Estimation of reference intervals from small samples: an example using canine plasma creatinine

Background: According to international recommendations, reference intervals should be determined from at least 120 reference individuals, which often are impossible to achieve in veterinary clinical pathology, especially for wild animals. When only a small number of reference subjects is available, the possible bias cannot be known and the normality of the distribution cannot be evaluated. A comparison of reference intervals estimated by different methods could be helpful. Objective: The purpose of this study was to compare reference limits determined from a large set of canine plasma creatinine reference values, and large subsets of this data, with estimates obtained from small samples selected randomly. Methods: Twenty sets each of 120 and 27 samples were randomly selected from a set of 1439 plasma creatinine results obtained from healthy dogs in another study. Reference intervals for the whole sample and for the large samples were determined by a nonparametric method. The estimated reference limits for the small samples were minimum and maximum, mean ± 2 SD of native and Box–Cox‐transformed values, 2.5th and 97.5th percentiles by a robust method on native and Box–Cox‐transformed values, and estimates from diagrams of cumulative distribution functions. Results: The whole sample had a heavily skewed distribution, which approached Gaussian after Box–Cox transformation. The reference limits estimated from small samples were highly variable. The closest estimates to the 1439‐result reference interval for 27‐result subsamples were obtained by both parametric and robust methods after Box–Cox transformation but were grossly erroneous in some cases. Conclusion: For small samples, it is recommended that all values be reported graphically in a dot plot or histogram and that estimates of the reference limits be compared using different methods.     

Reference intervals for biochemical and haematological parameters in mature domestic donkeys (Equus asinus) in the UK

Reference intervals (RIs) for haematology and serum biochemistry for donkeys in a temperate climate have been previously published using blood sample results from the resident population of a large donkey shelter in the UK. Periodic review of reference intervals is recommended to ensure their applicability to the patient population and changes in laboratory methods and technologies. The current study aimed to revise the previously published haematology and serum biochemistry values for the adult domestic donkey (Equus asinus) in the UK in the light of a change in analytical equipment at the Donkey Sanctuary laboratory, but also to refine the demography of the sample population with respect to age, physiology and clinical history. Clinical pathology results from 138 clinically healthy mature (4–24 years inclusive) female and castrated male donkeys selected from the resident population of the Donkey Sanctuary, were analysed retrospectively. The animals were blood sampled during the period February 2008 to June 2011 as part of a routine health screen prior to rehoming. Results for a total of 38 biochemical and haematological parameters were analysed including 3 previously unreferenced parameters in addition to those assessed in the previous study. The new reference intervals and median values show very poor transferability with recently derived reference intervals for non‐Thoroughbred horses and only limited transferability with reference intervals previously published for donkeys in the UK. Of particular note is a marked reduction in the upper reference limit for triglycerides of 2.8 mmol/l (from 4.3 mmol/l) since this parameter is used to decide when donkeys are at risk of developing hyperlipaemia. This study demonstrates the value of intermittent review of reference intervals and refinement of study populations. Notwithstanding the caution with which reference interval data from different laboratories should be compared, the lack of transferability of results between donkeys and horses highlights the importance of use of species‐appropriate reference intervals for clinical decision‐making.     

Reference values: a review

Abstract: Reference values are used to describe the dispersion of variables in healthy individuals. They are usually reported as population‐based reference intervals (RIs) comprising 95% of the healthy population. International recommendations state the preferred method as a priori nonparametric determination from at least 120 reference individuals, but acceptable alternative methods include transference or validation from previously established RIs. The most critical steps in the determination of reference values are the selection of reference individuals based on extensively documented inclusion and exclusion criteria and the use of quality‐controlled analytical procedures. When only small numbers of values are available, RIs can be estimated by new methods, but reference limits thus obtained may be highly imprecise. These recommendations are a challenge in veterinary clinical pathology, especially when only small numbers of reference individuals are available.     

