CA2421571A1 - System and method for determining clinical equivalence of test methods - Google Patents

Abstract

A system, method and computer readable medium of instructions are provided which are useful for determining whether an evaluation testing method or device is clinically equivalent to a reference testing method or device. The report can include a modified mean difference plot, a variability chart, confidence intervals for bias and a plot of the intervals. A graphical user interface is provided to allow data associated with the reference and evaluation methods or devices to be identified. A
level of variance in the reference method is determined. An observed bias between the evaluation methods or devices and the reference methods or devices is calculated. A
confidence interval for the bias is calculated. The biases between the evaluation methods or devices and reference method or device is compared relative to the level of variance in the reference method, and a report is generated including a conclusion about whether the evaluation methods or devices are clinically equivalent to the reference method or device.

Description

.! CA 02421571 2003-03-10 ar Patent Application for System and Method for Determining ~'linical Equivalence of Test Methods by Scott Pardo Karen Byron Tralerie Bush Julie Berube Shankang Qu and Henry Davis Field of the Invention [~O~1] The present invention relates to a system arid method for evaluating testing methods. In particular, the present invention relates. to a system and method fox comparing two or more testing methods,, systems, or products. The invention is particularly useful in the medical testing 1 field, to determine if the methods, and any results obtained therefrom, are clinically equivalent. However, the invention could easily be applied to any laboratory situation in which measurements made under different conditions are compared in order to determine whether the different conditions alter the results in a significant way.

Background of the Invention [0002] Clinical laboratories perform tests for doctors and healthcare professionals.
The laboratories perform tests on human blood, urine, plasma, serum or other body fluids in order to measure chemical or physical properties of the specimens.
The results of these tests are used by doctors and healthca:re professionals to make clinical decisions related to patient care and treatment. Be<;ause results are used to rn_ ake clinical decisions for patient care, dependable test results are of the utmost importance.

[0003) Clinical laboratories purchase supplies and products in order to perform these tests. For example, blood collection tubes, needles, diagnostic instruments, chemical reagents and other supplies are used during testing, and therefore must be periodically replenished. From time to time, some element of a testing procedure may change for a variety of reasons. For example, a new blood collection tube type may replace an older version, new blood collection tubes may include a new additive, or a new blood collection tube could be made of plastic rather than glass. Chemical reagents may be ordered tom a different supplier, or even a new batch of reagents could be considered a change in the testing procedure. Furthermore, the diagnostic instruments used to perform the testing themselves rr.~ay change. Newer models may replace older testing equipment. Also, hardware, software and firmware updates may be applied to the equipment.

[0004] Of course, the above-described list of variables in testing procedures is merely exemplary, and the list of possible variables is endless. It is important to recognize, however, that any change in testing procedure can potentially affect test results. Therefore, because the accuracy of test results is so important, there is a need for a way to gather and analyze empirical data to ,chow that the testing procedure using the new method, device or system does not signif cantly affect the testing results.
(0005p There is certain degree of variability i:n any testing procedure. By analyzing test data, it is possible to measure the variability in test results. In addition, a new test procedure or method may give results that are on the average different from a "reference" test procedure. This average difference is called bias. If the bias between a new test procedure and a reference method is small enough, and the variability ir~ the results using the new procedure is no greater than the variability of the old test procedure, the new test procedure can be considered clinically equivalent to the old test procedure, There is presently specialty software on the market for evaluating and validating testing methods. However, the e~cisting software products fall short in several respects.
[0006 Currently, most if not all clinical laboratories rely on a statistical technique called linear regression to compare testing methods, systems or products. The linear regression analysis is almost always accompanied by ;~ graphical representation called a scatter diagram. In a scatter diagram, the results from one method, system or product are plotted against the results from the "reference" method or system on a chart and linear regression analysis is used to determine a best-fit line on the chart to represent the data points. A perfect result on a scatter diagram would be a line having a slope of one and a vertical axis intercept of zero. Unfortunately, the degree to which the best-fit line fits the observed data depends ~on the number and frequency distribution of data values used. Therefore, the quGality of the best-fit line for its accuracy and usefulness may be manipulated by selecting individuals at either end of some analytic spectrum and including their results. in the data. Thus, while scatter diagrams and linear regressions may be helpful in determining the similarity of results between a reference and evaluation method, system or product, they are not sufficient.
~O~O?1 A commonly used quantity calculated b:y existing software pack-ages is called R2, sometimes referred to as the coefficient c~f determination. R2 can have a value between 0 and l, and represents the degree to which a straight line fits the data, relative to the total variability observed. A value of 1 indicates that all the points fit exactly on the same line. Often, R2 is seen as a measure of equivalence between the reference and evaluation methods, systems or products. Unfortunately, R2 is susceptible to a priori manipulation. For example, suppose two tests designed to measure cholesterol values in human bload are to be compared. Some patients may have very high cholesterol values while others may have very low cholesterol values.
If two methods for measuring cholesterol are being compared using a linear regression best-fit line, then a high value of RZ may be falsely interpreted as indicating equivalence of the two methods. In fact, the high value of RZ is may only be indicating that there are patients included in the study whose cholesterol values are at the high and low ends of the human spectrum. because RZ is susceptible to manipulation, it is not a good quantity to be dependedL upon for measuring the clinical equivalence of a new test method.

c~~

[0008] Still another disadvantage of current test validation methods, is that they typically validate only a single test method at a time. Thus, for a testing device which is capable of testing 30 separate analytes, using previous testing validation methods 30 separate validations will have to be performed. Accordingly, it would be advantageous to have a single sofrivare package which could validate all 30 testing methods at one time.
[0009] Therefore, there is a need for a test method validation system which reliably measures the accuracy and precision of a new testing method, determines whether the new testing method is clinically equivalent to a previous testing method, and is capable of validating a plurality of test methods at one time.
Summar~~ of the Invention (0010] The above needs are addressed and other advantages are realized in a method and system according to an embodiment of the present invention. The method according to an embodiment of the present invention comprises the steps of determining a level of variance in a reference method, determining the average difference in results of the reference and evaluation testing methods, and comparing the average difference between the two methods relative to the level of variability of the reference, and based on the comparison, generating a report indicating whether the evaluation testing method is clinically equivalent to the reference method.
The acceptable difference between results of the reference and evaluation testing methods can be calculated by comparing two sets of reference test data associated with the reference method. Alternatively, the acceptable difference can be defined by the user:

