14789	---
Na	146.00	146.70	2.64	146.00	146.63	2.86	146.5	146.53	2.83	0.5825	---
TP	3.00	3.09	0.31	3.00	3.10	0.31	3.00	3.10	0.29	0.75766	---
Alb	1.90	1. 18	0.25	1.90	1.91	0.25	1.85	1.90	0.25	0.4235	---
Glob	1.20	1.811	0.14	1.2	1.17	1.15	1.20	1.19	0.13	0.63467	---

* p-value < 0.05 is significant. These values are represented in italics.

. The values for most analytes were similar across all three time points. No significant differences were observed in the values obtained at T1, T2, or T3 for: AST, CK, UA, Glu, tCa, P, K, Na, TP, Alb, Glob, the calculated tCa x P product, and the Alb/Glob ratio. Although BA showed similar results at the first two time points (T1 and T2), the results for this analyte were significantly different at T3.

3.2. VS2 Portable vs. Cobas c501 Benchtop Analyzers

Descriptive statistics comparing the blood analyte values between both blood analyzers are summarized in

Table 2. Correlation, Passing–Bablok regression with constant and proportional bias, and Bland–Altman results for the agreement between the VetScan2 and the Cobas c501 analyzers.

Analytes	Correlation a	Passing–Bablok Linear Regression Analysis						Bland–Altman Plot
		y-Intercept	95% CI	Constant Bias b	Slope	95% CI	Proportional Bias c	WL d	%
AST	0.87	−7.06	−19.840 to 6.35	No	1.03	0.97 to 1.09	No	28/30	93.33
BA	0.0538	20.36	12.78 to 29.07	Yes	0.78	0.25 to 1.22	No	29/30	96.67
CK	0.9504	−28.39	−130.18 to 47.12	No	0.98	0.90 to 1.08	No	28/30	93.33
UA	0.9592	0.93	0.90 to 1.21	Yes	0.96	0.88 to 1.00	No	28/30	93.33
Glu	0.9577	0.35	−25.71 to 20.19	No	1.04	0.96 to 1.14	No	29/30	96.67
tCa	0.9203	−0.2	−1.15 to −1.05	Yes	1.00	0.89 to 1.08	No	29/30	96.67
P	0.8131	0.27	−0.89 to 1.14	No	0.91	0.80 to 1.06	No	29/30	96.67
K	0.7190	0.05	−0.95 to 1.13	No	0.83	0.61 to 1.05	No	29/30	96.67
Na	0.9212	1.50	1.50 to 50.00	Yes	1.00	0.67 to 1.00	No	29/30	96.67
TP	0.9595	−0.60	−0.97 to −0.10	Yes	1.17	1.00 to 1.28	No	29/30	96.67

^a Pearson correlation coefficient (R²). ^b Constant bias: the confidence interval for the y-intercept must include 0; otherwise, there is evidence of bias. ^c Proportional bias: the confidence interval for the slope must include 1; otherwise, there is evidence of bias. ^d WL indicates values that were within the limits of agreement based on the Bland–Altman plot. Denominators per analyte vary.

. Briefly, all analytes showed an agreement greater than 93% between the two machines when analyzed by the Bland–Altman test. Furthermore, all analytes showed a strong correlation (>70%) between the values obtained by the two machines—except for BA, which had a correlation coefficient of less than 10%. Constant bias was present in the BA, UA, tCa, Na, and TP values. The BA, Na, and TP values were constantly higher when measured by the VS2 than they were when measured by the Cobas c501 analyzer, while UA and tCa were constantly lower. Proportional bias was not detected in any of the blood analyte values.

3.3. Impact of Turkey Age on Blood Analyte Values

A total of 120 samples were analyzed from the brooder period—0 to 7 days of age. A total of 118 samples were collected from the 1- to 12-week-old turkeys in the grower period. The samples from two brooding poults and one growing poult were excluded due to quality check failures during the blood analysis. A total of 85% of the BA results fell below the dynamic range (35–200 µmol/L) of the rotor established by the vendor [9]. Box and whiskers plots were used to visualize the daily changes in blood analytes during the growth period (Figure 1). Analyte summary statistics including the mean, standard deviation, median, and number of samples analyzed were divided into brooder and grower categories and summarized in the Supplemental Table S1. The AST, CK, Glu, tCa, P, K, TP, and Alb values were significantly higher in birds during the growing period compared to those during the brooding period. In contrast, the UA and Na values were decreased in the birds during the growing period compared to those during the brooding period. No statistical differences were noted in the BA or Glob values when comparing the birds from the brooding and growing periods. Lastly, we report the calculated reference intervals for the brooding period and growing period in

Table 3. Reference intervals for healthy brooding turkeys (N = 118) between 1 and 7 days of age.