Hematologic and biochemical reference intervals for Norwegian crossbreed grower pigs

Background: Hematologic and biochemical reference intervals depend on many factors, including environment and age. Reference intervals for Norwegian grower pigs are lacking, and previously published reference intervals for similar pigs from other countries are now outdated due to significant changes in management and breeding on the pig farms. Objectives: The aim of this study was to determine updated reference intervals for hematologic and biochemical analytes in healthy crossbred grower pigs, and to compare the results among 3 different farms. Methods: Blood samples were collected from 104 clinically healthy pigs of the most common Norwegian crossbreed (Landrace Yorkshire sow × Landrace Duroc boar). The pigs were 12–16 weeks old, weighed 30–50 kg, of both sexes, and lived on 3 farms in eastern Norway. Automated hematologic and biochemical analysis were performed using ADVIA 2120 and ADVIA 1650 analyzers. Results: Five samples were excluded because of hemolysis (1) or outliers (4). Reference intervals were calculated using parametric or nonparametric methods, depending on data distribution. Mean, median, minimum, and maximum values were tabulated. Conclusions: The reference intervals calculated in this study will be useful for the diagnosis and monitoring of disease in this widespread crossbreed pig. Compared with previously published reference values, reference intervals for total WBC count, creatine kinase and alanine aminotransferase activities, and albumin, bilirubin, and urea concentrations in this study differed notably.     

Canine reference intervals for the Sysmex XT-2000iV hematology analyzer

Background: The laser‐based Sysmex XT‐2000iV hematology analyzer is increasingly used in veterinary clinical pathology laboratories, and instrument‐specific reference intervals for dogs are not available. Objective: The purpose of this study was to establish canine hematologic reference intervals according to International Federation of Clinical Chemistry and Clinical and Laboratory Standards Institute guidelines using the Sysmex XT‐2000iV hematology analyzer. Methods: Blood samples from 132 healthy purebred dogs from France, selected to represent the most prevalent canine breeds in France, were analyzed. Blood smears were scored for platelet (PLT) aggregates. Reference intervals were established using the nonparametric method. PLT and RBC counts obtained by impedance and optical methods were compared. Effects of sex and age on reference intervals were determined. Results: The correlation between impedance (I) and optical (O) measurements of RBC and PLT counts was excellent (Pearson r=.99 and .98, respectively); however, there were significant differences between the 2 methods (Student's paired t‐test, P<.0001). Differences between sexes were not significant except for HCT, PLT‐I, and PLT‐O. WBC, lymphocyte, and neutrophil counts decreased significantly with age (ANOVA, P<.05). Median eosinophil counts were higher in Brittany Spaniels (1.87 × 10⁹/L), Rottweilers (1.41 × 10⁹/L), and German Shepherd dogs (1.38 × 10⁹/L) than in the overall population (0.9 × 10⁹/L). PLT aggregates were responsible for lower PLT counts by the impedance, but not the optical, method. Conclusion: Reference intervals for hematologic analytes and indices were determined under controlled preanalytical and analytical conditions for a well‐characterized population of dogs according to international recommendations.     

Determination of haematological reference intervals in healthy adult greyhounds

Objective : The purpose of this study was to establish breed‐specific reference intervals for haematological measurands in non‐racing greyhounds. Suitability of the data for partitioning according to sex was also examined. Methods : Haematological data were collected from 304 healthy non‐racing greyhounds and analysed using non‐parametric methods. Results were compared with non‐breed‐specific canine reference intervals and also with greyhound reference intervals obtained by other investigators. Results : Compared with non‐breed‐specific reference intervals, the results showed comparable mean and upper limit and higher lower limit for erythrocyte count; higher values for haemoglobin, haematocrit and mean corpuscular volume; and lower values for total leucocyte count and absolute concentration of neutrophils, lymphocytes, monocytes, eosinophils and platelets. Partitioning according to sex was recommended by the statistical analysis for all analytes except haematocrit and total leucocyte count. Clinical Significance : In this study the reference intervals were derived from a large sample size. The results are in general agreement with previous reports, although higher values for low reference limits have been noted for the erythroid parameters, and lower values for upper reference limits have been observed for the total and different leucocyte counts. Breed‐specific reference intervals provide a useful clinical tool for haematological evaluations.     