The report comprises a plot of confidence intervals for bias, modified mean difference plot, a variability chart having a first axis representing accuracy and a second axis 'representing precision, a scatter diagram with best-fit regression line and associated statistics, as well as a conclusion as to whether th.e evaluation testing method is clinically equivalent to the reference method. The report also contains a summary statistics table, a table of the bias confidence intervals., the limits on allowable bias as optionally input by the user, and an Appendix containing all the input data.
Finally, in the method according to an embodiment of the present invention, data associated with the reference testing method and the evaluation testing method can be conveniently identified by the user through a graphical user interface.
Brief Description of the Drawings [0411] The invention will be better understood with reference to the following description and the attached drawings, in v~rhich:
[0012] Fig. 1 is a block diagram of a system according to an embodiment of the present invention;
[0013] Fig. 2 is a screen shot representing a graphical user interface used to identify data in a spreadsheet representing reference and evaluation data in order to determine equivalence of testing methods;
[0014] Fig. 3 is a flowchart illustrating a method according to an embodiment of the present invention;
[0015] Fig. 4 is an optional output of the system, showing confidence intervals for two evaluation test methods;

_7_ [0016] Fig. 5 is an optional output of the system, showing a mean difference plot for two evaluation test methods;
[001?] Figure 6 is a variability chart generated b~y the system, showing accuracy of an evaluated test method on the horizontal axis and precision of the evaluated test method on the vertical axis;
[0018] Figures 7 and g are correlation plots generated by a system according to an embodiment of the invention, illustrating the ro~etst,'_on hP~~zree;~
observations obtained using the evaluated test method and observations , obtained using the reference test method.
[0013] In the drawing figures, it will be understood that like numerals refer to like structures and method steps.
Detailed Descriution of the Invention [0020] A block diagram of an exemplary system 100 in accordance with an embodiment of the present invention is shown in Figure 1.. The system preferably includes a memory 102 for storing test data. The mc;mory 102 is accessible from a processor 104. Processor 104 receives inputs from input devices 106, such as a keyboard and mouse. Processor 104 also produces outputs which are displayed on an output device 108, which can be a monitor or printer, for instance. The memory preferably stores test data to be evaluated, but may store other information, such as program instructions for executing a program in accordance with an embodiment of the present invention. Data can be entered into memory 102 through user input devices 106, or alternatively, optional Iab equipment 110 can automatically store test results.
[0021] Processor 104 executes a set of machine instructions adapted to evaluate the test data' stored in memory 102. The test data relates to test results obtained using various testing methods on a common set of donors, as will be explained in greater detail below. The data includes at least one and preferably two results per subject using the reference (control) methods, and at least one result for each Pvah.:ation method. Processor 104 is adapted to perform a series of calculations which determine if the evaluation test methods are clinically equivalent to the reference method or methods. The calculations and steps performed by the processor to make this determination will 'oe described in greater detail below.
(0022] An exemplary set of data associated with two reference test method results and one result from each of two evaluation test methods per subject is reproduced in Appendix A. These data, as shown, are preferably stored u~ a spreadsheet program, such as Microsoft~ Excel. As shown, the data are stored in cells of a table identified by columns and rows. Rows ~-4 of the exemplary table contain information about the test, including the test name, the units appropriate to tike results of each test, and user-defined limits of equivalence (acceptable bias) for each test. As shown the limits can be expressed in exact quantities, such as 2 mmol/L for Sodium, or in percentages, such as 10% for AST.
[0023] As further shown in the table of Appendix A, Row 6 contains labels for each of the columns of data in rows 7 and above. Column A contains donor numbers, column B contains the main variable in the testing methods .(the blood collection tube type), Column C contains results of the tests for Sodium, Column D captains results of the tests for AST, and Column E contains results of the tests for Triglycerides.
There were three types of blood collection tubes used :in this study, Serum, SSTT"", and SST IIT"° As can further be observed from the exemplary table of Appendix A, specimens from 30 donors were tested, and each donor was tested for three analytes, Sodium, AST, and Triglycerides. For each donor, four blood specimens were drawn, tw'o wit h the Seitiiia type i~i'OG, arid une each wrth tile SSTT~" and SS j, Il'"" tubes, wlth each of the three analytes being measured in each specimen. Two specimens were drawn with the Serum tube, which in this case was considered to be the reference or control method. One specimen was drawn with each of the two evaluation devices.
'Thus there were twelve results (4 for each analyte) for each donor.
[0024] Tube type is the main variable in the exemplary test methods, but it should be understood that any variable could be evaluated, and blood collection tube type is chosen and discussed herein simply as an example. The serum tube was the reference or control device. The,first evaluation device in this example was a blood collection tube labeled SSTT"", and the second evaluation device was a blood collection tube labeled SST IIT""
[0025] The user interface will now be described in connection with Figure 2, which is a screen shot of a user interface according to an embodiment of the invention. In the preferred embodiment of the invention, a graphical computer interface such as the one shown in Figure 2 is provided. The invention is embodied in a computer program which acts as a plug-in to Microsoft~ Excel. Of course it will be understood by those of skill in the art that the invention could be programmed as an independent software application running on a personal computer, or embedded in hardware, or implemented in any other suitable manner. When the plug-in is activated, the user interface 200 shown in Figure 2 is presented.
[0026] The user interface 200 allows for the user to identify parts of a table, such as the spread sheet shown in Appendix A, which are xelated to reference and eqal»att'_nn tect t'~~eth~~g~ aid t~ ~hn~',3~~a, ~~a,~.~,,ri ,.~:r.~,18,13iv°. L'pti__ ~vns yr th'~. typ2 Of evaluation to be performed, as well as the types of outputs desired. The user then uses a mouse or other suitable input device to identify the corresponding portions of the table which contain the information needed by the program to perform the necessary calculations and generate the desired output.
[002?] For example, a portion of the user interface 200 is labeled "Study Information" 202. This portion includes Experiment Name 2.04, Analyte Names 206, and Analyte Units 208. The user has the option of typing the cell range corresponding to "Experiment Name" directly into the space provided for in the user interface at 204, or to click a button 210 allowing the user to use a mouse to identify the corresponding cell range within the Excel worksheet. Since the Experiment Name in this example is "Anaplot Vest" at cell A6 of the table in Appendix A, cell A6 would be identified by the user in field 204 of the user interface. Similarly, cells would be identified as corresponding to the "AnalS~te Names" at 206 of the user interface 200. Cells C4-E4 would be identified as corresponding to "Analyte Units"
208 in the user interface 200.

