Point-of-care glucometers are commonly used in veterinary hospitals to quickly measure blood glucose concentrations in whole blood samples obtained from cats. Although use of POC glucometers is generally recommended only for healthy patients,1 they are often used for assessment of critically ill cats because of their ease of use and accessibility. Accurate measurement of blood glucose concentration is clinically important, particularly in critically ill patients with abnormal blood glucose concentrations.
There are interspecies differences in the hematologic distribution of unbound and hemoglobin-bound glucose.2 In humans, blood glucose is distributed approximately equally between the hemoglobin-bound and unbound states.3 In comparison, dogs and cats have substantially more unbound than hemoglobin-bound glucose.2 Point-of-care glucometers designed for veterinary use (eg, in dogs and cats) are intended to improve the accuracy of blood glucose measurement by accounting for such differences.1
Although a POC glucometer that uses a species-specific algorithm provides reliable and accurate measurements of blood glucose concentrations in healthy dogs and cats, inaccurate results can occur with some POC glucometers when used for patients of various species with PCV, Hct, or hemoglobin values outside the reference ranges.1,4–16 The authors assume that similar inaccuracies in measurement of blood glucose concentration can occur in cats with hemodilution or hemoconcentration, although the magnitude of the effect in this species is unknown. A previous study14 by our group evaluated the effects of PCV on canine blood glucose measurements with the same POC glucometer used in the study reported here and showed that a correction formula that adjusted results for a patient's PCV was useful for mitigating the effect of PCV on POCgluc.
The purpose of the study reported here was to determine the effect of a wide range of PCVs on POCgluc in feline blood samples. Additional goals were to develop and evaluate a correction formula to improve the accuracy of these measurements in a convenience sample of hospitalized cats. We hypothesized that POCgluc would correlate closely with LABgluc for samples with PCVs within the reference range and that hemodiluted and hemoconcentrated samples (ie, with PCVs below or above this range) would yield falsely increased and decreased POCgluc values, respectively. Lastly, we hypothesized that the use of a correction formula to adjust for a given PCV would improve the correlation between POCgluc and LABgluc measurements in feline blood samples with a wide range of PCVs.
Materials and Methods
Sample collection and processing for healthy blood donor cats
Blood samples were collected from 4 healthy cats that were routinely used as blood donors at the University of Georgia Veterinary Teaching Hospital for use in developing the correction formula. The blood collection was compliant with a protocol approved by the university's institutional animal care and use committee. The cats were deemed healthy on the basis of history and a physical examination performed on the day of sample collection. Standard aseptic jugular venipuncture technique was used following sedation with midazolam hydrochloride (0.2 mg/kg [0.09 mg/lb], IV) and ketamine hydrochloride (3 mg/kg [1.4 mg/lb], IV). A 19-gauge winged infusion needlea and a 60-mL syringe containing 625 U of sodium heparin anticoagulantb were used to obtain 50 mL of blood from each cat.
Blood samples were processed immediately following collection by use of previously described methods.14 In brief, 3 micro-Hct tubes/cat were filled with blood and centrifuged in a micro-Hct centrifugec at 11,800 × g for 3 minutes. Baseline measurements of PCV (in triplicate) and total protein concentration (single measurement) for each cat were obtained with a micro-Hct capillary tube reader cardd and refractometer,e respectively. Duplicate baseline POCgluc measurements were obtained with a POC glucometerf designed for veterinary use. The glucometer was operated by 1 author (AK) according to the manufacturer's instructions for use with feline blood samples. All test strips were of the same lot number. All measurements for PCV (reference range, 35% to 50%) and total protein (reference range, 6.5 to 8.2 g/dL) were performed by one of the authors (SLL).
The remainder of each 50-mL blood sample was then immediately transferred to a disposable blood transfer bagg and centrifugedh at 6,500 × g for 6 minutes. A manual plasma extractori was used to decant the plasma into a glass beaker, and the packed RBCs were transferred to a separate glass beaker. A PCV measurement was obtained for the packed RBCs from each cat by use of the described procedures. Twelve suspensions containing various quantities of packed RBCs (0% and approx 8%, 13%, 20%, 26%, 32%, 40%, 50%, 58%, 64%, 70%, and 80%) were prepared in glass tubesj with aliquots of packed RBCs (0.5 to 2 mL/tube) that were resuspended with various quantities of plasma. Because the PCVs of the packed RBCs varied slightly among the healthy blood donor cats (range, 77% to 87%), there was slight variability in the PCVs of the prepared suspensions. All suspensions were made by one of the authors (SLL). One total protein measurement and duplicate PCV and POCgluc measurements were obtained for each suspension by the described procedures. Because there was no manufacturer-established reference range for the POC glucometer used in the present study, the baseline POCgluc of each healthy cat served as the basis for comparison with POCgluc measurements obtained for that cat's suspensions.