Reference values for serum biochemical parameters in free-ranging harp seals

Background — The harp seal (Phoca groenlandica) is one of the most important predators in the Northeastern Atlantic ecosystem. Establishing biochemical reference intervals is important for evaluating the health status of harp seals kept in captivity and for evaluating the effects of environmental changes on the health of populations in the wild. Objective — The purpose of this study was to determine reference values for serum biochemical parameters in wild adult harp seals using readily available current methods. Methods — Blood samples were obtained from 14 adult female harp seals and 9 suckling pups on the pack ice of the Greenland Sea in early March 1998. Seven seals were humanely killed on the ice by permission of the Norwegian Directory for Fisheries and in conjunction with several other research projects. The seals were sampled within 15 minutes postmortem. Remaining seals were captured alive and sampled via the extradural intravertebral vein. Serum biochemical parameters were measured using a Technicon Axon analyzer and included electrolytes (sodium, chloride, potassium, magnesium, and calcium), substrates (free fatty acids, triglycerides, fructosamine, and glucose), end products (urea and uric acid), and proteins (total protein, globulins, and albumin). Serum protein electrophoresis also was done. Data were tested for normality and reference limits were calculated as mean ±1.96 × SD. Results between groups were compared using 2‐tailed t‐tests. Results — Serum levels of glucose and triglycerides were lower, but serum levels of urea were higher in dead animals than in animals that were captured alive. Serum levels for 7 of 17 parameters were significantly different in pups compared with adults. Separate reference intervals were calculated for adult seals and seal pups. Conclusion — Both sampling method and age should be considered when evaluating the results of analysis of serum parameters in wild and captive harp seals.     

Hematology and serum clinical chemistry reference intervals for free‐ranging Scandinavian gray wolves (Canis lupus)

Background: Scandinavian free‐ranging wolves (Canis lupus) are endangered, such that laboratory data to assess their health status is increasingly important. Although wolves have been studied for decades, most biological information comes from captive animals. Objectives: The objective of the present study was to establish reference intervals for 30 clinical chemical and 8 hematologic analytes in Scandinavian free‐ranging wolves. Methods: All wolves were tracked and chemically immobilized from a helicopter before examination and blood sampling in the winter of 7 consecutive years (1998–2004). Seventy‐nine blood samples were collected from 57 gray wolves, including 24 juveniles (24 samples), 17 adult females (25 samples), and 16 adult males (30 samples). Whole blood and serum samples were stored at refrigeration temperature for 1–3 days before hematologic analyses and for 1–5 days before serum biochemical analyses. Reference intervals were calculated as 95% confidence intervals except for juveniles where the minimum and maximum values were used. Results: Significant differences were observed between adult and juvenile wolves for RBC parameters, alkaline phosphatase and amylase activities, and total protein, albumin, γ‐globulins, cholesterol, creatinine, calcium, chloride, magnesium, phosphate, and sodium concentrations. Conclusion: Compared with published reference values for captive wolves, reference intervals for free‐ranging wolves reflected exercise activity associated with capture (higher creatine kinase activity, higher glucose concentration), and differences in nutritional status (higher urea concentration).     

Evaluation of a modified thrombelastography assay initiated with recombinant human tissue factor in clinically healthy horses

