Example: barber

Evaluation of a Point-Of-Care (POC) Glucose Meter Suitable ...

Department of clinical Pathology, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada, 2 Department of Pathobiology and Laboratory Medicine, University of Toronto and 3 Gamma-Dynacare Medical Laboratories, Ontario, Canada Evaluation of a Point-Of-Care (POC) Glucose Meter Suitable for Use in Complex Tertiary care Facilities1,2 Chan, , 1 Rozmanc, M, 2,3 Seiden-Long, 1 Kwan of clinical Pathology, Sunnybrook Health Sciences Centre, Toronto, Ontario, M4N 3M5, 2 Department of Pathobiology and Laboratory Medicine, University of Toronto and 3 Gamma-Dynacare Medical Laboratories, Ontario, Canada. Objective: To evaluate the performance and suitability of a single Glucose Meter for use in a complex tertiary care facility (TCF);Introduction: POC Glucose meters are commonly used in TCF, despite some known limitations and concerns.

Department of Clinical Pathology, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada, 2. Department of Pathobiology and Laboratory Medicine, University of Toronto and

Tags:

  Clinical, Suitable, Care, Meters, Glucose, Glucose meter suitable

Information

Domain:

Source:

Link to this page:

Please notify us if you found a problem with this document:

Other abuse

Transcription of Evaluation of a Point-Of-Care (POC) Glucose Meter Suitable ...

1 Department of clinical Pathology, Sunnybrook Health Sciences Centre, Toronto, Ontario, Canada, 2 Department of Pathobiology and Laboratory Medicine, University of Toronto and 3 Gamma-Dynacare Medical Laboratories, Ontario, Canada Evaluation of a Point-Of-Care (POC) Glucose Meter Suitable for Use in Complex Tertiary care Facilities1,2 Chan, , 1 Rozmanc, M, 2,3 Seiden-Long, 1 Kwan of clinical Pathology, Sunnybrook Health Sciences Centre, Toronto, Ontario, M4N 3M5, 2 Department of Pathobiology and Laboratory Medicine, University of Toronto and 3 Gamma-Dynacare Medical Laboratories, Ontario, Canada. Objective: To evaluate the performance and suitability of a single Glucose Meter for use in a complex tertiary care facility (TCF);Introduction: POC Glucose meters are commonly used in TCF, despite some known limitations and concerns.

2 clinical conditions associated with extreme body chemistries, use of potentially interfering medications & biologics, etc, are common in TCF and pose unique challenges to POC Glucose testing. The ability to use a single, yet reliable and accurate POC Glucose Meter that can satisfy different clinical needs, will not only be cost effective, but also improve patient care and safety. Methods: The StatStrip Glucose Meter from Nova Biomedical, MA, was evaluated for use in a 1300-bed (125 critical care ), 3-site, teaching TCF in Toronto, involving 4 technologists and 20 front-line nurses (Table 1); Intra-assay and long term imprecision were determined using aqueous, commercial stabilized whole blood (SWB) and fresh whole blood materials. A CV<5% is deemed to be acceptable; Lab Glucose measurements were made on the Modular analyzer from Roche, using a Glucose oxidase method (CV<2%); Analytical interferences were tested for 9 substances at 3 Glucose concentrations.

3 A difference in Meter Glucose measurements greater than the larger of 10% or mmol/L from the control sample is considered significant. Effects of endogenous parameters (Hct, Hb, pH, pCO2, pO2, HCO3, lactate, Na, K, Cl and iCa measured on the ABL800 Flex, Radiometer) on Meter Glucose measurement were assessed by backward stepwise linear regression (SigmaStat from SPSS), and discrepancies from lab measurement were compared at different levels of contributing parameters. Meter and lab Glucose measurements were correlated using simple linear regression analysis and Pearson s correlation coefficient was calculated ( , SPSS). Meter inaccuracy was assessed by: Bias plot with ISO 15197/CSLI criteria [1,2] Consensus grid analyses (EP evaluator release 8, David G. Rhoads Associates Inc). Modified Locally Smoothed-Median Absolute Difference (LS-MAD) [3] Statistical significance were accepted at p< 1.

4 Sampling procedures for Glucose Measurements by Meter and Central Laboratory Operator refers to personnel who operated the Meter to measure the Glucose levels using the blood sample described. clinical Services Sample Collection & M eter Operators Neonatal ICU Meter : capillary blood sample by heel prick; nurse operator Lab: plasm a from capillary sam ple by heel prick collected into a microtainer tube Diabetic Clinic Meter : capillary blood sample by finger prick; nurse operator Lab: separate venous blood into grey top tubes CVICU Meter : arterial blood drawn from Arterial line into a syringe; nurse operator Lab: plasma from arterial blood drawn from arterial line into a second syringe and then into a green top tube Stat Lab (Operating Rooms and ICUs) Meter : arterial blood in gas syringes; laboratory technologist operator Lab: same arterial blood for ABL 800 Flex (electrolytes & blood gas parameters), then spun down for plasma to be used for Glucose measurement on the Roche Modular Delivery Suite Meter : cord blood collected into gas syringes; nurse operator Lab: plasma from the same cord blood sample Dialysis Unit Meter : arterial/venous blood drawn from fistula or central venous catheter into a syringe; nurse operator Lab: plasma from sample collected into a second syringe & then dispensed into a green top tube 1 Department of clinical Pathology, Sunnybrook Health Sciences Centre, Toronto, Ontario, M4N 3M5, 2 Department of Pathobiology and Laboratory Medicine, University of Toronto and 3 Gamma-Dynacare Medical Laboratories, Ontario, Canada (Cont d)Table 2.