The remaining portions of each suspension were then centrifugedc at 1,500 × g for 5 minutes. The plasma was decanted and immediately frozen at −20°C. Three days later, a clinical laboratory biochemical analyzerk was used to batch analyze the plasma samples for glucose concentration (reference range, 66 to 142 mg/dL) at the clinical pathology laboratory of the veterinary teaching hospital, which was accredited by the American Association of Veterinary Laboratory Diagnosticians. Prior to freezing the plasma in preparation for testing at the clinical pathology laboratory, all samples, including plasma, packed RBCs, and RBC suspensions, were kept on ice for no longer than 2 hours after sample collection for processing and handling.
Prospective sample collection from hospitalized cats
Following completion of blood collection from healthy cats, a convenience sample of 16 cats admitted to the veterinary teaching hospital was obtained prospectively over a 6-month period. Standard venipuncture technique was used to obtain heparinized blood samples from hospitalized cats during routine diagnostic evaluation. Because blood samples were obtained during standard patient care, no client consent or approval by the clinical research committee or the institutional animal care and use committee was required. Total protein concentration, PCV, LABgluc, and duplicate POCgluc measurements were obtained for each blood sample according to the procedures described for healthy cats.
Statistical analysis
Data analyses were performed with commercially available statistical software.1 Values of P < 0.05 were considered significant. Duplicate measurements were used to calculate mean PCV and POCgluc for each suspension from each healthy cat. A paired t test was used to compare mean POCgluc and LABgluc values for suspensions from healthy cats at various PCVs to identify the PCV range beyond which these values differed significantly.
Development of the correction formula—A correction formula for POCgluc was developed on the basis of the slope and intercept from a simple linear regression model describing the relationship between mean PCV (predictor) and the difference in glucose concentrations between methods (mean POCgluc minus LABgluc; outcome). The goodness of fit of the linear regression model was assessed by evaluation of the coefficient of determination (R2).
Evaluation of the correction formula—A paired t test was used to compare mean POCgluc and LABgluc values for samples from healthy and hospitalized cats; corrected POCgluc and LABgluc values were similarly compared. Pearson correlation coefficients (r) were calculated to evaluate the linear association between mean POCgluc and LABgluc and between corrected POCgluc and LABgluc for healthy and hospitalized cats.
Results
Healthy cats
At baseline, the mean PCV and mean total protein concentration for the healthy cats were 34% (range, 32% to 38%) and 7.1 g/dL (range, 6.6 to 8.0 g/dL), respectively. The mean POCgluc at baseline was 83.4 mg/dL (range, 81.5 to 84 mg/dL).
The mean LABgluc measurement for all suspensions was 70.4 mg/dL (range, 59 to 79 mg/dL); all measurements were within the laboratory reference range. The mean PCV values obtained for suspensions ranged from 0% to 87%. Total protein measurements were within the reference range for all suspensions and were not evaluated further. The POC glucometer failed on 2 attempts to produce a POCgluc measurement for the suspension with a mean PCV of 87%, resulting in an error message. The mean POCgluc measurement for all other suspensions was 77.4 mg/dL (range, 20.5 to 130.5 mg/dL).
The mean difference between POCgluc and LABgluc was 24.3 mg/dL (range, −55.5 to 41.5 mg/dL; P = 0.07). The mean difference between POCgluc and LABgluc was not significantly (P = 0.07) different from 0 for suspensions with PCVs within the reference range (35% to 50%); however, the mean difference between these values was significantly (P = 0.04) different from 0 when PCVs were outside the range of 35% to 55%. Compared with LABgluc, false increases in mean POCgluc were detected as mean PCV decreased, and false decreases in mean POCgluc were detected as mean PCV increased (Figure 1). The mean POCgluc results obtained for each cat displayed a similar slope over the range of mean PCV values (Figure 2). On the basis of the slope and intercept of the linear regression model (Figure 3), a correction formula was developed to adjust POCgluc for a given PCV:
Corrected POCgluc = POCgluc + ([1.17 × PCV] – 50.2)
Comparison of mean POCgluc (diamonds) and LABgluc (circles) measurements in suspensions of packed RBCs from 1 (shown as a representation) of 4 healthy blood donor cats in a study to assess the effects of PCV on POCgluc and develop and evaluate a formula that adjusts POCgluc for a given PCV in cats. Packed RBCs were resuspended in various quantities of plasma to achieve a range of PCVs. Duplicate measurements of each suspension were used to identify the mean values of POCgluc and PCV.