Background: Thrombelastography (TEG) is used to evaluate the viscoelastic properties of blood during clotting and provides a global assessment of hemostasis and clot lysis. TEG analysis initiated with recombinant human tissue factor (TF) has not been evaluated in clinically healthy horses. Objectives: The purpose of this study was to determine whether TEG results are affected by the time elapsed between sampling and analysis (storage time) of equine blood samples and to establish a preliminary equine reference interval for a modified TEG assay, using recombinant human TF to initiate coagulation. Methods: Citrated blood samples were obtained from 20 clinically healthy adult horses. Thirteen samples were stored for 30, 60, and 120 minutes at room temperature before TEG analysis. Coagulation was initiated by adding 20 μL of CaCl₂ to 330 μL of blood and 10 μL of diluted recombinant TF for a final dilution of 1:3600. Reaction (R) and clotting (K) times, angle (α), and maximum amplitude (MA) were compared between time points. A preliminary reference interval (minimum–maximum values) was determined using data from all 20 horses after 30 minutes of sample storage. Results: There was a significant effect of storage time on R, K, and α but not MA. Reference intervals were: R, 3.65–6.4 minutes; K, 1.8–5.45 minutes; α, 33.4–66.2°; MA, 41.2–64.1 mm; lysis at 30 minutes post‐MA (LY30), <2.75%; and lysis at 60 minutes post‐MA (LY60), 1.55–9.5%. Conclusions: TEG can be performed on equine citrated blood samples using recombinant human TF to activate clot formation. TEG parameters were significantly affected by storage time, suggesting an incomplete inhibition of coagulation in citrated blood.     

Haematology and plasma biochemistry of Stamboek pre-pubertal gilts in Italy: reference values

Blood samples were taken between February and April from 105 healthy Stamboek pre-pubertal gilts, aged 1-3 months, which were housed at a modern pig farm in northern Italy. The blood was analysed for nine haematological and nine selected haematochemical variables by means of automated and semiautomated blood analysis apparatus. After detection and rejection of outliers, the data were submitted to reference limits evaluation, also taking into account the limits for the red blood cell volume histogram as the anisocytosis index. Some haematological reference values deal with previously published data; in the haematochemical parameters, several discrepancies between evaluated limits and existing reference limits were noted, mainly for aspartate aminotransferase, alanine aminotransferase and lactate dehydrogenase levels and total protein concentration. The results confirm the relevance of age in determining blood reference intervals and that 'normal' values should be determined by each laboratory, taking into account the age of subjects, the sample size and methods of analysis.     

Protein electrophoresis of psittacine plasma

BACKGROUND: Although protein electrophoresis (EPH) has been widely applied in human and veterinary medicine, it has only recently been implemented in the analysis of avian samples. OBJECTIVE: The purpose of this study was to examine the application of protein EPH to the analysis of psittacine plasma samples. Our goals were to describe protein fraction mobility, establish reference intervals for some common species, determine the coefficient of variation (CV) of the chosen method, and examine the effects of sample handling and sample condition. METHODS: Heparinized plasma samples from several common psittacine species (minimum sample size 50 each) were examined using the Beckman Paragon system and SPEP-II gels. Total protein was measured by refractometry. Reference intervals (95%) were calculated by the rank methods. RESULTS: Fraction migration patterns were found to vary among common psittacine species. Day-to-day CV for the EPH fractions ranged from 2.2% to 10.5%; within-run CV ranged from 4.8% to 10.8%; and total CV ranged from 3.2% to 14.8%. The highest CV was noted for the poorly defined alpha-globulin fraction. Prolonged refrigeration, repeated freeze-thawing, hemolysis, and lipemia altered the results. CONCLUSIONS: Protein fractions from psittacine species were variable in terms of migration pattern and protein concentration, which necessitates the use of species-specific reference intervals. Avian protein electrophoretic patterns and values should be interpreted based on knowledge of the CV associated with the technique as well as on the effects of sample handling and condition.     

Reference intervals for feline cerebrospinal fluid: biochemical and serologic variables, IgG concentration, and electrophoretic fractionation

Reference intervals are reported for feline CSF biochemical and serologic variables, IgG concentration, and electrophoretic fractionation, derived from 58 clinically normal adult cats that did not have histologic lesions of the CNS. There was no apparent effect of age on any variable. The CSF total protein concentration was significantly (P = 0.012) greater in males than in females, but all other variables were unaffected by gender. The only variable that had a statistically significant correlation with its corresponding blood concentration was IgG. Blood contamination of the CSF affected the following CSF variables: total protein concentration, activities of lactate dehydrogenase and creatine kinase, IgG ratio, and gamma-globulin percentage. The reference intervals proposed for feline CSF were derived from 33 cats with CSF RBC count less than 31 cells/microliters. Reference limits for CSF with 31 to 1,700 RBC/microliters also are reported.     