Analyte	Reference Interval	90% Confidence Interval		Units
		Lower Limit	Upper Limit
AST	128.7–253.7	117.0–135.0	239.6–276.0	U/L
BA	0.0–79.6	0.0	65.3–96.0	μmol/L
CK	132.9–1619.7	113.0–147.6	1221.7–1800.0	U/L
UA	1.20–9.31	1.10–1.59	8.91–10.0	mg/dL
Glu	190.6–357.4	172.0–207.6	330.6–545.0	mg/dL
tCa	8.90–12.41	8.60–8.99	12.20–12.70	mg/dL
P	2.50–9.20	2.20–2.89	8.81–9.40	mg/dL
K	2.60–6.41	2.30–2.80	6.21–7.80	mmol/L
Na	136.0–149.1	130.0–139.0	149.0–151.0	mmol/L
TP	2.20–3.40	2.00–2.40	3.21–3.60	g/dL
Alb	1.10–2.00	1.00–1.20	1.90–2.00	g/dL
Glob	0.70–1.82	0.70–0.90	1.70–2.60	g/dL
tCa-P product	23.63–106.28	18–92–26.55	103.54–180.56	mg2/dL2
Alb/Glob ratio	0.67–2.30	0.38–0.73	2.00–2.57

and

Table 4. Descriptive statistics and reference intervals for healthy grower turkeys (N = 117) that were 2–12 weeks old.

Analyte	Reference Interval	90% Confidence Interval		Units
		Lower Limit	Upper Limit
AST	128.7–338.2	122.0–138.8	303.2–357.0	U/L
BA	0.0–38.1	0.0	35.0–47.0	µmol/L
CK	247.2–5018.0	177.0–321.9	3548.3–7191.0	U/L
UA	0.60–5.86	0.30–1.10	5.41–8.70	mg/dL
Glu	245.9–350.2	237.0–254.0	323.8–389.0	mg/dL
tCa	10.49–12.61	10.20–10.70	12.40–12.80	mg/dL
P	3.38–9.20	2.50–6.18	8.90–9.20	mg/dL
K	2.80–6.50	2.40–3.30	6.31–7.10	mmol/L
Na	142.0–156.1	141.0–142.0	155.0–160.0	mmol/L
TP	2.60–3.91	2.60–2.70	3.80–4.40	g/dL
Alb	1.50–2.41	1.40–1.50	2.30–2.60	g/dL
Glob	0.90–2.21	0.80–0.90	1.70–2.60	g/dL
tCa-P product	36.83–106.89	26.75–67.34	103.05–114.08	mg2/dL2
Alb/Glob ratio	0.71–2.33	0.65–1.05	2.11–2.38

, respectively.

4. Discussion

We evaluated possible time-dependent changes in the biochemical blood analyses for up to 60 min only. We did not investigate changes beyond this time, as the samples do not need to be shipped to a lab for testing when using a portable analyzer. We considered 1 h to be a realistic time delay from the time of collection to the time of on-site testing All of the analytes examined by the VS2 portable analyzer showed stability up to 60 min post collection; the only exception was BA. The BA values showed a statistically significant increase over time; however, the difference at the final time point was interpreted as clinically negligible. Compared to our previous studies in chickens, the turkey blood biochemistry values appear to be more stable over time [6]. The differences in the effects of time and storage on the biochemistry analytes may reflect physiological differences between chicken and turkey blood, environmental conditions, or sample collection and handling practices. Proper storage and handling techniques are vital for timely and accurate analysis [10]. The storage of samples at 4 °C aids in the analyte stability since cooler temperatures minimize post-collection blood cell metabolism and evaporation [11]. This may be especially important for poultry veterinarians, as birds are brooded at relatively high temperatures. Higher environmental temperatures are anticipated to promote changes in blood chemistry analytes stored at “room temperature” faster than in samples stored properly under refrigeration or on ice.

All of the blood analyte values show a strong agreement and correlation between the portable VS2 analyzer and the conventional Cobas c501 benchtop analyzer. However, we found that the VS2 consistently measured the BA, Na, and TP levels as higher than the Cobas c501, while UA and tCa were consistently lower with the VS2, underscoring the need for instrument-specific reference intervals for the VS2 and Cobas c501. We detected no evidence of proportional bias in the analytes measured in this study. Our findings differ somewhat from a previous study in broiler chickens [6], where the researchers found constant and proportional biases in four out of ten analytes in chicken blood, including AST, UA, Glu, and K.