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Comparison of mean POCgluc (diamonds) and LABgluc (circles) measurements in suspensions of packed RBCs from 1 (shown as a representation) of 4 healthy blood donor cats in a study to assess the effects of PCV on POCgluc and develop and evaluate a formula that adjusts POCgluc for a given PCV in cats. Packed RBCs were resuspended in various quantities of plasma to achieve a range of PCVs. Duplicate measurements of each suspension were used to identify the mean values of POCgluc and PCV.
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Comparison of mean POCgluc (diamonds) and LABgluc (circles) measurements in suspensions of packed RBCs from 1 (shown as a representation) of 4 healthy blood donor cats in a study to assess the effects of PCV on POCgluc and develop and evaluate a formula that adjusts POCgluc for a given PCV in cats. Packed RBCs were resuspended in various quantities of plasma to achieve a range of PCVs. Duplicate measurements of each suspension were used to identify the mean values of POCgluc and PCV.
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Mean POCgluc measurements obtained over a range of measured PCVs in suspensions of RBCs from 4 healthy blood donor cats (each represented by a different symbol).
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Mean POCgluc measurements obtained over a range of measured PCVs in suspensions of RBCs from 4 healthy blood donor cats (each represented by a different symbol).
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Mean POCgluc measurements obtained over a range of measured PCVs in suspensions of RBCs from 4 healthy blood donor cats (each represented by a different symbol).
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Results of linear regression analysis for the difference between POCgluc and LABgluc (Δgluc) versus measured PCV for the same sample suspensions from healthy cats as in Figure 2. Data points represent results for each cat at each measured PCV value; the solid line represents the linear regression equation (y = −50.2 + 1.17x; R2 = 0.92).
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Results of linear regression analysis for the difference between POCgluc and LABgluc (Δgluc) versus measured PCV for the same sample suspensions from healthy cats as in Figure 2. Data points represent results for each cat at each measured PCV value; the solid line represents the linear regression equation (y = −50.2 + 1.17x; R2 = 0.92).
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Results of linear regression analysis for the difference between POCgluc and LABgluc (Δgluc) versus measured PCV for the same sample suspensions from healthy cats as in Figure 2. Data points represent results for each cat at each measured PCV value; the solid line represents the linear regression equation (y = −50.2 + 1.17x; R2 = 0.92).
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
After application of the correction formula to mean POCgluc values, the mean difference between corrected POCgluc and LABgluc was 3 mg/dL (range, 0.05 to 9.4 mg/dL; P = 0.81). The correlation between corrected POCgluc and LABgluc (r = 0.72; P < 0.001) was higher than that between mean POCgluc and LABgluc (r = 0.59; P < 0.001; Figure 4).
Correlation between LABgluc and mean POCgluc (A) and corrected POCgluc (CorrPOCgluc; B) values for the same sample suspensions from healthy cats as in Figure 2 (circles) and blood samples from 16 hospitalized cats (squares). The correction formula derived from linear regression was as follows: corrected POCgluc = POCgluc + ([1.17 × PCV] – 50.2). The line of perfect agreement between the measurement methods is depicted. A wide distribution of data points away from the line was evident when comparing POCgluc with LABgluc. The correlation between corrected POCgluc and LABgluc was significant for both healthy (P < 0.001) and hospitalized (P < 0.001) cats.
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Correlation between LABgluc and mean POCgluc (A) and corrected POCgluc (CorrPOCgluc; B) values for the same sample suspensions from healthy cats as in Figure 2 (circles) and blood samples from 16 hospitalized cats (squares). The correction formula derived from linear regression was as follows: corrected POCgluc = POCgluc + ([1.17 × PCV] – 50.2). The line of perfect agreement between the measurement methods is depicted. A wide distribution of data points away from the line was evident when comparing POCgluc with LABgluc. The correlation between corrected POCgluc and LABgluc was significant for both healthy (P < 0.001) and hospitalized (P < 0.001) cats.
