Neonatal foals are immunocompetent but immunologically naïve at birth.1,2 Immediately after birth, production of autogenous immunoglobulins commences in response to antigen exposure, but substantial immunoglobulin concentrations are not achieved until approximately 2 months of age.1,3 During this time, most foals have humoral immunity derived from passively acquired immunoglobulins that were absorbed after ingestion of colostrum during the first 24 hours after birth.1–4 Because neonatal foals that do not receive adequate amounts of colostral antibodies are at a markedly increased risk for development of an infection, serum or plasma IgG concentrations are routinely measured between 12 and 24 hours of age to determine whether foals have FPTI.2,5–7 Failure of passive transfer of immunity is often defined as complete (IgG concentration, < 400 mg/dL) or partial (IgG concentration, 400 to 800 mg/dL), and the prevalence of FPTI in neonatal foals ranges from 3% to 17%.1,5,8–15 Although the exact concentration of serum or plasma IgG necessary for adequate protection against infectious disease varies depending on each individual foal's health and environment, serum or plasma IgG concentrations ≥ 800 mg/dL are generally considered sufficient.1
Foals are most susceptible to infection during the first few days of life, particularly if adequate passive transfer of immunity was not achieved2,6,7; thus, rapid detection of FPTI is important because it is a risk factor for foal morbidity and death.6 Rapid determination of serum or plasma IgG concentrations in neonatal foals allows for timely medical intervention to minimize complications associated with FPTI. As a result, numerous quantitative and semiquantitative diagnostic tests have been developed for the assessment of passive immunity in foals.16 The single RID method is a quantitative test for measurement of serum IgG concentration, and historically, its use has been considered the gold standard. However, the RID test is expensive, it is not readily available, it is susceptible to human error during measurement of the precipitin-ring diameter, and its results are temperature dependent and not available until 18 to 24 hours after the test is initiated.1,2 Additionally, the serum IgG concentration determined via RID kits from different manufacturers can vary substantially for the same sample.16 The immunoturbidimetric assay is an alternative method for the measurement of serum or plasma immunoglobulin concentrations in which an antibody directed against a specific analyte (ie, IgG) binds to that analyte and forms a complex that causes a change in the turbidity of the sample.17 The change in turbidity reduces the amount of light transmitted through the sample; the residual light transmitted through the sample is measured via spectrophotometry and provides a measure of the analyte concentration.17 Numerous immunoturbidimetric assays, including an IgG assay, have become automated so that the test can be performed on most automated biochemical analyzers, thereby minimizing potential human errors in measurement, improving precision, and shortening the turnaround time for results, compared with results obtained via an RID method. In fact, automated immunoturbidimetric assays have been recommended as an alternative to the RID method as the gold standard for measurement of serum or plasma IgG concentration.18–20
Hypoproteinemia and hypoalbuminemia are other common clinicopathologic abnormalities in foals with diseases such as enterocolitis, internal parasitism, and proliferative enteropathy (Lawsonia intracellularis infection).21–23 Although measurement of serum or plasma total protein and albumin concentrations is readily available, the ability to provide an accurate and affordable POC measurement of total protein and albumin concentrations may expedite the development of treatment plans, allow more frequent monitoring of total protein and albumin concentrations, and be beneficial for the care of critically ill foals.
A quantitative immunoturbidimetric assay for the measurement of serum or plasma IgG, total protein, and albumin concentrations has recently become commercially available by means of a POC portable analyzer. Although determination of serum IgG concentration in horses by the use of an immunoturbidimetric method via automated biochemical analyzers has been validated,18–20 the use of an immunoturbidimetric method via a POC portable analyzer has not been investigated. In veterinary medicine, it is important to evaluate the accuracy of a diagnostic test in the population for which it will be used before it is recommended for making decisions about or monitoring clinical response to treatment. The purpose of the study reported here was to investigate the performance of a POC portable analyzer that uses an immunoturbidimetric method to measure plasma IgG, total protein, and albumin concentrations in a group of ill foals and compare those results with results obtained via standard laboratory methods.
Materials and Methods
Animals—One hundred foals < 7 days of age with various medical problems that were admitted to the Hofmann Equine Neonatal Intensive Care Unit at the University of Florida Veterinary Medical Center were enrolled into the study. All study procedures were approved by the University of Florida Institutional Animal Care and Use Committee, and owner consent was obtained for each foal prior to study enrollment.
Sample collection—From each foal at the time of hospital admission, a venous blood (5 mL) sample was collected into a blood collection tube that contained heparin. Blood samples were centrifuged for 15 minutes at 5,000 × g. The plasma was removed from each sample and stored at −80°C until analyzed.
Determination of plasma IgG, total protein, and albumin concentrations via standard laboratory methods—Each plasma sample was thawed, and plasma IgG,a total protein,a and albuminb concentrations were measured with automated biochemical analyzers. Specifically, plasma IgG concentration was determined via an immunoturbidimetric assay that used goat anti-equine IgG antiserum,12–14,c total protein concentration was determined via a biuret assay,d and albumin concentration was determined via a bromocresol green dye–binding assay.e Quality control for the automated biochemical analyzers was performed daily.