Serum protein concentrations from clinically healthy horses determined by agarose gel electrophoresis

Background: Serum protein electrophoresis is a useful screening test in equine laboratory medicine. The method can provide valuable information about changes in the concentrations of albumin and α‐, β‐, and γ‐globulins and thereby help characterize dysproteinemias in equine patients. Reference values for horses using agarose gel as a support medium have not been reported. Objectives: The purpose of this study was to establish reference intervals for serum protein concentrations in adult horses using agarose gel electrophoresis and to assess differences between warm‐blooded and heavy draught horses. In addition, the precision of electrophoresis for determining fraction percentages and the detection limit were determined. Methods: Blood samples were obtained from 126 clinically healthy horses, including 105 Thoroughbreds and 21 heavy draught horses of both sexes and ranging from 2 to 20 years of age. The total protein concentration was determined by an automated biuret method. Serum protein electrophoresis was performed using a semi‐automated agarose gel electrophoresis system. Coefficients of variation (CVs) were calculated for within‐run and within‐assay precision. Data from warm‐blooded and draught horses were compared using the Mann–Whitney U test. Results: Within‐run and within‐assay CVs were <5% for all protein fractions. No significant difference was found between warm‐blooded and heavy draught horses and so combined reference intervals (2.5–97.5%) were calculated for total protein (51.0–72.0 g/L), albumin (29.6–38.5 g/L), α₁‐globulin (1.9–3.1 g/L), α₂‐globulin (5.3–8.7 g/L), β₁‐globulin (2.8–7.3g/L), β₂‐globulin (2.2–6.0 g/L), and γ‐globulin (5.8–12.7 g/L) concentrations, and albumin/globulin ratio (0.93–1.65). Conclusion: Using agarose gel as the supporting matrix for serum protein electrophoresis in horses resulted in excellent resolution and accurate results that facilitated standardization into 6 protein fractions.     

ASVCP reference interval guidelines: determination of de novo reference intervals in veterinary species and other related topics

Reference intervals (RI) are an integral component of laboratory diagnostic testing and clinical decision‐making and represent estimated distributions of reference values (RV) from healthy populations of comparable individuals. Because decisions to pursue diagnoses or initiate treatment are often based on values falling outside RI, the collection and analysis of RV should be approached with diligence. This report is a condensation of the ASVCP 2011 consensus guidelines for determination of de novo RI in veterinary species, which mirror the 2008 Clinical Laboratory and Standards Institute (CLSI) recommendations, but with language and examples specific to veterinary species. Newer topics include robust methods for calculating RI from small sample sizes and procedures for outlier detection adapted to data quality. Because collecting sufficient reference samples is challenging, this document also provides recommendations for determining multicenter RI and for transference and validation of RI from other sources (eg, manufacturers). Advice for use and interpretation of subject‐based RI is included, as these RI are an alternative to population‐based RI when sample size or inter‐individual variation is high. Finally, generation of decision limits, which distinguish between populations according to a predefined query (eg, diseased or non‐diseased), is described. Adoption of these guidelines by the entire veterinary community will improve communication and dissemination of expected clinical laboratory values in a variety of animal species and will provide a template for publications on RI. This and other reports from the Quality Assurance and Laboratory Standards (QALS) committee are intended to promote quality laboratory practices in laboratories serving both clinical and research veterinarians.     

Evaluation of a novel haematology analyser for use with feline blood

A novel haematology analyser was evaluated for its use with feline samples. Complete blood cell counts, a five-part differential count, and reticulocyte numbers were determined, and the results compared with reference data. Coefficients of correlation, Passing–Bablok regression analysis and Bland–Altmann difference plots with biases and 95% limits of agreement are reported. Precision and linearity were also studied. The instrument demonstrated very low imprecision, and the tested range of linearity exceeded the reference ranges provided by the manufacturer. For all parameters except monocytes (r = 0.65), the analyser results correlated well with reference methods. Compared with the manual count of aggregated reticulocytes, the instrument showed good agreement with a positive bias. The optical platelet count correlated well with the manual chamber count. In conclusion the analyser was found to be highly reliable for the analysis of feline blood samples in a large veterinary laboratory.     