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Correlation between LABgluc and mean POCgluc (A) and corrected POCgluc (CorrPOCgluc; B) values for the same sample suspensions from healthy cats as in Figure 2 (circles) and blood samples from 16 hospitalized cats (squares). The correction formula derived from linear regression was as follows: corrected POCgluc = POCgluc + ([1.17 × PCV] – 50.2). The line of perfect agreement between the measurement methods is depicted. A wide distribution of data points away from the line was evident when comparing POCgluc with LABgluc. The correlation between corrected POCgluc and LABgluc was significant for both healthy (P < 0.001) and hospitalized (P < 0.001) cats.
Citation: Journal of the American Veterinary Medical Association 254, 10; 10.2460/javma.254.10.1180
Hospitalized cats
The mean PCV and mean total protein concentration for hospitalized cats were 36% (range, 10% to 69%) and 7.3 g/dL (range, 5.4 to 9.8 g/dL), respectively. On the basis of LABgluc measurements, 7 of 16 hospitalized cats were classified as hyperglycemic and none were classified as hypoglycemic. The mean values for POCgluc and LABgluc were 169.3 mg/dL (range, 56 to 541 mg/dL) and 151.8 mg/dL (range, 72 to 388 mg/dL), respectively. The mean difference between POCgluc and LABgluc for hospitalized cats was 41.5 mg/dL (range, −153 to 68 mg/dL; P = 0.22). The mean difference between corrected POCgluc and LABgluc was 25.9 mg/dL (range, −127.4 to 44.5 mg/dL; P = 0.33), with the greatest differences observed at the highest blood glucose concentrations. The correlation between corrected POCgluc and LABgluc (r = 0.97; P < 0.001) was higher than that between mean POCgluc and LABgluc (r = 0.93; P < 0.001; Figure 4).
Discussion
Results of the present study showed that the measurements obtained by means of a veterinary POC glucometer differed significantly from the laboratory values when suspensions of feline RBCs with PCVs outside the range of 35% to 55% were tested. This manifested as measurements of blood glucose concentration that were falsely increased in hemodiluted blood samples and falsely decreased in hemoconcentrated samples. Previous reports5,14 suggest that PCV and Hct of blood samples from dogs affect POCgluc measurements. Compared with results for our previous study14 of dogs, the smaller values obtained for the slope and intercept of the linear regression model in the present study indicated that PCV had a lower impact on POCgluc measurements for feline samples than for canine samples. This finding was likely a reflection of the differences in the hematologic distribution of unbound and hemoglobin-bound glucose between dogs and cats.1,3
The POC glucometer used in the present study measured glucose concentration in a whole blood sample and estimated the plasma glucose concentration under the assumption that the patient's PCV was within the reference range. This estimate of the plasma glucose concentration may be inaccurate for patients with PCVs that are outside the reference range. The concentration of glucose is higher in plasma than in whole blood because of the displacement of glucose by the lipid membranes and hemoglobin proteins of RBCs in whole blood samples. When a whole blood sample is placed on the glucometer test strip, the plasma filters through to the reagent layer where the plasma glucose is involved in an electrochemical reaction, generating an electrical current that is proportional in magnitude to the glucose concentration of the sample. A high PCV results in the availability of less plasma to react with the reagent layer, with resultant lower glucose concentration estimates; conversely, a low PCV results in the availability of more plasma for the reaction, with resultant higher glucose concentration estimates.
The purpose of the correction formula developed in the present study was to convert POCgluc measurements into a corrected POCgluc value that more closely approximated LABgluc. The use of corrected blood glucose measurements may help prevent clinical errors, such as failing to treat an anemic animal for hypoglycemia or inappropriately treating a patient with hemoconcentration for hypoglycemia.