[0028] A type of Mean Difference Limit Calculation i.s selected using the user interface 200 at 212. The choices are Replicated Control Calculation 214, Bland Altman 216, Given Variability 218, and No Control Limits 220. ~nly one of the four selection can be selected. Also, a choice between Constant CV 222 and Constant SD
224 is provided in this section 212. The types of Mean Difference Limit Calculations will be discussed in further detail below.
[0029] A portion of the user interface 200 is provided to allow for the selectio!~ of desired outputs 226. The possible selections preferably include Confidence Limits for Bias 228, Mean Difference Plot 230, Chevron Plot 232, Correlation Plot 234, and Data in Appendix 236. A checkbox for each type of output to be included is provided, and selecting any of the output types will cause the output to be included in the report generated by the system. The Clinical Criteria for Bias Limits 238 can also be set, either by entering the criteria directly in the space provided, or by referring to cells in a table which contain the clinical criteria for bias limits, such as an Excel worksheet.
[0030] A section of the interface 200 is provided for identifying certain relevant data 240. The data identified in this section includes a Donor ~ Column 242, a ContIEval ID Column 244, and a Data Range 246. In the present example, Donor ID
Column would refer to column A of the table reproduced in. Appendix A. This is the column of data containing donor IDs. Cont/Eval ID Column 244 refers to the column in the table which contains the names of the reference and evaluation variables for each donor. In this example, column B of the table in Appendix A would be identified. Column B contains the labels for the blood collection tubes used in each test (Serum, SSTTM, and SST IIT~). The data to be evaluated, including reference data and evaluation data as appropriate, are identified in the Data Range 246 field. In this example, columns C, D, and E are identified as corresponding to the test results for both the reference and evaluation tests. These columns contain the actual test data for the three analytes tested, and for each of the 30 donors. The interface 200 also includes a field for Control ID 248 and Evaluation ID 250. A, "Select All but Control"
button 252 is provided. Finally, an "OK" button 254, a "cancel" button 256, an "add COmparlSOn" button _ _2~~ and a "restt~r~ pnpr v~l_pPC" h~attr,_n_ 26Q are prr~~Ilr~ed.
[0031 The method according to an embodiment of the present invention will now be described in connection with the flowchart of Fi;,u~ure 3. At step 300 a reference method is conducted. Observations from the reference method are recorded at 302.
The reference method forms the basis for compariison to the evaluation method.
Preferably, the reference method is performed at least twice, and observations of both reference methods are recorded. In this manner, the variability between successive runs o~ the same method can be measured. At step 304, the evaluation method is performed, and observations are recorded at 306. Preferably, the observations are recorded into a table, such as a lVlicrosoft~ Excel worksheet, to facilitate accessing the data fox calculations to be performed by the statistical analysis program.
More than one evaluation method may be performed and recorded. Advantageously, according to an embodiment of the present invention, any number of evaluation methods can be evaluated simultaneously.
[0032] At 308 the statistical analysis program is started. Preferably, this produces an interface as described above in connection with Figure 2. Various data are identified in the user interface 200 at 310. Preferable, the data identified in the user interface 200 include the donor 1D's associated with 'the data, the Control/$valuation IDs, and the columns of data for the tests perforrried. A sample table of data is provided at Appendix A.
[0033] Also in the interface 200, the types of mean difference limit calculations desired are selected 3i2. The types available are Replicated Control Calculation 214, Bland Altman 216, Given Variability 218, and No Control Limits 220. Also to be selected are constant CV 222 or constant ~D 224. T_f 1_z_ept;cated Control Caoulation 214 is selected, the statistical program calculated the acceptable variability in the evaluation data based on the variability between the at least two sets of reference data.
Bland Altman 216 selects a Bland Altman mean difference calculation. Given Variability 218 allows the user to select the acceptable variability. Finally, No Control Limits 220 allows the user to select a set of calculations without control limits.
(0034] At step 314, the user selects the desired set of outputs to be generated.
These selections are available at 226 of the user interface 200. The user's choices comprise Confidence Limits for Bias 228, lvlean Differencf; Plot 230, Chevron Plot 232, Correlation Plot 234, and Data in Appendix 236. Examples of each type of data will be described in greater detail below.
[0035] Once all data have been identified, and calculations and outputs have been selected, the user selects the "OK" button 254 at step 316 to begin the calculations selected.
(0036] A series of equations appropriate to the various selections available to the user are shown at Appendices B and C. Appendix B shows the set of equations associated with determining the slope and intercept in a correlation plot 234.
Different equations axe provided for different combinations of calculation type, and the kind and number of reference and evaluation data sets, as well as the type of variation selected. Appendix C shows the set of equations used to generate Chevron Plot data.
The Chevron Plot will be described in greater detail below in connection with Figure 4.
[003?] At step 318, the system determines based on statistical analysis, whether the evaluation data indicates that the evaluation method is clinically equivalent to the reference method or methods. Finally, at step 320 the selected outputs are generated, along with conclusions reporting whether the evaluation method is clinically equivalent or not.
[003] Various outputs will now be described. The outputs described were based on the sample data provided in the table of Appendix A. A complete sample report is reproduced in Appendix D', and this report includes each of the types of outputs to be described in the foregoing description, for each of the three analytes tested in the reference and evaluation methods shown in Appendix A. For brevity, the outputs will each be described once in connection with one of the three analytes, AST.
[0039] Figure 4 illustrates the Confidence Limivts for Bias output, selected by checking Confidence Limits for Bias 228 in the user interface 200. The output shown in Figure 4 corresponds to the analyte AST which was tested for each donor and for each reference and evaluation test method. The 95% confidence interval for bias gives a feasible range of possible values for the average bias or difference between results obtained using a reference method or device and an evaluation method or device.
Thus, if the 95% confidence interval for bias in AST between SSTTM and serum tubes is (5%, 8%), then there is 95% confidence that the true difference is somewhere between 5% and 8%. The confidence interval for each of the evaluation methods, SSTTM and SST IITM, are shown to be well within the 10% limits designated, indicating equivalence between the evaluation and reference devices.
[0040 Figure 5 illustrates a mean difference plot generated by the program according to an embodiment of the present invention. Data for each of the evaluation methods, SSTTM and SST IITM, are plor_r_ed. Fact, point represents a differe nce bet::Teer~
the result observed using the reference method and the result observed using the evaluation method.
[0041] Figure 6 illustrates a Chevron Plot generated by t:he program according to an embodiment of the present invention. The Chevron Plot is a measure of bias (accuracy) and precision. Each evaluation experiment is plotted. Evaluation methods with a combination of good accuracy, and good precision a~~e preferred.
Regions are deSlgnated as "Crood", "Satisfactory", "Unsatisfactory" and ~~PoOr" SO that the user can easily see which classification applies to each of the evaluation methods.
Of course, it will be understood that while the Chevron Plot is the preferred manner of presenting accuracy and precision data, any graphical or non-graphical method of presenting accuracy and precision data is considered to be within the scope of the present invention.
[0042]' Figures 7 and 8 illustrate correlation plots generated according to an embodiment of the present invention. Figure 7 correlates reference (Serum) results with the first evaluation method (SSTTM). Figure 8 correlates reference results with the second evaluation method (SST IITM). Regression is performed on the data and a regression line is plotted. An ideal line with slope equal to l and intercept equal to zero is also produced for comparison.
[0043 A sample report generated by the system according to an embodiment of the invention is reproduced in Appendix D. The report includes the various outputs selected in the user interface 200 as described above for each analyte tested.
Also, the report includes conclusions about the clinical equivalence of the evaluation methods for each of the analytes evaluated. 1n this mamn_e_r, _n_ew _test methods (including existing test methods with new components, such as blood collection tubes, chemical reagents or analytical instruments), can be evaluated, and a lab can quickly and definitively determine that test results using the new method are clinically equivalent to previous test results. If the new method is shown not to be clinically equivalent, steps can be taken to correct the problem.
[0044] While the invention has been described by means of specific embodiments and applications, numerous modifications or variations could be made thereto by those skilled in the art without departing from the scope of the invention as set forth in the appended claims and equivalents thereof.