Determination of plasma IgG, total protein, and albumin concentration via a POC portable analyzer—Additionally, for each sample, plasma IgG, total protein, and albumin concentrations were measured via immunoturbidimetric spectrophotometry by use of a commercially available POC portable analyzerf in accordance with the manufacturer's instructions. Briefly, the POC analyzer was calibrated by use of the control vial while the plasma samples and reagent vials were warmed to room temperature (approx 22°C) for 20 minutes. To each reagent vial, 10 μL of plasma was added, then the vial was mixed gently and incubated at room temperature for 5 minutes. Subsequently, the reagent vial was placed in the POC analyzer and covered with a plastic shroud to inhibit ambient-light interference with the analyzer's spectrophotometric function. At analysis completion, the concentration of plasma IgG, total protein, or albumin was recorded. Quality control was performed for the POC analyzer in accordance with the manufacturer's instructions daily. For each analyte evaluated throughout the study, the reagent vials (IgG,g total protein,h and albumini) originated from the same lot number.
Statistical analysis—For the purpose of the present study, plasma IgG, total protein, and albumin concentrations determined via the standard laboratory methods were considered the gold-standard, or definitive, measurements for all samples. On the basis of plasma IgG concentration results obtained via the standard laboratory method, 2 different cutoff values were used to define FPTI (plasma IgG concentration, < 400 mg/dL and < 800 mg/dL) in the study foals, and the sensitivity, specificity, accuracy, and likelihood ratios for plasma IgG concentration results obtained via the POC analyzer for diagnosing FPTI were determined at each of those 2 cutoffs. For results obtained via both the standard laboratory method and the POC analyzer, a positive test result for FPTI was defined as a plasma IgG concentration < 400 mg/dL or < 800 mg/dL, respectively, whereas a negative test result was defined as a plasma IgG concentration ≥ 400 mg/dL or ≥ 800 mg/dL, respectively Although sensitivity and specificity are fixed characteristics of a diagnostic test, predictive values for positive and negative test results vary with the prevalence of the disease in the population. Therefore, after the sensitivity and specificity of the POC analyzer for diagnosis of FPTI were determined, the respective predictive values for positive and negative test results were calculated for prevalences of FPTI ranging from 0% to 100%.
An ROC curve was generated to determine the optimal plasma IgG concentration determined via the POC analyzer to use as the cutoff for diagnosis of FPTI in foals at each of the 2 cutoffs for the standard laboratory method used to define FPTI (plasma IgG concentration, < 400 mg/dL and < 800 mg/dL).24
The agreement between the plasma IgG, total protein, and albumin concentrations obtained via the POC analyzer and those obtained via the standard laboratory methods was evaluated via the Bland-Altman method.25 For each sample and given analyte (IgG, total protein, and albumin), the bias was calculated as the analyte concentration determined via the standard laboratory method minus the corresponding analyte concentration determined via the POC analyzer. The mean bias and 95% limits of agreement (mean ± 1.96 SD) were calculated. A positive bias value indicated that the portable analyzer underestimated the concentration of an analyte, whereas a negative bias value indicated that the portable analyzer overestimated the concentration of an analyte, compared with the concentration of that analyte determined via the standard laboratory method.
For each sample and analyte (IgG, total protein, and albumin concentrations), measurements were classified as low (IgG concentration, ≤ 400 mg/dL; total protein concentration, ≤ 4.5 g/dL; and albumin concentration, ≤ 2.5 g/dL), intermediate (IgG concentration, > 400 to < 800 mg/dL; total protein concentration, > 4.5 to < 7.0 g/dL; and albumin concentration, > 2.5 to < 3.5 g/dL), or high (IgG concentration, ≥ 800 mg/dL; total protein concentration, ≥ 7.0 g/dL; and albumin concentration, ≥ 3.5 g/dL). Then a Kruskal-Wallis 1-way ANOVA by ranks was performed to evaluate the effect of the magnitude (low, intermediate, or high) of a measurement on the bias of that measurement for each analyte. When appropriate, pairwise multiple comparisons were performed via a Dunn test. For plasma IgG concentration, bias was significantly associated with the magnitude of the measurement; therefore, the relative bias was calculated for each sample as follows: (IgG concentration determined via the standard laboratory method—IgG concentration determined via the POC analyzer)/(0.5 × [IgG concentration determined via the standard laboratory method + IgG concentration determined via the POC analyzer]) × 100. Similar to bias, a positive value for relative bias indicated that the POC analyzer underestimated the concentration of an analyte, whereas a negative value for relative bias indicated that the POC analyzer overestimated the concentration of an analyte, compared with the concentration of that analyte determined via the standard laboratory method.
The agreement between measurements obtained via the standard laboratory methods and the POC analyzer was also assessed via the Lin concordance correlation coefficient, which compares 2 techniques that measure the same variable without the inherent bias of establishing a gold standard.26,27 The Lin concordance correlation coefficient (ρc) integrates a measurement of precision and accuracy: ρc = ρ × Cb, where ρ is the Pearson correlation coefficient, which measures how far each observation deviates from the best-fit line (ie, precision), and Cb is a bias correction factor that measures how far the best-fit line deviates from a 45° line through the origin (ie, accuracy); a ρc equal to 1 indicates perfect agreement between the 2 techniques.26,27 All analyses were performed with commercial statistical software,j and values of P < 0.05 were considered significant.