In hospitalized cats, both POCgluc and corrected POCgluc were significantly correlated with LABgluc; however, the corrected POCgluc values were more closely correlated with LABgluc than were the POCgluc values. Although the mean differences between POCgluc and LABgluc and between corrected POCgluc and LABgluc were not significantly different from 0 for hospitalized cats, the greatest differences between corrected POCgluc and LABgluc were observed at the highest blood glucose concentrations. On the basis of LABgluc measurements, there was a high prevalence of hyperglycemia among hospitalized cats. Similar to results for the previous study14 of dogs, our correction formula was less useful for prediction of blood glucose concentration in profoundly hyperglycemic cats. This may have been the result of a relatively greater increase in the amount of glucose that was unbound versus hemoglobin bound in patients with marked hyperglycemia as well as the limits of the internal glucometer algorithm for estimating blood glucose concentration.15 The lack of significant differences between POCgluc and LABgluc values for hospitalized cats in the present study may have been attributable to the small sample size. Additional studies that include blood samples from feline patients with a wide range of PCVs and blood glucose concentrations are needed to evaluate the frequency with which our correction formula might prevent clinical errors.
A study limitation was that the influence of factors known to interfere with POC glucometer readings, including Pao2, Pco2, blood pH, and certain drugs such as mannitol or dopamine, was not examined for the clinical samples.17–20 The use of capillary samples can result in inaccurate blood glucose measurements owing to the effects of hypoperfusion, shock, and increased use of glucose by tissues in critically ill patients.11,16,21,22 However, the effect of capillary sample use on performance of the correction formula was not evaluated in the present study. The use of venous blood samples in the present study, which was consistent with the POC glucometer manufacturer's guidelines,1 should minimize any potential effect of the sample collection technique. It should also be noted that the correction formula described in the present study is applicable only to the POC glucometer that was tested. Finally, the effects of sedation, hemolysis, icterus, lipemia, sample storage time, and other factors that could potentially influence POCgluc and LABgluc measurements and results obtained with the correction formula were not evaluated in the present study.
The correction formula used in the present study incorporated PCV, which is inherently subject to some user interpretation. To mitigate this variability, the same investigator obtained all PCV readings for the present study. We chose to use PCV, rather than Hct measurements, because PCV evaluation is commonly performed in-house in nearly all veterinary clinics.
Reduction of error rates in glucose concentration measurement resulting from anemia has been achieved in human medicine by the use of correction formulas for several different types of POC glucometers and by use of multichannel instead of single-channel glucometers.10,23,24 Although the use of correction formulas for POCgluc is quick and simple, the authors advocate for incorporation of species-specific formulas to correct for PCV effects into the intrinsic algorithm of the POC glucometer. The incorporation of a corrective formula could potentially be helpful in mitigating PCV-related inaccuracies when POC glucometers that are marketed to veterinarians are used for measurement of blood glucose concentration in cats and is deserving of further study.
Acknowledgments
Dr. Koenig has previously received research support (supplies) from Abbott Animal Health.
No funding was received for this study. The authors declare that there were no conflicts of interest.
Presented in abstract form at the 2015 International Veterinary Emergency and Critical Care Symposium, Washington, DC, September 2015.
The authors thank Dr. Deborah Keys for assistance with statistical analyses.
ABBREVIATIONS
| LABgluc | Plasma glucose concentration as measured by a clinical laboratory biochemical analyzer |
| POC | Point-of-care |
| POCgluc | Blood glucose concentration as measured by a point-of-care glucometer |
Footnotes
Exel International scalp vein butterfly set, Thermo Fisher Scientific, Waltham, Mass.
Heparin sodium injection USP, 1,000 U/mL, Hospira, Lake Forest, Ill.
Triac centrifuge, BD, Franklin Lakes, NJ.
Micro-Hct capillary tube reader, Veterinary Information Network, Davis, Calif.
JorVet J-351 refractometer, Jorgenson Laboratories Inc, Loveland, Colo.
AlphaTRAK 2, Abbott Laboratories, Abbott Park, Ill.
Teruflex transfer bag, Terumo Medical Corp, Somerset, NJ.
Sorvall RC-3, Thermo Fisher Scientific Inc, Waltham, Mass.
Manual plasma extractor, Fenwal Inc, Lake Zurich, Ill.
BD Vacutainer glass blood tube, Becton, Dickinson and Co, Franklin Lakes, NJ.
Hitachi P-Modular 800, Roche Diagnostics, Indianapolis, Ind.
SAS, version 9.3, SAS Institute Inc, Cary, NC.
References
1. Alpha TRAK 2 [package insert]. Abbott Park, Ill: Abbott Laboratories, 2012.
2. Coldman MF, Good W. The distribution of sodium, potassium and glucose in the blood of some mammals. Comp Biochem Physiol 1967;21:201–206.