-16b-p ~ B C D

~1 2 Stud Name:

3 Anaplot Limits '2 10% 10%
Test (+l-):

4 ~ ~ t9nits:mmoUL_ UIL mg/dL

6 Donor Tutae Sodium AST Tri 7 1 ~ Serum 137 18 183 8 1 Serum 137 17 182 9 .1 SST 138 19 180 10 1 SST ! 137 17 184 I

11 __ Serum 140 12 78 2 .

12 2 Serum 140_ 12 79 13 _ SST ' 138 13 81 14 2 SST II 140__ 13 80 15 3 Serum 141 _ .42 126 _ 16 _ Serum 140 41 . 126 __ a ~ 3 ~

18 __ SST II 140 42 125 19 4 Serum 142 ~ 18 133 20 _ Serum 142 17 136 22 __ SST I 142 17 136_ ___ __ I
~ 4 ~

23 5 Serum 139 16 101 24 5 Serum 138 16 _102 25 _ SST ~ 138 ~ 17 '100 _ 5.

26 . 5 I l ~ 139 6 _ 101 ' ~ST 1 27 _ _ _ 13~ _ 54 28'_ ~ Serum ! 140 _ ; 57 -- i Serum 18 ~
6~ ~_~ ; 19 6 ~

29 ___6____~ SST ~ . 139 18 55 30 6 SST lllf __ _ 57 j I 139 ~ 18 ~

31 7 J 140 25 _ 126 Serums 32 __ __ ~ Serum 140 24 ' ~. 125 33 ~ 7 SST 140_ 24 - ~ 123_ 34 _ _ SST II ~ _ _24 125 35 ___ Serum 140_ 18 204 ~ 7-_ ~ ~ 139 8 ~~

36 _ Serum ! 13_9 18 204 ~8 37 _ SST I 139 18 ' 202 __ 38 8 SS.T..11 _ 139 _ 17 205 , 39 9 Serum 136 ~ 20 150 40 9 -- ,Serum _136 21 ~ 148 41 9 '' SST 136 ~ 21 152 42 9 ASST II 136 22. 153 _ 43 10 Serum 14_0 _17 97 ~ __ __ 44 _ '_ __ ~ 17 _ 96 45 -_ ~_ Serum _ 141 16 _96 1~0 _ ' _ ~_. ~-10 ~ SST_ 14_1 _~.__.. . _ i - .__ 46 ~ 0 i SST 141 20 95 II

47 11 Serum _13_9 19 _ 84 ~ ~ ~
~

48 _ _ 141_ 20 ~ ~, 84 __ erum ~ _.
_ _ ;
~ 11 S
-49 _ _ _ 142 __20 84 50 ~11 ~~ _ ' 14_1 19 6 5 ____11 SST_ ~ 140 21 ~ 159 " ~ SS_T II~ ~
1 12 ~ Serum -16c-A B C D E
~

52 12 Serum 140 20 15fi 53 12 ~ _ 142 21 161 SST

I

_ _ _ 139 J 18 64 55 13 Serum ' _ Serum 140 14 63 57 13 SST ~ 141 15 62 59 14 Serum 142 18 181 60 14 Serum 140 18 ~ 182 61 ~ _ 14 SST 141 19 189 62 14 _ SST II 141 18 178 63 15 _ Serum 143 20 163 64 __15 Serum _ 21 164 65 _15__~ SST 143 20 165 67 16 _ ~ Serum 140 20. 330 - I I

68 Serum _ _ 332 _ SST 141 21 329 70 16 SST i 141 21 333 I

71 _ 17 ' Serum 140 26 199 72 Serum 149 25 201 ' __ _ ~ 140 __ 200 _ SST ' 26 ' _ __ 73 17 _u y ~

__ _ _ 24 J 164 _ ___ ~ Serum 141 76 _-_ 18 Serum 141 24 ~ 163 ,., V

77 _ ~SST~ 140 23 162 _18 _, 78 18 j 141 _ 163 SST i1 24 ' _ _ 141 20 ~ 70 79 Serum ~ ~

_ ;__ Serum139 _ 69 _ ~ 20 80 ,19 V
~

81 __ 19 SS ! 19 ~ 64 __ _ T 140 ' _ -82 19 _ _ j _ f 140 T 19 68 _ SST II

83 _ 20 I Serum ~ 143 19 _ 105 84 20 ~ _ erum _ 20 106 _ _ 9 _ 105 _ SST II 14 ' 20 104 86 ~ _20 2 ~
~

21 Serum ~ 140 88 _21 Serum ~ 139 18 93 89 21 - SST_ 139 18 ' 93 90 ' ~ ~ ~

__.21 _ SST II 139 94 .

91 22 Serum 140 22 152 92 __22 Serum 140 22 ~ 152 93 22 SST _ 21 __153_ 94 SST ~ 141 21 ~~ 158 _22 __~ 1l 139 _~
~

_ __ _ _ 130 95 __ 142 _ _23 ' Serum 12 96 _ 23 __ Serum ~ ~ 13 133 ~ 141 97 23 _ 140 13 131 SST

98 23 _ _ 13 133 _.~ _._ SST 11 139 ~
__ _.