Results
Plasma IgG concentration—Of the 100 plasma samples evaluated, the plasma IgG concentration determined via the standard laboratory method was < 400 mg/dL in 31, 400 to < 800 mg/dL in 23, and ≥ 800 mg/dL in 46. Thus, when FPTI was defined as a plasma IgG concentration < 400 mg/dL as determined via the standard laboratory method, the POC analyzer had a sensitivity of 80.7%, specificity of 100%, accuracy of 94%, and likelihood ratio for a negative test result of 0.19 (the likelihood ratio for a positive test result could not be determined because there were no false-positive results [ie, the denominator was 0]). When FPTI was defined as a plasma IgG concentration < 800 mg/dL as determined via the standard laboratory method, the portable analyzer had a sensitivity of 75.9%, specificity of 100%, accuracy of 87%, and likelihood ratio for a negative test result of 0.24 (the likelihood ratio for a positive test result could not be determined because there were no false-positive results [ie, denominator was 0]). The predictive values for positive (IgG concentration, < 400 mg/dL and < 800 mg/dL, respectively) and negative (IgG concentration, ≥ 400 mg/dL and ≥ 800 mg/dL, respectively) test results were determined for the entire range (0% to 100%) of FPTI prevalence and plotted (Figure 1) for the POC analyzer.
Estimated predictive values for positive (plasma IgG concentration, < 400 mg/dL [A] or < 800 mg/dL [B]; dashed line) and negative (plasma IgG concentration, ≥ 400 mg/dL [A] or ≥ 800 mg/dL [B]; solid line) test results as determined via a POC portable analyzer for the entire spectrum of FPTI prevalence in neonatal foals. For each of 100 ill foals (age, < 7 days) admitted to a neonatal intensive care unit at a veterinary teaching hospital, plasma IgG, total protein, and albumin concentrations were determined via immunoturbidimetric, biuret, and bromocresol green dye–binding assays, respectively, performed by automated biochemical analyzers (standard laboratory methods) and via immunoturbidimetric assays performed by a POC portable analyzer. Plasma IgG, total protein, and albumin concentrations determined via the standard laboratory methods were considered the gold-standard, or definitive, measurements for all samples. The sensitivity and specificity of the plasma IgG concentration determined by the POC portable analyzer for diagnosis of FPTI were calculated, and then the predictive values for positive and negative test results obtained via the POC portable analyzer were calculated for the entire spectrum of FPTI prevalence.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
Estimated predictive values for positive (plasma IgG concentration, < 400 mg/dL [A] or < 800 mg/dL [B]; dashed line) and negative (plasma IgG concentration, ≥ 400 mg/dL [A] or ≥ 800 mg/dL [B]; solid line) test results as determined via a POC portable analyzer for the entire spectrum of FPTI prevalence in neonatal foals. For each of 100 ill foals (age, < 7 days) admitted to a neonatal intensive care unit at a veterinary teaching hospital, plasma IgG, total protein, and albumin concentrations were determined via immunoturbidimetric, biuret, and bromocresol green dye–binding assays, respectively, performed by automated biochemical analyzers (standard laboratory methods) and via immunoturbidimetric assays performed by a POC portable analyzer. Plasma IgG, total protein, and albumin concentrations determined via the standard laboratory methods were considered the gold-standard, or definitive, measurements for all samples. The sensitivity and specificity of the plasma IgG concentration determined by the POC portable analyzer for diagnosis of FPTI were calculated, and then the predictive values for positive and negative test results obtained via the POC portable analyzer were calculated for the entire spectrum of FPTI prevalence.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
Estimated predictive values for positive (plasma IgG concentration, < 400 mg/dL [A] or < 800 mg/dL [B]; dashed line) and negative (plasma IgG concentration, ≥ 400 mg/dL [A] or ≥ 800 mg/dL [B]; solid line) test results as determined via a POC portable analyzer for the entire spectrum of FPTI prevalence in neonatal foals. For each of 100 ill foals (age, < 7 days) admitted to a neonatal intensive care unit at a veterinary teaching hospital, plasma IgG, total protein, and albumin concentrations were determined via immunoturbidimetric, biuret, and bromocresol green dye–binding assays, respectively, performed by automated biochemical analyzers (standard laboratory methods) and via immunoturbidimetric assays performed by a POC portable analyzer. Plasma IgG, total protein, and albumin concentrations determined via the standard laboratory methods were considered the gold-standard, or definitive, measurements for all samples. The sensitivity and specificity of the plasma IgG concentration determined by the POC portable analyzer for diagnosis of FPTI were calculated, and then the predictive values for positive and negative test results obtained via the POC portable analyzer were calculated for the entire spectrum of FPTI prevalence.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
The Kruskal-Wallis 1-way ANOVA by ranks results indicated that the bias for the plasma IgG concentration determined via the POC analyzer was significantly (P < 0.001) associated with high (≥ 800 mg/dL) measurements of plasma IgG concentrations. However, relative bias was not significantly (P = 0.380) associated with the magnitude of the plasma IgG concentration determined via the POC analyzer, and the mean relative bias was −32.1% (95% limits of agreement, −82.2% to 18.1%; Figure 2). The Lin concordance correlation coefficient was 0.86 (95% CI, 0.80 to 0.90), which suggested moderate agreement between the plasma IgG concentration measured via the standard laboratory method and that measured via the POC analyzer.