3. Mackay EM. The distribution of glucose in human blood. J Biol Chem 1932;97:685–689.
4. Johnson BM, Fry MM, Flatland B, et al. Comparison of a human portable blood glucose meter, veterinary portable blood glucose meter, and automated chemistry analyzer for measurement of blood glucose concentrations in dogs. J Am Vet Med Assoc 2009;235:1309–1313.
5. Paul AE, Shiel RE, Juvet F, et al. Effect of hematocrit on accuracy of two point-of-care glucometers for use in dogs. Am J Vet Res 2011;72:1204–1208.
6. Solnica B, Skupien J, Kusnierz-Cabala B, et al. The effect of hematocrit on the results of measurements using glucose meters based on different techniques. Clin Chem Lab Med 2011;50:361–365.
7. Jamaluddin FA, Gunavathy M, Yean CY, et al. Variability of point-of-care testing blood glucometers versus the laboratory reference method in a tertiary teaching hospital. Asian Biomed 2012;6:67–74.
8. Tang Z, Lee JH, Louie RF, et al. Effects of different hematocrit levels on glucose measurements with handheld meters for point-of-care testing. Arch Pathol Lab Med 2000;124:1135–1140.
9. Barreau PB, Buttery JE. The effect of the haemoatocrit value on the determination of glucose levels by the reagent-strip methods. Med J Aust 1987;147:286–288.
10. Pidcoke HF, Wade CE, Mann EA, et al. Anemia causes hypoglycemia in intensive care unit patients due to error in single-channel glucometers: methods of reducing patient risk. Crit Care Med 2010;38:471–476.
11. Hussain K, Sharief N. The inaccuracy of venous and capillary blood glucose measurement using reagent strips in the newborn period and the effect of haematocrit. Early Hum Dev 2000;57:111–121.
12. Krinsley JS, Grover A. Severe hypoglycemia in critically ill patients: risk factors and outcomes. Crit Care Med 2007;35:2262–2267.
13. Ginsberg BH. Factors affecting blood glucose monitoring: sources of errors in measurement. J Diabetes Sci Technol 2009;3:903–913.
14. Lane SL, Koenig A, Brainard BM. Formulation and validation of a predictive model to correct blood glucose concentrations obtained with a veterinary point-of-care glucometer in hemodiluted and hemoconcentrated canine blood samples. J Am Vet Med Assoc 2015;246:307–312.
15. Tonyushkina K, Nichols JH. Glucose meters: a review of technical challenges to obtaining accurate results. J Diabetes Sci Technol 2009;3:971–980.
16. Dungan K, Chapman J, Braithwaite SS, et al. Glucose measurement: confounding issues in setting targets for inpatient management. Diabetes Care 2007;30:403–409.
17. Tang Z, Du X, Louie RF, et al. Effects of drugs on glucose measurements with handheld glucose meters and a portable glucose analyzer. Am J Clin Pathol 2000;113:75–86.
18. Karon BS, Griesmann L, Scott R, et al. Evaluation of the impact of hematocrit and other interference on the accuracy of hospital-based glucose meters. Diabetes Technol Ther 2008;10:111–120.
19. Louie RF, Tang Z, Sutton DV, et al. Point-of-care glucose testing: effects of critical care variables, influence of reference instruments, and a modular glucose meter design. Arch Pathol Lab Med 2000;124:257–266.
20. Rebel A, Rice MA, Fahey BG. The accuracy of point-of-care glucose measurements. J Diabetes Sci Technol 2012;6:396–411.
21. D'Orazio P, Burnett RW, Fogh-Andersen N, et al. Approved IFCC recommendation on reporting results for blood glucose: International Federation of Clinical Chemistry and Laboratory Medicine Scientific Division, Working Group on Selective Electrodes and Point-of-Care Testing (IFCC-SD-WGSEPOCT). Clin Chem Lab Med 2006;44:1486–1490.
22. Atkin SH, Dasmahapatra A, Jaker MA, et al. Fingerstick glucose determination in shock. Ann Intern Med 1991;114:1020–1024.
23. Mann EA, Salinas J, Pidcoke HF, et al. Error rates resulting from anemia can be corrected in multiple commonly used point-of-care glucometers. J Trauma 2008;64:15–21.
24. Girouard J, Forest JC, Massé J, et al. Multicenter evaluation of the Glucometer Elite XL meter, an instrument specifically designed for use with neonates. Diabetes Care 2000;23:1149–1153.