The term “metabolic profile” has been defined as the analysis of blood biochemical parameters on determined periods of life to evaluate the health of an entire herd with the goal of preventing metabolic and nutritional diseases [12]. In these instances, changes in the entire group, flock or herd are the priority over the individual animal. In this context, we evaluated the blood results in two age categories based on the different metabolisms of the birds. During their first week of life, poults rely mostly on nutrients absorbed from the yolk sac [13], which provides the young bird with a diet primarily composed of fatty acids [14]. By the second week of age, the yolk sac has been used up, and the primary source of energy is the carbohydrates. We speculate that these dietary and metabolic differences between the two age groups could influence some of the blood analytes.

Regarding the individual analytes, some interesting trends were noted. For example, both the AST and CK enzymes were significantly lower during brooding (first week of age) than during the growing period. On day 1 of age, right after hatching, the CK levels were higher and decreased thereafter (Figure 1). We suspect that the increase in the CK values at the day of hatch is caused by the heavy usage of muscles during the pipping process. The CK is kept relatively constant from 8 to 50 days of age. After that, there is a steady increase in the CK levels. Further investigation is needed to determine whether the increase was due to muscle damage, the presence of muscle myopathies, or a lack of exercise. The possible causes of increased CK values include muscle myopathies [15,16], dehydration and restricted feeding practices [17], and restrictive movement for prolonged periods of time [18,19]. Given the age-related increases in CK reflecting muscle involvement, we interpret age-related changes in AST to be due to similar mechanisms rather than liver disease, as the BA concentration did not increase over time (Figure 1). It is important to note that the BA upper limit results must be interpreted cautiously, as only 21% of the samples from the brooding period and 11% of the samples from the growing period were above the dynamic range of the rotor (35–200 µmol/L) [9]. The upper limit for the BA reference interval was 79 µmol/L during the first week, while the upper value of the BA reference interval was 38 µmol/L during the growing period. This higher BA concentration in the first week might be associated with the yolk absorption and lipidic metabolism of poultry, as BA synthesized from cholesterol in the liver [20]. During the first week, the primary source of energy for poultry is fatty acids [14]. The reference intervals calculated from the measurable samples suggest that BA are higher in the brooding birds compared to BA in the growing birds.

The concentrations of tCa and P increased from days 1–7 of age (Figure 1) and then remained relatively constant for the rest of the growth period. The average tCa in the growing meat-type chickens [21,22] are in the same range as the average tCa for the turkeys in the growing period. In the laying chickens, the calculated reference range for tCa was lower during the first week of life (tCa 8.90–12.41 mg/dL) and the growing period (10.49–12.61 mg/dL) compared to that for adults [2]. Regarding P, the average concentration was comparable to the average concentration of this analyte in the blood of broiler chickens [21]. On the other hand, the lower limit of the calculated reference interval was higher in the turkeys—for both the brooder (2.50 mg/dL) and growing (3.38 mg/dL) periods—than that reported in adult backyard hens [2]. Yet, the P values below 4 mg/dL during the growing period may be outliers and should be interpreted cautiously. We do not know what causes such a large variation in P concentration at 7 weeks of age (Figure 1). Of all the samples tested from the growing turkeys (N = 117), only three samples had a P concentration below 4 mg/dL, and those were from the same flock at 43 days of age. The tCa-P product was above 80 mg²/dL² in 25% of the samples from the turkeys during the brooding period and in more than 75% of the samples from the turkeys during the growing period. This elevation was mainly associated with a higher concentration of P. In mammals, including humans, the tCa-P product >60–80 mg²/dL², linked to higher P, is considered abnormal [23,24], as there is a tendency for calcium phosphate to precipitate in blood vessels, joints, and soft tissues. On the other hand, in our unpublished experience, the tCa-P product in chicken and turkey laying hens is well above 100 mg²/dL², which is primarily driven by tCa being higher than 20 mg/dL. Further investigation is warranted to determine whether the P concentration in young turkeys is high, whether the product of these two analytes in poultry should be interpreted differently from those in mammals, and whether the source of dietary P can cause changes in the P concentration in the blood of turkeys. A recent study in dogs demonstrated that the source of inorganic phosphate in the diet was linked to elevated serum P [25].

Based on the results from the study, we conclude that the VS2 can be used as an additional diagnostic tool in poultry medicine and management. The design and portability of the machine makes it ideal for field applications. This machine has been previously validated in chickens, and we now show that it can be used to analyze the blood from turkeys, as all of the evaluated analytes were comparable between the VS2 and the reference laboratory method. The turkey blood biochemical values were stable over a 60 min time interval. The initial reference intervals established in this paper will allow field veterinarians to start making decisions on metabolic conditions using physiologic groups and may help determine the proper course of action to follow.