Bland-Altman plot of plasma IgG concentrations measured concurrently via a standard laboratory method and a POC portable analyzer. Relative bias was calculated as (IgG concentration determined via the standard laboratory method—IgG concentration determined via the POC analyzer)/(0.5 × [IgG concentration determined via the standard laboratory method + IgG concentration determined via the POC analyzer]) × 100. The solid line represents the mean relative bias, and the dashed lines represent the limits of agreement (mean ± 1.96 SD). See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
Bland-Altman plot of plasma IgG concentrations measured concurrently via a standard laboratory method and a POC portable analyzer. Relative bias was calculated as (IgG concentration determined via the standard laboratory method—IgG concentration determined via the POC analyzer)/(0.5 × [IgG concentration determined via the standard laboratory method + IgG concentration determined via the POC analyzer]) × 100. The solid line represents the mean relative bias, and the dashed lines represent the limits of agreement (mean ± 1.96 SD). See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
Bland-Altman plot of plasma IgG concentrations measured concurrently via a standard laboratory method and a POC portable analyzer. Relative bias was calculated as (IgG concentration determined via the standard laboratory method—IgG concentration determined via the POC analyzer)/(0.5 × [IgG concentration determined via the standard laboratory method + IgG concentration determined via the POC analyzer]) × 100. The solid line represents the mean relative bias, and the dashed lines represent the limits of agreement (mean ± 1.96 SD). See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
When FPTI was defined as a plasma IgG concentration < 400 mg/dL as determined via the standard laboratory method, the area under the ROC curve for the portable analyzer was 0.995 (95% CI, 0.954 to 1.00), which suggested that measurement of plasma IgG concentration via the POC analyzer was very accurate. The optimal cutoff value for plasma IgG concentration measured via the POC analyzer for diagnosis of FPTI was 620 mg/dL. At that cutoff, the sensitivity and specificity of the POC analyzer for diagnosis of FPTI were 96.8% (95% CI, 83.3% to 99.9%) and 94.2% (95% CI, 85.8% to 98.4%), respectively, and the likelihood ratios for positive and negative test results were 16.7 and 0.03, respectively. When FPTI was defined as a plasma IgG concentration < 800 mg/dL as determined via the standard laboratory method, the area under the ROC curve for the POC analyzer was 0.990 (95% CI, 0.946 to 1.00), which suggested that measurement of plasma IgG concentration via the POC analyzer was very accurate. The optimal cutoff value for plasma IgG concentration measured via the POC analyzer for diagnosis of FPTI was 1,150 mg/dL. At that cutoff, the sensitivity and specificity of the POC analyzer for diagnosis of FPTI were 96.3% (95% CI, 87.3% to 99.5%) and 95.7% (95% CI, 85.2% to 99.5%), respectively, and the likelihood ratios for positive and negative test results were 22.2 and 0.04, respectively.
Plasma total protein concentration—The median plasma total protein concentration as measured via the standard laboratory method was 4.9 g/dL (range, 1.1 to 8.9 g/dL). Unlike plasma IgG concentration, the magnitude of the plasma total protein concentration was not significantly (P = 0.626) associated with bias. The mean bias for plasma total protein concentration measured via the POC analyzer, compared with that measured via the standard laboratory method, was −0.33 g/dL (95% limits of agreement, −1.2 to 0.54 g/dL; Figure 3). The Lin concordance correlation coefficient was 0.92 (95% CI, 0.88 to 0.94), which suggested moderate agreement between the plasma total protein concentration measured via the standard laboratory method and that measured via the POC analyzer.
Bland-Altman plot of plasma total protein concentration measured concurrently via a standard laboratory method and a POC analyzer. Bias was calculated as the total protein concentration determined via the standard laboratory method—total protein concentration determined via the POC analyzer. The solid line represents the mean bias, and the dashed lines represent the limits of agreement (mean ± 1.96 SD). See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
Bland-Altman plot of plasma total protein concentration measured concurrently via a standard laboratory method and a POC analyzer. Bias was calculated as the total protein concentration determined via the standard laboratory method—total protein concentration determined via the POC analyzer. The solid line represents the mean bias, and the dashed lines represent the limits of agreement (mean ± 1.96 SD). See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
Bland-Altman plot of plasma total protein concentration measured concurrently via a standard laboratory method and a POC analyzer. Bias was calculated as the total protein concentration determined via the standard laboratory method—total protein concentration determined via the POC analyzer. The solid line represents the mean bias, and the dashed lines represent the limits of agreement (mean ± 1.96 SD). See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
Plasma albumin concentration—The median plasma albumin concentration as measured via the standard laboratory method was 2.9 g/dL (range, 0.5 to 4.7 g/dL). Similar to plasma total protein concentration, the magnitude of plasma albumin concentration was not significantly (P = 0.528) associated with bias. The mean bias for plasma albumin concentration measured via the POC analyzer, compared with that measured via the standard laboratory method, was −0.88 g/dL (95% limits of agreement, −1.98 to 0.21 g/dL; Figure 4). The Lin concordance correlation coefficient was 0.29 (95% CI, 0.21 to 0.37), which suggested poor agreement between the plasma albumin concentration measured via the standard laboratory method and that measured via the POC analyzer.