- __...___._..r__-_ __ 99 24 . 140 I 14 ~ 595 Serum 100 24 _ I 140 15 T 594 ~ Serum ~

101 24 _ , ~ 14 585 102 24 SST I 142 ~ 15 597 I

-16d-A B C D E
10325 __ Serum 141 _ 16 _120 1 25 Serum 142 16 121 _ __25 __ _ SST 144 1g _ _117 107_ 26 Serum 142 _ 31 331 10826 Serum 142 31 337 10926 SST 143 _ 3 110__26 ~ SST II 141 31 338 11127 ~ Serum 142 _ 32 87 11227 !T _ 142 ~ 20 86 Serum ~ 22 11327 SST 143 19 ~ 87 114_ SST II 141 20 87 115_ Serum 143 _ 171 28 . _ 11628 . Serum 143 21 170 117_ SST 143 20 167 _ 28 !
~

~

119_ Serum 139 39 230 .29 ~

12029 Serum 141 37 233 121. 29 SST 1 37 230 12229 SST II _ _ .39 235 _ 123_ erum 144 _ 15.5 30 ~ 19 12430 _ 144 19 ~ 158 Serum 125__,- __ __ 144 20 _ _ 126- 30 _ ~ SST 141 20 ~ 150 _ __ 152 30 ~ SST II

' ~ '~

-16f Evaluation Variation Limvit Interce t, Sio a .

mil Control.
t_i Replicated . single single constant sd Control Replicated Single single constant cv Control Replicated single paired constant sd Control Replicated single paired constant cv : 2 Control t sd ,0 t t zAve (c - c1)2 +(c -c2 ~ ) Replicated fired an _.
Pa singie cons Control - -)2 '~'(C~C2~2~1C 2 3 Ave[~(c-c O

Replicated aired single constant cv l ~ ,t Control _ d t ZAve[(c - c~)2 +(c -c2 ) 2~~O

Replicated paired paired constant s y Control - -. )2 +(C-C2)Z~1C 2 3 Ave[{(c-~

licated fired constant cv , Rep paired pa O,t Control d ~ G~CZY'~C

Bland Attman S~ngtP i 1 COnstants t s.ng.e t 2 Vc~~ [c' Bland Aitman single single constant cv ' , onstant sd t ~ Vaa~ [ct-eI , 0 sin i e paired c Bland Altman J

.

l aired constant cv 0, t ~ Yap [ci-e]
p ' Bland Altman sang constant sd single t 2 Tar [c -el] , 0 Bland Altmari paired ~

'ed single constantcv- 0,r ~h~a~'[c-el~
n pair it ma Bland A

an paired paired constant sd t 1Y~'[c-e~~'0 It m Bland A

an paired paired constant cv 0, t ll~ax'[c-e~ 1 Alt m Bland Given single single constant sd t ~ gaveh,0 Variabilit ven ~ ~ g GNen single single constant cv ~

a Variabili stant sd t ~ giver~,0 Given single paired co Variabiti t~give~t Given single paired constant cv , Variabiti Given paired single constant sd ~ t 3 given, 0 Variability Given wired single 0, t ~ given constant cv Variability Given paired paired t ~ give~a, 0 constant sd Variability ' , ~:-- ~ ~ -16i- ~:.....~

'on Variation Chevron Score-, ,,,- .,, .
t~ ............:..:
:: v::::::;'."v?:!,;.!!:.?::
;:::::: ::..'k:i:?.:::'.,'.:.;
rol Evalua t "_:<-'-:<::?:::._ ~.. :. .".,~:.?:,'-::.,:h,::.,:.,.??k Con :(<v. :,:,z::-..::
::;....,....s...,."..
~ ~ ....~U-.\:.:..
I :.:,......,..:..:..
:.a,:y,,.
.>.
;T;_ ":.v$:i:':::~:i<i it. ......
. .. n:;:'.::-::!:.
.... ..w :v.::.....:;
.:.'~;'.:>v.
;:'>.;;:~~-:
>v::.~,~,:~
:;;t~!:::.s;:;:~:::
n{: ;':::i::!;:::;:.,:
~ :L??:::.t?:i'.-???:::.
~:::.sw: n..:.,;...:..
' tan .:z?.??,.:_:.::!.':.:?:;:;::::::<;;;;;;.':;;:.:",:.":::::::::.:....:::::......:
.:::::.:.,:.-.
:::.? :.?.;, tsd in 1e cons ..........,....
'.???e'??:.?:?!::,,..::;:
s :<>K;;::,;
: t:
g .?!:-?~:;::~:;
..:.. ...
::.::. .
1e .e-- .::
r.:E~,::~::::;
:, ...-' .
.,.
Sin .:: -:
:-: .:.,,..
.. t ::~::w.:.,:.
~,.:....,.,~:
...
ontrol 9 :::~!:<
:,:.:~:._ ..~.:,:,.,...:..;~..a,.
.
cated C ::_:~:..:"._~.~:,~,::,:.,."..:..;...:......:..
:..........:::.:.a..:::::.!,::?:::
::.::..::.:.:
Repl~ .:::;;;.!
~;;::::.!!:::::?.::::.:~,::.::;:.??>!:::;::>.?::.!...:;..

:..::.:::::.::.........;....
0 StantCV :;:::;:.::.:::
:: .:;.W.,,:~:..
:.. ...
c n ~:;::::_:::...
.. .' .<.:
_ .... ..
:..:...:.,:;:':,.::?:::::.::,:;,::~:::?,;:~:~!?~::::::~~:;;,:a=

sin 1e .:;:;!:'..::.::~:.?:w.::
1e 9 . ,.:c.
,. ,. :- ..::, .":::.:.! ?.
::,?.';?:':
,..,, l s n "a .T
:.-.::c::..:."
.
trop 9 ,::....
~c:'v:~n<c...
..,.., ...a...~.....
ated Con ';:'iii:v':.'.':a::~...,....uC.!'~i~.':'S..T.T\SC....

. .:~:.Z:::::::.?
Repl~c :=a!..,..
_. ~ :.:.;.:.:?:::
:.~.:,~.:.::::
:; v.?,:::..:
::..;?;;::::,.
: >.:-?..::...
..:: . !::!':
.::....: ...
:. ;:
' t nt sd :>::!::>!?
::_:::'? ~::::<:?:::::::.:;::,>.,;?
:.:.:::,.::....
....; .. -......:.
::. ,..,.-:.,?..
:<..::!?:-?....:
_. , ' red cons a :;..:.,--,..
.:... .........
.;,:::.:::::::::;:.:;.:.::.:.?:.!?'.!?::!:.::.._ ;.:::,.:?.>.::,:.?:;.:'':
1e pal .::.~.::-:
sin ,;';'~:a" s.
tr :''~ <::::
''43:t:: :
:;:T . .:,:..,.
O1 ~ ":vj;Gi'k:RV::
":;\l..v ed Con <!_w::,:,:..
. ~ ~'c:.......:
....;:w...~...............
a hcat :';::~;;~;;,;:~:~.:..~,.,~.:.::?....,:....-~............:.;:::.:.:....:.::.....,.....::..:...::::.',,'...:

R P- . .;~::::.::::.:;~:.::~.:::,:.,..,.!::-!!!:'!:::?!:;:'?-..:;.?:<;?:>;:::::::.~................
:...-...
Replicated Control single pavred constantcv :;..'. :;~.
..,..........
Replicated Control paired single constantsd !e1 ~ct~/
2Ave[(c--C~~2'~'CC"C2j2~
y2 Re licatedControl paired single constantcv lCt-C~~/ 2AVet~(C-C1~2-f-~C-C2~2~IC, lP

Replicated Control aired constantsd paired P g1-cl)l 2Ave[(c-ci)2+(C-C2~2 a Iicated Controlpaired pairedconstantcv Rp {ga -CJ~I
3 AVe[~(C-C1~2 +~C-CZ)~l Bland Altman single singleconstant Bland Altman single singlesd /A
Bland Altman sir'~gle pairedconstant Bland Altman single pairedcv /A
Bland Altman paired singleconsta"t Bland Altman paired singlesd /~
Bland Altman paired pairedconstant Bland Aitman paired pairedev /A
Given Variabilitysingle singleconstant Given Variabilitysingle singlesd /A
Given Variabil'dysingle pairedconstant Given Variabilitysingle pairedcv /l~
Given Variabilitypaired singleconstant Given Variabilitypaired singlesd /l~
Given Variabilitypaired pairedconstant - en Variabititpaired pairedcv /!~
constantsd Giv Y ~Bl -Cl)l~$ivel2 constant cv (et -C1} /
. (~Cl$ZVPYI~
in 1e constant l sd tBT
9 -Ct~ l ~ $IV~3~~
constant cv ~8j - Cl~ l ~~ $averi<) constant sd ~81 -C7~ l ~~ $lVBYd~
constant cv ~8i -CI~ l ~.~C$IVEYi~
constantsd (2= -C3)l~~$rVe~M
' constant cv ~8=
-C~.~ l ~'~ C$iVuF2) ............
.....:._:;;::.......t:~..........;..:.:..:,:::::,:.~::.:...
.. .. ..
.. .:.:':,:...:~
:.a:.:::
::':_ '::
:::':.
:::::: ::..:::
:::: ;'c.?
:::.::l:.....
_: y.\:..:.
:.r: 1:~,.::>...Zr '.-.t7:.'S..\:i:
.:..;::.w:...:::;;,::'.;:.....;i:.'..'.~::.::.!;:?-:;.:.-:::.::::~i::::-.;;::~:::....~
::v.: ~!;?:p:4::
i,?:~:.;:.?::::;;???;i?'?'.'.,..:ky:':;:'L?
"4:"::'t-?-0.?
iy'-:;:::::
o,:.?;y'wy~y>...
constantsd :y,;1,:!'.;:::'v::>',.:p:?qY;,l;::.;.,':.-.,,:.\.\,..y.y.r:.4.n:_ ::, ! ;:.:..
:
_ -~
-:
:
. ~
v;
:
"

r ILim>ts e ;
No Contro sing : ..:.
:
CY :
in ie constan .::::.?
1e 9 :
sin .. . .. .:..:. :
miks g :
No Control ' c: ::::.::
?:.: :.:.
a.~:,;.. ,.:~:.,::,~:-:',-.:,;.:,:.?:;,:;....<.:,::,;:..::?:~':":y>...,.>~>:.
:..:,a:.:. ; ,;:::,~.:.::.";.
' onstant .v.?..:..:.'~..:~'?:;-.::
asred sd : ..:::: ;.:.:.:..
.:.,.:::::::.:.::.:.<::::.~:!::,:,::>'!:
.?.::: :. .. ...::a,::<:?:.
1e P :
ot Lsm~ts sin 's:<::v'Y;~~::\:-''<::.::,-~i:?'=ri:':9.'~.::iOx:::a:-:?:i.?f...::-.5:..:vhv...n,.,n5.,.
NO COn 9 ':::~:::::::.a..?...~:......
, :v::,~:-:<
nt CV :_-?':~:..~:_:~.::x:.::~
ired cons ~.v .
sin 1e p :'..?,~::?'',;';:w!:<:;:.::;<>.::;>.~~:-."
mits 9 ....,:..:,,.~;..>..:
No Control '- -y:i.W:W:~,~: ~h::.:-?
... ...........:n~: ...T.
- :.t..:..:.... .. ....
:!!:.:::'.~'v:'~..::..::::~-0::
n 1e t Sd :~u:;::' ~::::::'.:~
ed 9 consta ::....::.': ,:: ~:w::
.~; :!:,'.: r1,.::?".
i Limits pan :!:;v::.?!;?'. '4:''v?;
No Coat S t:!?: Yi : ;iL_ '.:-. W;?:;:.; ..::.,.: n:,:
::.:w.~: :.y..:~:y..,.v.v.S...:e.
.:_.;.?,...: ,.....nv..,.
1e onstant .... .r\.a...
ed 9 cv .........,.,~,.........
it .....,~
s .
..............,..".~::.:::n.......;.,,.;.:::.::...::4'":.
lm ............... .........
of L a ..:: .:..; ............
NO Contr . .._. .. ..'... ..;.;
.. . .: ..:.. .. ..
.
aired ,.:',;.',_:.c~;.:.
~,.:',:;:_;,~.;t..,:.~..:,?:c:.:.:?.:sc Limits aired P xr:ax;:::
:;;::>.'.5:#:~:;:.;.:~::~.::..:..;
No Contr 1 P nstantsd >::..::. ....?...:..
:: ::.:...; :: ~::::::
:: .:.,.-::: ''2;: ;::!::'.-';;:''?:::?::.c:::;;:
w~ ~~:'::::.,ws:':.-:;:::_,:?::;.:h::~~'-' o Control Limitsaired paired ,.:...:~.?:w: ..::: .,....:,..
~a,:
nstant cv t.::....::....:.........
..
_ -.,,..:s:..;;:~:
,;.:;::: ~..,~.:,,?: .:.:,?<:
.::..-?.,.?: ,:,,~~.
::'~<<:.. ;a:;:,.:: :;:~::.:..::........,:
... .?.....
1':~:..w,..:'i.'.?:'~s:hn....'........,......
3 ........~..~.....
..t.::::;....:.''......:;:
:.~~::::::::..v:Y.:...., ..,v... :':::.':'?: ':::s:::.vv.
:: ::h:; ::- '::-3::':.:.'' :.;;;;;;::,v -:.<:: .. ,::.,,. .,::,v:
'::,':.:z:,v.K;:,'vi\:::;t-\si:\;::v;s i'C:si:: v=::'u: u:.;v.Qv.,.::L.:.
.. .. ..
:.?._T.:::.:.......,:..,:.,..
.. :<;.?!;.~:: >?:<:,?:.~;:';::.::,;;
':a:~:Y:~:.<~.y, _ . ~.:---.~.:c...,..~s,~v..
. x.v, :~,:,:~<:~~.--:}:::~:~::.,,:,..
;:.-..:.a.!.,.;.,..::.~?..~:.:..::
..:?:?. ... ... :a.
...:..: >;.?,.y;.;.,;:::::;;?:;;2s??:?
x:::'?:' ::;;': ::..'.:.
: :..;.::.: ~.'2:.. .:.~::.:;?:Z.:;i,:.,.
'<; ~~;::t::.:;:.'':.?.,~:,,:"-..:..
~~.?~..:.. t t '~
:... . . ,: ~:...:
.. :....:.. . :..;.~
.';:.,.., ..,;t.,;;<,.-.Yr.
:.~:;.:?:.::
.. ;, .
:,::<;. ,;-?',?i,,:;w:.
":::..:..,. ._ .w..
:,::<-:.. . .::,;;:w:::
..\....., ..T.. ....:;.!'..
::~i:'?:..,::b:::::'.'-:\..:;.:J.;::::~:"N:,.;.'.:.:;.:'.?;::::~::.?:::::.~.:::.:.:..::::r.:-.'~~
:::;ors?'.~~.;:!:,.!:?:!::a!?:y::y.:.:::;:.,.::
??_:.-?:'c?:s? ~:::go::?';.:';;:'<:;'.?.:..:.
::.r.;
?.:.?;;:~!::.?:.?: -.:c<...::
n..:c.. ::.a...~.;;.....:.::
:;.::-:~'".:,o-:
'~:'t-:...: ...'. ~:>y''~"'~:;.:?"':,.rt W . ~;
:~.'~::::.:_:. ,'.vt::$:._':::;::,:vc:'::,..-:.-::\:,:w'::...~:i?:;v...
..,~.;\;.,a,.:> :: .
?....y,.:.~,::;i'..:::::.
':: ...;~.;;;.:::::;.~:.::,:k::::~:.:.,...
'':::--5.: '.:w ;.?::2-?a?:;;::~<::;',::'t.Y~:!
sy' _;,;~ ~~,<.:":~'c':~:'::;.:.:,:.,~..-:~:,.::.:~>:x:-::.