5 Imprecision Study Material N Glucose Concentration, mmol/L Imprecision, % CV Aqueous 70 70 Stabilized whole blood 83 83 Arterial Whole Blood (within run) 20 20 Stabilized whole blood (within run) 22 22 22 Table 1. Sampling procedures for Glucose Measurements by Meter and Central LaboratoryTable 2. Imprecission StudyTable 3. Potentially Interfering Substances and Concentrations Tested Substances Tested Units ConcentrationsGlucose Concentration, mmol/L Glucose Concentration, mmol/L Glucose Concentration, mmol/L Haematocrit, % 27 49 68 Hb, g/L 0 2 10 * Acetaminophen, mol/L 0 330 660 Bilirubin, mol/L 0 257 855 Ascorbic acid, mol/L 0 290 590 Maltose, mmol/L 0 Galactose, mmol/L 0 Lactate, mmol/L 0 10 20 Uric acid, mmol/L 0 -OH Butyrate, mmol/L 0 15 30 Each Glucose concentration represents the mean of four Glucose measurements using four different Glucose meters .

6 A difference of the larger of 10% or mmol/L is considered significant and is denoted by *. Table 3. Potentially Interfering Substances and Concentrations TestedResults:Imprecisions were all < 5% (Table 2);Interferences: all were insignificant except total hemoglobin (Hb) at 10 g/L, showing a 9-18% reduction in measured Glucose values (Table 3);Backward stepwise linear regression identified pCO2 & Na as contributing factors on the Meter Glucose measurement, but effects were negligible (mean difference over lab value< mmol/L);Correlation of Meter and lab results: slope= (95% CI ), intercept= ( ), r= (Figure 1);Consensus grid analysis: 97% and 3% in regions A and B respectively (Table 4), and 97% met ISO 15197 & CSLI (formerly NCCLS) criteria (Figure 2 & 3).LS-MAD showed median differences were stable up to about 10 mmol/L and started to increase afterwards (Figure 4).

7 1 Department of clinical Pathology, Sunnybrook Health Sciences Centre, Toronto, Ontario, M4N 3M5, 2 Department of Pathobiology and Laboratory Medicine, University of Toronto and 3 Gamma-Dynacare Medical Laboratories, Ontario, Canada (Cont d)Table 4. Relative Accuracy of the Meter Glucose Measurement 1 Consensus Grid acceptance criteria: 95% of data within Region A & no sample outside region B 2 ISO/CSLI criteria: 95% of data with a difference mmol/L for Glucose concentration mmol/L or 20% if Glucose concentration is > mmol/L Consensus Grid1 % Error % in Region clinical Service N Mean 95th centile A B ISO/CSLI2 Criteria Met, % Neonatal ICU 56 98 2 95 Diabetic Clinic 43 95 5 95 CVICU 45 93 7 98 Stat Lab (Operating rooms and ICUs)

8 163 98 2 99 Delivery Suite 47 100 0 96 Dialysis 32 94 6 94 Overall 386 97 3 97 Figure 1. Linear Regression N=386 Meter = ( ) *Lab + ( ) r = concentrations covered: Mean SD (range)Lab: ( ) mmol/LMeter: ( ) mmol/LTable 4. Relative Accuracy of the Meter Glucose MeasurementFigure 1. Linear RegressionLab, mmol/L2624222018161412108642 Meter , mmol/L302826242220181614121086421 Department of clinical Pathology, Sunnybrook Health Sciences Centre, Toronto, Ontario, M4N 3M5, 2 Department of Pathobiology and Laboratory Medicine, University of Toronto and 3 Gamma-Dynacare Medical Laboratories, Ontario, Canada (Cont d)Figure 3.

9 Consensus Grid AnalysisFigure 2. Bias Plot with ISO LimitsBias Plot with ISO 15197 LimitsN=386-8-6-4-2024680510152025303540 Lab Glucose (mmol/L) Meter - Lab Glucose (mmol/L)Discussion & Conclusions: The imprecision (CV<5%) is acceptable, and will rarely lead to major errors in insulin dose as suggested previously [4]. CV will be ideal but is not achievable consistently by any known meters on the market. The StatStrip Meter is relatively free from common interferences seen in TCF such as Hct, maltose, etc. The effects of free Hb (as in hemolysis), pCO2 and Na are marginal and will not compromise the Meter performance significantly. The Meter results compare very well with plasma values from the Laboratory, and meet the suggested acceptance criteria from CSLI and ISO, although fall short of those proposed by ADA (5% total error).

10 No significant error (Type 1 & 2) was found with the current Tight Glycemic Control target of mmol/L. Overall, StatStrip Meter is found to be a good candidate for general use in a complex of clinical Pathology, Sunnybrook Health Sciences Centre, Toronto, Ontario, M4N 3M5, 2 Department of Pathobiology and Laboratory Medicine, University of Toronto and 3 Gamma-Dynacare Medical Laboratories, Ontario, Canada (Cont d)Locally-smoothed Median Absolute Difference Glucose (mmol/L)Median Absolute Difference, MAD (mmol/L)MADE rror Lim itUpper CILower CIFigure 4. Modified LS-MAD N=386 Note: mmol/L ( current Tight Glycemic Range of 5-8) was used instead of mmol/L as suggested by Kost, et al [3]. A error limit (calculated based on Meter and lab CV) shown instead of a fixed mmol/L : NCCLS. Point-Of-Care Blood Glucose Testing in Acute and Chronic care Facilities; Approved Guideline 2nd Ed.