Bland-Altman plot of plasma albumin concentrations measured concurrently via a standard laboratory method and a POC analyzer. Bias was calculated as the albumin concentration determined via the standard laboratory method—albumin concentration determined via the POC analyzer. The solid line represents the mean bias, and the dashed lines represent the limits of agreement (mean ± 1.96 SD). See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
Bland-Altman plot of plasma albumin concentrations measured concurrently via a standard laboratory method and a POC analyzer. Bias was calculated as the albumin concentration determined via the standard laboratory method—albumin concentration determined via the POC analyzer. The solid line represents the mean bias, and the dashed lines represent the limits of agreement (mean ± 1.96 SD). See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
Bland-Altman plot of plasma albumin concentrations measured concurrently via a standard laboratory method and a POC analyzer. Bias was calculated as the albumin concentration determined via the standard laboratory method—albumin concentration determined via the POC analyzer. The solid line represents the mean bias, and the dashed lines represent the limits of agreement (mean ± 1.96 SD). See Figure 1 for remainder of key.
Citation: Journal of the American Veterinary Medical Association 242, 6; 10.2460/javma.242.6.812
Discussion
Results of the present study indicated that the use of a POC portable analyzer provided accurate measurement of plasma IgG and total protein concentrations within 5 minutes after initiation of the tests and may be beneficial for the rapid assessment of passive transfer of maternal antibodies in neonatal foals. During the critical care of ill neonatal foals, it is often necessary for rapid determination of various blood variables to identify clinicopathologic abnormalities (ie, electrolyte derangements), develop a list of differential diagnoses, and formulate a treatment plan. The increased availability of POC analyzers to veterinary clinicians has expedited and improved stall-side evaluation of ill horses and foals.28 Point-of-care analyzers that have been evaluated in horses include handheld devices that measure concentrations of glucose, lactate, cardiac troponin, and hemoglobin.29–33 Portable ultrasonography and POC blood gas and electrolyte analyzers have also become readily available, which allow immediate stall-side imaging and evaluation of blood acid-base and electrolyte status.34–37 Because foals with FPTI have a higher risk of developing an infection than do foals with adequate concentrations of maternal antibodies, rapid and accurate measurement of serum or plasma IgG concentrations is important for the management and care of neonatal foals.5–7 Treatment of FPTI in foals is time dependent; the capacity of the intestine to absorb colostral antibodies is greatest within the first 12 hours after birth and then begins to decrease.38 Failure of passive transfer of immunity is generally defined as a serum or plasma IgG concentration < 400 mg/dL measured at 18 to 24 hours of age, and the treatment for FPTI is IV administration of plasma.39 Thus, measurement of serum or plasma IgG concentration is most advantageous if results can be obtained within a few hours, if not less, after sample collection.
In the present study, when FPTI was defined as a plasma IgG concentration < 400 mg/dL as determined via the standard laboratory method, the sensitivity and likelihood ratio for a negative test result for the POC analyzer were 80% and 0.19, respectively. When FPTI was defined as a plasma IgG concentration < 800 mg/dL as determined via the standard laboratory method, the sensitivity and likelihood ratio for a negative test result for the POC analyzer were 76% and 0.24, respectively. This indicated that the likelihood of a foal having FPTI decreased approximately 4- to 5-fold given a negative test result (ie, IgG > 400 or 800 mg/dL, respectively) with the POC analyzer. Given the performance parameters of the POC analyzer for detecting FPTI in the present study and assuming an FPTI prevalence of 15% in neonatal foals,7–15 the probability that a neonatal foal with a negative test result (plasma IgG concentration, ≥ 400 mg/dL or ≥ 800 mg/dL) has FPTI (ie, false-negative result) is 3.4% and 4.5%, respectively. The specificity of the POC analyzer for detecting FPTI was 100%; no foals with adequate passive transfer of immunity were misclassified as having FPTI (ie, had a false-positive result). Thus, use of the POC analyzer for measurement of plasma IgG concentration in neonatal foals is unlikely to result in overdiagnosis of FPTI and unnecessary administration of plasma to foals with adequate passive transfer of immunity, and only a small proportion (3.4% to 4.5%) of foals with FPTI will be misdiagnosed (ie, have false-negative results).
A diagnostic method used to screen a population for any disease, including FPTI in foals, should have high sensitivity because this will ensure a high predictive value for a negative test result and allow for identification of most diseased individuals. In the present study, the sensitivity of the POC analyzer was substantially improved if the plasma IgG concentration cutoff used to define FPTI in foals was increased. When FPTI was defined as a plasma IgG concentration < 400 mg/dL as determined via the standard laboratory method, increasing the cutoff for plasma IgG concentration measured via the POC analyzer from < 400 mg/dL to < 620 mg/dL improved the sensitivity of the POC analyzer for detection of FPTI from 80% to 97%. When FPTI was defined as a plasma IgG concentration < 800 mg/dL as determined via the standard laboratory method, increasing the cutoff for plasma IgG concentration measured via the POC analyzer from < 800 mg/dL to < 1,150 mg/dL improved the sensitivity of the POC analyzer for detection of FPTI from 76% to 96%. However, in both instances, the increase in sensitivity was accompanied by a decrease in specificity.