<.:~ '.~_ ~
-16k-nal~t lest Study Background The study was conducted to compare SST.ar_d SST' II (the evaluation treatments) to Serum (the control treatment). The study had 30 subjects. Each subject had two controls and one of each evaluation. The Correlation, Bias, Mean Difference, and Zap Plots are attached for the following analytes: Sodium, AST, and Trig. The acceptable imprecision for the Mean Difference and Zap Plots was calculated from the control values.
ConCItiSlon The table summarizes the conclusion for each treatment and analyte.
Equivalence is based on the observed bias and the criteria given for the average bias. If the 95%
confidence limits for the average bias are within the criteria; then the evaluation is equivalent to the control.
Analyte Criteria SST SST I!
Conclusion Conclusion Sodium 2 mmol/i_ equivalente~~ui~ealent AST '!0% equivalent equivalent Trig 10% equivalent equivalent ~. ~ -161-Subject Data Summary The following information ~ Provides a means of checking adequacy of the ana.iytic ranges in the study ~ Provides a means of verifying that missing observations were roughhr uniform across treatments Serum SST SST II

Mean 140.47 140.63 140.37 __...__.._..__._.__._......._._____._..__..............._.___.._._..__.__._....
.__..._.._.._..._..._____.___..
..
_ _. ._...__....._.._..____.._...__..___...__..
Std Deviation 1_.73 .__ 1.5 --_. .._.__.__....... 1 __..._.. _.__.___ 1.9_7 .
___..._ _.___.._.

Sodium minimum 136_ ~36 _ ___.__._. __._.._..._._.____.____._____.._ 136_.____ __..__..._._...- _ - . ..
...

Max 144 . .______.__.___._.._.__.._.___ _ im_um_ .. 143 y. ___ .
_.__ _ .___ Missin Count 1' 0 _ _Mean 20_.67 '19.93 20.30 I
-m I _..._.....
w_~__ -_ ~-~~.~

Std Deviation 6.4~ ~.~ ___.__.6.6_.__.._..
_..___.____...___~..__....__._.._.___._.____.________.___.__.______._ ___._____.._ __ ______._.._..._.._._.._..._._....__..__.___.____..._.....
AST Minimum 12 13 13 ___.._._._.___.._.._..__..__.___.__......._._.__.___._______.._.___..__.__._ .
_ _..

. ..____._.___._____._____ Maximum _ 42 _. 42 ~ -.._ _ - ~.~~~ _____...

___.__ __ Missin Count ~, ~ __. _._ ___ _ __M_ean_ 16_1.4.5 160.23 ~ 1 ' _-._ _... 62.3 -.~ ~~ . 3 ~__ Std D_e_viation104.92 1 _ .....__~ .____.__.___~4.92 _ _--~.___ ____..__._.____....__.____105.56-..
..
..
.

Trig M_i_nimum 5~ 5_ ...._.._._____5 ~ ._._ .. __._...___..
~ ____ _..
~ .._..._ .-_ 595 5g5 597 ~.--_ _._. _..___ _Maxim_um_ ~

Count , _ _~
Mlss~n vcnciu~ion used ~I~ A~ar~age Sia~
The table summarizes the conclusion for each treatment and analy~te.
Equivalence is based on the bias and the criteria given for the average bias. If the 95% confidence limits for the bias are within the criteria, then the evaluation is equivalent to the control.
Anaiyte Criteria SST Bias SST S'>T II SST i1 Biias Bias 95% CI Conclusibn Bias 95% Conclusion CI

Sodium 2 mmoIIL _0.2, 0.5 equivalent -~0.4, 0.3 equivalent AST 10% t-2, 2 equivalent t~11 3 . equivalent Trig 10% -~ 10) equivalent 011 equivalent CA

a a t. J

i.

~xp Name: Sodium ~xp Name:
Sodiurrl 'tiff rannw P!nt M

_, 30 domf5 30 dM~rs e 2 SsNm ohxrvutiotw 2 9AUrn and obselVaNmo and 9 evaluation 9 aValuatbn ohsortatian obaervat:m ~ d"~

rxar damn ' C.Oltib! Eaxd CV =C.'36% ConhW
8axd GV.
C.536%

~ ~ C=ood AB%
cap mdua V,VITIh evaluateodad I:mYe t~ ~ Satis~Y

0 7 ~ ~a ~' C~ ~y,ti, s~vaoW
um~

i Urratisfxrory a......................,......g."~.......~~.,...~.~..~..~..~...................
..........i~ 751595%dproduca ~Mlvln c~rai atwclsd IIm2a ~~~ ~ ~ssr p~d~
''~ Csrii vith~lnwaua7 ~ ~ ~ lsoelodlimks.

