1.

Current Analyzer Competing Analyzer Level I Level 2 Level 1 Level 2 = anjeA MESSY Assay value = Assay value = Assay value = 2.50 ng/dl 5.00 ng/dl 2.50 ng/dl 5.00 ng/dl 2.50 5.00 2.48 4.98 2.48 5.01 2.46 4.80 2.55 5.05 2 45 4.99 2.51 4.98 2.49 4.95 2.52 5.00 2.40 4.89 2.50 5.03 2.47 4.93 2.51 4.99 2.57 5.10 2.52 5.01 2.36 5.00 2.49 4.98 2.45 4.92 2.50 5.00 2.47 4.97 2.49 4.99 2.46 4.80 2.51 5.01 2.52 5.15 2.50 4.97 2.46 4 83 2.48 5.01 2.39 4.79 2.49 5.07 2.41 5.01 2.50 4 98 2.40 4.89 2.47 4.99 2.43 4.91 2 49 5.07 2.44 4.99 2.50 5.00 2.39 4.90 2.49 5.00 2.49 5.00 The mean, 5.D. and C.V. for each level of control on each analyzer are Current Level 1 Analyzer Level 2 Competing Level 1 Competing Analyzer Level 2 Mean: 2.5 - Mean: 5.002 - Mean: 2.45 - Mean: 4.94 50: 0.017 - SD: 0.019 - 5D: 0.05 - SD: 0.095 CV: 0.68 - CV: 0.38 - CV: 2.041 - CV: 1.923 How do you measure accuracy? Precision? (MVS) (2) which analyzer appears to be most accurate? Explain why. (MV6) (2) Which analyzer appears to be more precise? Explain. (MVB) (2] Does either analyzer appear to be more accurate or precise at level I (normal) or level 2 (elevated] ranges of measure? If so, how did you determine this? (MVS) (2)026 7.30 6.11 027 2.64 1.08 028 4.93 4.26 029 3.77 1.4 Comparison Study DED 7.78 6.83 031 6.24 4.10 032 The competing vendor has been generous enough to provide you with analyzers to run your 8.22 7.63 comparative study. Following is a table with the replicate data that was generated after running 033 4.87 4.13 40 samples from a population of normal outpatient as well as psychiatric unit admits over a 7- 034 2.01 0.76 day period, on your current analyzer and the competing analyzer. This and additional information 035 0.98 0.50 will provide you with data to decide whether the test competitor results are sufficiently similar 036 10.6 7.51 to the reference analyzer to support moving to the competitor. 037 128 10.6 038 11.6 8.31 039 9.74 7.6 Patient # epicatechin levels epicatechin levels 040 3.22 2.00 Current Instrument #1 Competitor Instrument #2 (ng/ml) "Reference" (ng/ml) "Test Competitor" 001 |3.4: 3.18 Be careful to place your data in the appropriate columns, test is the new instrument, Comp is the 002 5.0 4.0 current reference. 003 4.74 1.4 This analysis will provide you will a series of statistical results, each of which will provide 004 7.03 6.02 information about how the two methods compare as well as any error that may be present in 005 5.17 5.01 the competing analyzer. The paired data calculator will provide you with: 006 4.97 3.96 Comparison plat 007 4.22 3.07 Slope of the best-fit line 3.91 3.14 Y-intercept 009 4.0 3.56 So standard deviation of points about the best-fit line; we will not be using this 010 5.76 3.63 calculation 011 6.0 4.18 " : Pearson correlation coefficient 012 6.29 .7 Mean: average of results of the test method 013 2.41 2.01 Bias: average difference between the two methods 014 1.90 1.80 Space same as Seed standard deviation of points about the best-fit line 015 12.27 9.94 T-value: indicator of whether enough paired sample measurements have been collect to 016 8.88 6.93 know whether the observed bias is real or statistically significant; we won't be using this 017 3.36 2.84 calculation here. 018 1.16 0.51 019 2.05 1.48 1. What general observations can you make about the best-fit line generated in your 020 4.7 3.05 comparison plot? Feel free to copy and insert your comparison plot. (MV2) (1) 021 3.97 3.60 022 2.88 2.16 023 2.97 2.33 2. What does the slope tell you about what type of error might be present in the new 024 5.01 3.54 instrument? [MV2,3) (1) 025 6.52 5.02 2 3. What does the y-intercept of your line tell you about how the two methods compare to each other? (MV2) (2) 4. What would the ideal y-intercept be? Explain. (MV3) (2)Test Method Slope 0.8149 y-intercept -0.1861 S 0.6055 0.9712 on Plotter mparison Plot ta Calculator Test Method Mean 4.2037 Bias -1.1833 SD diff 0.8151 t-value -9.1815Reference Range Verification you need to determine if the reference range for serum epicatechin levels will be similar enough to each other. This can be a significant issue if it involves notification of healthcare providers of changing reference ranges. They generally don't like that to happen. Your current analyzer has a reference mean of 2.5 and range of 2.35-2.65 ng/dl (mean +/- 2 5.D.), which is what you have previously verified and currently use in your laboratory. Thirty samples from healthy blood donors were assayed using the competing analyzer and the following results were obtained: 2.28 ng/dl 2.36 2.48 2.39 2.18 2.73 2.22 2.50 2.25 2.67 2.48 2.30 2.27 2.40 2.39 2.41 2.44 2.33 2.32 2.67 2.44 2.40 2.58 2.83 2.60 2.41 2.12 2.24 2.30 2.35 Mean: 2.411 -SD: 0.167 - CV: 6.927 18. Ideally, what is the optimum number of samples that should be used to verify a reference range? How are samples collected and run during this evaluation? (MV10) (3) 20. What would the calculated mean and reference range be on the competing analyzer? (MV9) (1) 21. Is the competitor's mean similar enough to the current analyzer that you would not need to make changes to your reference range? How do you know if the two means are significantly different? (Apply the 2-tailed Student t-test to this question). (MV7, 8) (4)