Results of the present study suggested that measurement of plasma IgG concentration via the POC analyzer had a similar, if not slightly better, sensitivity and specificity, compared with that of other methods used to measure plasma or serum IgG concentration in foals.16,40 Semiquantitative laboratory methods used to measure serum IgG concentration in foals include zinc sulfate turbidity, latex agglutination, glutaraldehyde coagulation, and ELISA.16 Of those semiquantitative tests, the most commonly used test to detect FPTI in foals is the ELISA because it provides results within 10 minutes after test initiation and is commercially available, easy to use, and inexpensive. Although results of multiple studies40–42 indicate that the ELISA is fairly accurate, in 1 study40 this test had good sensitivity but poor specificity when used in a population of hospitalized foals. Moreover, because of the semiquantitative nature of the ELISA, the results are categorized into 1 of 3 classifications (ie, < 400 mg/dL, 400 to 800 mg/dL, or ≥ 800 mg/dL) rather than provided as the specific IgG concentration. In a study16 in which 5 methods (zinc sulfate turbidity, glutaraldehyde coagulation, a quantitative colorimetric assay, and 2 semiquantitative immunoassays) for the measurement of serum IgG concentration in neonatal foals were compared, the sensitivity of the 5 assays for identifying foals with serum IgG concentrations < 400 mg/dL ranged from 89% to 100%, and the specificity ranged from 79% to 96%; the sensitivity of the 5 assays for identifying foals with serum IgG concentrations < 800 mg/dL ranged from 52% to 98%, and the specificity ranged from 57% to 100%. Thus, the performance of the POC analyzer evaluated in the present study was roughly equivalent, and in some instances superior, to those other 5 methods for identifying neonatal foals with plasma IgG concentrations < 400 mg/dL (sensitivity, 80%; specificity, 100%) and < 800 mg/dL (sensitivity, 76%; specificity, 100%) and has the additional benefit of the rapid provision of quantitative results.
An automated immunoturbidimetric method was used as the standard laboratory method in the present study because investigators of other studies19,20 reported a high correlation between IgG concentration results obtained via this method and those obtained via an RID method. Additionally, the use of the immunoturbidimetric method yielded consistent (ie, repeatable) results when the same sample was tested multiple times during 1 day and when the same sample was tested once per day during a period of several days.19 Historically, the RID method has been used as the standard laboratory method for determination of serum IgG concentration against which other test methods are evaluated; however, results of another study16 suggest that there is substantial variation in agreement among commercially available RID assays, especially at serum IgG concentrations > 400 mg/dL. Moreover, results for the RID method are measured manually and reported in increments of 100 mg/dL, whereas the results of the immunoturbidimetric method are measured automatically and reported in increments of 1 mg/dL (ie, more precise). Other advantages of the immunoturbidimetric assay when compared with the RID assay include a shorter turnaround time (< 1 hour) for results and elimination of human error in measurement of the precipitin-ring diameter.20 In fact, investigators of 1 study20 concluded that the automated immunoturbidimetric method was as reliable as the RID assay for determination of plasma IgG concentration and was acceptable for use as a standard reference laboratory method.
In veterinary patients, measurement of serum or plasma total protein concentration is often used to estimate hydration status, monitor response to fluid therapy, and assess total protein content within the blood. Use of a handheld medical refractometer represents a readily available, low-cost, and rapid method for assessment of serum or plasma total protein concentration, and results obtained via that method are highly correlated with those obtained via standard laboratory methods.43,44 In the present study, the mean bias for measurement of plasma total protein concentration via the POC analyzer was −0.33 g/dL, which indicated that the POC analyzer overestimated total protein concentration slightly, compared with that measured via the standard laboratory method. The reference range for plasma total protein concentration in healthy neonatal foals is 4.3 to 8.1 g/dL45; thus, the magnitude of overestimation (0.33 g/dL) of plasma total protein concentration via the POC analyzer was small and likely to be clinically irrelevant. Moreover, the Lin concordance correlation coefficient (0.92) between the POC analyzer and the standard laboratory method for the measurement of plasma total protein concentration indicated good agreement, which suggested that the POC analyzer might be useful for the determination of plasma total protein concentration during the clinical evaluation of ill foals. However, the 95% limits of agreement for measurement of plasma total protein were wide (–1.2 to 0.54 g/dL), which indicated that the POC analyzer could overestimate plasma total protein concentration by as much as 1 g/dL, especially when the total protein concentration was high (ie, > 7 g/dL). Given that the serum or plasma total protein concentration can be accurately and affordably determined via a handheld refractometer, the purchase of a POC portable analyzer for the sole purpose of determination of plasma total protein concentration may not be justified.
In the present study, the POC analyzer overestimated plasma albumin concentration by a mean of 0.88 g/dL, compared with the plasma albumin concentration determined via the standard laboratory method. The reference interval for plasma albumin concentration in healthy foals is 2.5 to 3.6 g/dL45; thus, care must be taken in the interpretation of results obtained from the POC analyzer. For example, the POC analyzer could yield a plasma albumin concentration of 2.6 g/dL for a neonatal foal in which the actual plasma albumin concentration as determined via the standard laboratory method was 1.7 g/dL. This could result in the attending clinician making the assumption that the foal's plasma albumin concentration was adequate and formulating an inappropriate treatment plan. Furthermore, the 95% limits of agreement for measurement of plasma albumin concentration were wide (–1.98 to 0.21 g/dL), compared with the reference interval for plasma albumin concentration in healthy foals. Therefore, we concluded that the POC analyzer was an unacceptable method for determination of plasma albumin concentration in neonatal foals.