Canine hematology and biochemistry reference values.

Reference hematology and biochemistry values for 53 variables are presented from 51 clinically healthy dogs, 26 female and 25 male, approximately six to 24 months of age and of mixed breed. These dogs were sampled because of their good health status and the opportunity to collect the volume of blood required to complete the variable analysis of interest. Collection of blood specimens and laboratory analysis was done in a standard described manner, the latter including a continuing quality control program. For each variable the data were examined for homogeneity and when present, outliers (n = 9) were excluded. Parametric analysis was used to calculate the reference interval for those variables which had a Gaussian distribution or could be changed to a Gaussian distribution by any of four transformations. For those variables in which Gaussian distribution was not present or attained, nonparametric analysis was used. Due to the small size of the population sample, the uncertainty of breed and the exact age of each dog, breed, age and sex effects were not examined. Reference values should be used to assist interpretation of observations obtained from an animal or animals of comparable origin, i.e. similar subpopulation, and only if the same laboratory techniques are followed. Until each laboratory is able to generate reference values using adequate sample size and current methodology for the numerous subpopulations of interest, reference intervals such as these are useful to clinicians and researchers.     

Serum biochemical reference intervals for wild dwarf ornate wobbegong sharks (Orectolobus ornatus)

Background: Sharks are important to sport and commercial fishing, public aquaria, and research institutions. However, serum biochemical reference values have been established for few species. Objective: The aim of this study was to establish serum biochemical reference intervals for wild‐caught dwarf ornate wobbegong sharks (Orectolobus ornatus). Methods: Fifty wobbegongs were caught, and their health status, sex, length, and weight were evaluated and recorded. Following collection of blood, serum biochemical analytes were measured and analyzed using standard analytical and statistical methods. Combined samples generated means, medians, and reference intervals. Results: For the measured analytes, means (reference intervals) were as follows: sodium 287 (284–289) mmol/L, chloride 277 (274–280) mmol/L, potassium 5.2 (5.0–5.3) mmol/L, total calcium 4.6 (4.5–4.7) mmol/L, magnesium 1.9 (1.7–2.0) mmol/L, inorganic phosphate 1.8 (1.7–1.9) mmol/L, glucose 2.6 (2.4–2.8) mmol/L, total protein 46 (45–47) g/L, urea 396 (392–401) mmol/L, creatinine ≤0.02 mmol/L, total bilirubin 2.0 (1.9–2.1) μmol/L, cholesterol 1.3 (1.2–1.4) mmol/L, triglyceride 0.5 (0.4–0.6) mmol/L, alkaline phosphatase 24 (21–28) U/L, alanine aminotransferase 3 U/L, aspartate aminotransferase 28 (25–31) U/L, creatine kinase 49 (38–59) U/L, and osmolarity 1104 (1094–1114) mmol/L. Serum values were not affected by sex, length, or weight. Conclusions: Established reference values will assist with clinical evaluation and treatment of dwarf ornate wobbegongs in aquaria, research institutions, and the wild.     

Reference limits for biochemical and hematological analytes of dairy cows one week before and one week after parturition

Since dairy cows during the transition period have multiple endocrine and metabolic changes, it is necessary to determine the reference limits of laboratory analytes in normal transition cows. Reference limits for the weeks before and after calving were determined in dairy cows. Animals that had adverse clinical outcomes after calving and cows that were culled or had mastitis within the first 7 days after calving were excluded. All biochemical analytes (β-hydroxybutyrate, fatty acids, glucose, cholesterol, urea, calcium, and phosphorus) were statistically different between precalving and postcalving groups. The hematological analytes were not significantly different except for eosinophils. The data from precalving and postcalving cows were significantly different from reference limits in a university-associated laboratory derived from early- and mid-lactation cows. Different reference limits for precalving and postcalving dairy cows should be determined for biochemical analytes to ensure appropriate interpretation of results.