Results of the present study suggested that use of the POC portable analyzer is an acceptable method for rapid, accurate, and quantitative determination of plasma IgG concentration in neonatal foals during the assessment of FPTI. The POC analyzer also provided acceptable results for determination of plasma total protein concentration; however, purchase of a POC portable analyzer for the sole purpose of measuring serum or plasma total protein may be unwarranted because use of a handheld refractometer for determination of total protein concentration is easier and more cost-effective. The POC portable analyzer was not an effective method for determination of plasma albumin concentration in neonatal foals.
ABBREVIATIONS
| CI | Confidence interval |
| FPTI | Failure of passive transfer of immunity |
| POC | Point-of-care |
| RID | Radial immunodiffusion |
| ROC | Receiver operating characteristic |
Cobas Mira L, Roche Diagnostics Corp, Indianapolis, Ind.
Randox Daytona, Randox Laboratories Ltd, Antrim, Northern Ireland.
Goat anti-equine IgG antiserum, Midland BioProducts Corp, Boone, Iowa.
Total protein reagent, Sigma-Aldrich, St Louis, Mo.
Albumin (liquid), Randox Laboratories Ltd, Antrim, Northern Ireland.
MBC QTII Instrument, Midland BioProducts Corp, Boone, Iowa.
Equine IgG Kit, Midland BioProducts Corp, Boone, Iowa.
Total Protein Kit, Midland BioProducts Corp, Boone, Iowa.
Albumin Kit, Midland BioProducts Corp, Boone, Iowa.
MedCalc for Windows, version 12.3.0.0, MedCalc Software, Mariakerke, Belgium.
References
1. Sellon DC, Hines MT, Johnson JR. Immunologic disorders: equine immunodeficiency diseases. In: Smith BP, ed. Large animal internal medicine. 4th ed. St Louis: Mosby, 2009; 1667–1671.
2. Giguère S, Polkes AC. Immunologic disorders in neonatal foals. Vet Clin North Am Equine Pract 2005; 21: 241–272.
3. Holznagel DL, Hussey S, Mihalyi JE, et al. Onset of immunoglobulin production in foals. Equine Vet J 2003; 35: 620–622.
4. Sheoran AS, Timoney JF, Holmes MA, et al. Immunoglobulin isotypes in sera and nasal mucosal secretions and their neonatal transfer and distribution in horses. Am J Vet Res 2000; 61: 1099–1105.
5. Tyler-McGowan CM, Hodgson JL, Hodgson DR. Failure of passive transfer in foals: incidence and outcome on four studs in New South Wales. Aust Vet J 1997; 75: 56–59.
6. Robinson JA, Allen GK, Green EM, et al. A prospective study of septicaemia in colostrum-deprived foals. Equine Vet J 1993; 25: 214–219.
7. McGuire TC, Crawford TB, Hallowell AL, et al. Failure of colostral immunoglobulin transfer as an explanation for most infections and deaths of neonatal foals. J Am Vet Med Assoc 1977; 170: 1302–1304.
8. Morris DD, Meirs DA, Merryman GS. Passive transfer failure in horses: incidence and causative factors on a breeding farm. Am J Vet Res 1985; 46: 2294–2299.
9. Raidal SL. The incidence and consequences of failure of passive transfer of immunity on a Thoroughbred farm. Aust Vet J 1996; 73: 201–206.
10. Kohn CW, Knight D, Hueston W, et al. Colostral and serum IgG, IgA, and IgM concentrations in Standardbred mares and their foals at parturition. J Am Vet Med Assoc 1989; 195: 64–68.
11. LeBlanc MM, Tran T, Baldwin JL, et al. Factors that influence passive transfer of immunoglobulins in foals. J Am Vet Med Assoc 1992; 200: 179–183.
12. Baldwin JL, Cooper WL, Vanderwall DK, et al. Prevalence (treatment days) and severity of illness in hypogammaglobulinemic and normogammaglobulinemic foals. J Am Vet Med Assoc 1991; 198: 423–428.
13. Clabough DL, Levine JF, Grant GL, et al. Factors associated with failure of passive transfer of colostral antibodies in Standardbred foals. J Vet Intern Med 1991; 5: 335–340.
14. Haas SD, Bristol F, Card CE. Risk factors associated with the incidence of foal mortality in an extensively managed mare herd. Can Vet J 1996; 37: 91–95.
15. Perryman LE, McGuire TC. Evaluation for immune system failures in horses and ponies. J Am Vet Med Assoc 1980; 176: 1374–1377.
16. Davis R, Giguère S. Evaluation of five commercially available assays and measurement of serum total protein concentration via refractometry for the diagnosis of failure of passive transfer of immunity in foals. J Am Vet Med Assoc 2005; 227: 1640–1645.
17. Sacks DB. Analytes: carbohydrates. In: Burtis CA, Ashwood ER, eds. Tietz fundamentals of clinical chemistry. 5th ed. Philadelphia: WB Saunders Co, 2001; 427–461.
18. Bauer JE, Brooks TP. Immunoturbidimetric quantification of serum immunoglobulin G concentration in foals. Am J Vet Res 1990; 51: 1211–1214.
19. McCue PM. Evaluation of a turbidimetric immunoassay for measurement of plasma IgG concentration in foals. Am J Vet Res 2007; 68: 1005–1009.
20. Davis DG, Schaefer DMW, Hinchcliff KW, et al. Measurement of serum IgG in foals by radial immunodiffusion and automated turbidimetric immunoassay. J Vet Intern Med 2005; 19: 93–96.
21. Magdesian KG. Neonatal foal diarrhea. Vet Clin North Am Equine Pract 2005; 21: 295–312.
22. Clayton HM, Duncan JL, Dargie JD. Pathophysiological changes associated with Parascaris equorum infection in the foal. Equine Vet J 1980; 12: 23–25.
23. Frazer ML. Lawsonia intracellularis infection in horses: 2005–2007. J Vet Intern Med 2008; 22: 1243–1248.
24. Greiner M, Pfeiffer D, Smith RD. Principles and practical application of the receiver-operating characteristic analysis for diagnostic tests. Prev Vet Med 2000; 45: 23–41.
25. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307–310.
26. Lin LI-K. A concordance correlation coefficient to evaluate reproducibility. Biometrics 1989; 45: 255–268.
27. Lin LI-K. A note on the concordance correlation coefficient. Biometrics 2000; 56: 324–325.
28. Wilkins PA. The equine neonatal intensive care laboratory: point-of-care testing. Clin Lab Med 2011; 31: 125–137.
29. Hollis AR, Dallap Schaer BL, Boston RC, et al. Comparison of the Accu-Chek Aviva point-of-care glucometer with blood gas and laboratory methods of analysis of glucose measurement in equine emergency patients. J Vet Intern Med 2008; 22: 1189–1195.
30. Hackett ES, McCue PM. Evaluation of a veterinary glucometer for use in horses. J Vet Intern Med 2010; 24: 617–621.
31. Tennent-Brown BS, Wilkins PA, Lindborg S, et al. Assessment of a point-of-care lactate monitor in emergency admissions of adult horses to a referral hospital. J Vet Intern Med 2007;21: 1090–1098.
32. Kraus MS, Jesty SA, Gelzer AR, et al. Measurement of plasma cardiac troponin I concentration by use of a point-of-care analyzer in clinically normal horses and horses with experimentally induced cardiac disease. Am J Vet Res 2010; 71: 55–59.
33. Chevalier H, Posner LP, Ludders JW, et al. Accuracy and precision of a point-of-care hemoglobinometer for measuring hemoglobin concentration and estimating packed cell volume in horses. J Am Vet Med Assoc 2003; 223: 78–83.
34. Grosenbaugh DA, Gadawski JE, Muir WW. Evaluation of a portable clinical analyzer in a veterinary hospital setting. J Am Vet Med Assoc 1998; 213: 691–694.
35. Looney AL, Ludders J, Erb HN, et al. Use of a handheld device for analysis of blood electrolyte concentrations and blood gas partial pressures in dogs and horses. J Am Vet Med Assoc 1998; 213: 526–530.
36. Silverman SC, Birks EK. Evaluation of the i-STAT hand-held chemical analyzer during treadmill and endurance exercise. Equine Vet J Suppl 2002;(34): 551–554.
37. Viljoen A, Saulez MN, Carstens A, et al. The impact of ultrasound during emergency after-hour admission in horses. J S Afr Vet Assoc 2010; 81: 216–218.
38. Raidal SL, McTaggart C, Penhale J. Effect of withholding macro-molecules on the duration of intestinal permeability to colostral IgG in foals. Aust Vet J 2005; 83: 78–81.
39. Sellon DC. Secondary immunodeficiencies of horses. Vet Clin North Am Equine Pract 2000; 16: 117–130.
40. Metzger N, Hinchcliff KW, Hardy J, et al. Usefulness of a commercial equine IgG test and serum protein concentration as indicators of failure of transfer of passive immunity in hospitalized foals. J Vet Intern Med 2006; 20: 382–387.
41. Bertone JJ, Jones RL, Curtis CR. Evaluation of a test kit for determination of serum immunoglobulin G concentration in foals. J Vet Intern Med 1988; 2: 181–183.
42. Pusterla N, Pusterla JB, Spier SJ, et al. Evaluation of the SNAP Foal IgG test for the semiquantitative measurement of immunoglobulin G in foals. Vet Rec 2002; 151: 258–260.
43. George JW, O'Neill SL. Comparison of refractometer and biuret methods for total protein measurement in body cavity fluids. Vet Clin Pathol 2001; 30: 16–18.
44. Carlson GP, Harrold DR. Relationship of protein concentration and water content of equine serum and plasma samples. Vet Clin Pathol 1977; 6: 18–20.
45. Bauer JE. Normal blood chemistry. In: Koterba AM, Drummond WA, Kosch PC, eds. Equine clinical neonatology. Philadelphia: Lea & Febiger, 1990; 602–622.