_q ~ v p ~ ~ ~ .a9aCJr -z- - ~ ssrn ,_ f ! . .
~, ~~a ms ~aa tas ~so 8ias Average Yd=val .
Control) .

e~ /0 4cnriaencav lnterv~!I~ ror o!ds Sodiuttt 2 mg7m4 i o .......................:........T..............................................
.........T................................~

>_ ~ 1 1 m° r -2 mglm!
SST-Serum . SST II-Serum - ' . .

-1 ~~-~

Exp Name; Sod;t<m Enp Name; Sodif~rn IS4 ~

, Sfady InfOrmafion:

30 donors 2 Serum abser~a6ons per donor t SST observation per donor 14z ~

1 missing Serum observation SST . ~ ~ ~ OmissingSST abservations - (evaluation) ~ . 30 comparisons - 90 expected observations 940 ~ ~ ~ 89 actual observafions ~ m ~ Regressionlnfortrafion:

' Intercept=-2.08 (-37.3, 33.1) 138 ~~ . ' ra ~ . Slope =1.02 (0.765.1.27) 1.
R squared = 7t-1M

w--- RegresslonLhe --- ieoci una Simultaneous Test of 6ntercept and Slope:

p-value ~= 65.2 k 136 - - AccepE Ideal Line Hypothesis t4%

Serum .. (controp xp Name: Sc:dium Exp Name: 5odiurn c~-a::~,-;=s a 144 . .
, j .

Shady tnformatlon:

30 donors 2 Serum absrvations perdanor 1 SST I! observation pet donor 942 o ~
~

. 9 missing Serum abservaflan SST it 0 missing SST
~ '~ ~' ' ' d! observations u!

(evaluation)' 30 comparisons -' ' ~ 90 expected observations 140 ~ ~ 89 actual obse.
rations Regression Information:

Intercept--. 36.5 X6.77, 66.2) t38 Slope=0.74 f,0.a23,0,951) R squared = 64..7 /o '-'--- Regeession Line ' Simultaneous Test ' ' -' meat une of Intercept ! and Slope: , 136 ~ ' - ~ p-Yelue = 5.16 Accept Ideal Line Hypothesis '138 938 140 942 ' Serum (control) Exp Nama: ~,ST ~xp dame: AST

0 ~~ 30 dcrrors 2Se-~ma6senation:and.. ~Sanimotrxe;vetionsatM
! nvswation 7 oSservation , evawzUon observaton g . pardawr ..

. ~rd~~

CanvdBa;adCV=d,q~q5 ContrMB aCV=C.R596 4 ~"'' Good r ss% x rossct wihln, ~vatrnctxnmit Satkxtary U ' s5%. 3g% of pradad O ndMin aroluatc tad IImII:

0 'c'~ ai2.'aK Ut;acfiy v 75%. 95% of p aiuct '. m .~awaWd~im;~

v= SST E; CJ i~r 'J SST cT5!cw~dvaa II

within ~riroluatawcfad O limits ~ SS $

~ SST 9!

$

o so zo ao ao sa co Eras AversOe: h(C-val+COnsoi) 95/a Confidence Intervals 4or Sins SST

_: -~ _ "
-y~~_ ~xp Blame: fiST ~xp P~ame: AST

Study Information:

4C 30 dcnors 2 Serum observations per donor 1 SST obseraation per donor D missing Senarn ob:;eevatinns 9 missing SSTobse~,vation 30 29 corxiparisons - SST 80 expecfed observations (evaluation) 89 actual observations Regression Intorma6on:

20 Intercept=9.2 (-0.9&,2.55) Slope:_D,939 (D.873, 1.09) R squared = 86.9 ' Sirnulfaneoi.as Test of Intercept sand S:ope:

1D . p-value.:: 619945 Accept Ideal t_;ne Hypothesis 90 20 30 a0 Serum (control) I_xp Blame: AST '~xla Blame: ~fiST

Seudy Information:

40 30 donors 2 Serum observatiora per donor 1 SST .I obsee~~afion per donor '0 missing Sen.;m observaRions D missing SST II observations ' 30 30 comparisons 90 expected observations (evaluation) 90 actual observations Regression InfomaNon:

2 eh ~
~

~~9) Slope .02 {0.95 1.09) R squared = 97.4 Simultaneous Tesi of (ntE:rcepEand Slope:

p-value = 64.2~to Accept Idaal L;ne Hy?othesis Serum (conNo4) Ex~ hlame: Trig ~xo iVamp: Trsg 30 30 uJnors 80 dmaaxs Serum cb.,orvations arM
~

2 Serum ofManV tlonsaM
9 mra~.uecon cte:ervm9an 9 evaluaNAn ahsetwuPiat per donor perdanof 20 CwY.rd Hosed CV=9.a7k Canirol Boned CV--t.W

C,ood >Y. oromd,iC
IO
within avalu t~aE l:mt Satis4ucPafy G: - c 95%-6Gb ofWOducP
o ni w <

,.e eua g .~ a vaad ilmrt:
%'~'.: UnatisFaclaf'~

077 7s G.-651L ofVfcuuci g' SSv1 a ,.arym.'.awo<.~doar Ir=

a , ~ _ E ~ Pox z Jai f( - 7s!sdwedcc:
w p ;0 ioYinarswacemcteai;.-.,fh e~~JT

-20 ~ ~ST ~i -sa a zoo Sao Hco aaa ~oaa ~'ias ~

.average: '.4(Eval r c;ontrap ~ , 5 a '/a o P

SST-Serum SST II-Serum ' i<

W! ~j~

~~

Exp iVame: Trig . ~xp f~iar~re: -1 rig soo Study Infannation:

30 donors 2 Serum obsereations per donor 1 SST observation per donor 400 0 missing Sonrm observations Sgr 0 missing SST obsewations (evaluation) 30 comparisons ~s - 90 expected observ:.ztions 90 actual observaRions ' 200 - Regression In;orma6bn:

Intercept=0.636 ((L737,2.51j Slope==0.987 (0.979, 0.995) iao ft squared = 100 Simultaneous TasE
of Intercept and Stape:

0 p-value --. 0.0957 Reject ;deal Line Hypothesis Serum (control) F~~~iame;'F,ri.~n,. uxp(:iaiie8: rii "

Study Information:

30 donrxs 2 Serurn obse, ~aEions per donor 1 SST ti ebservatlon per donor 400 0 missing Setllm obsorvations SST II 0 missing SST If observations (evaluation) - 30 comparisons 90 expeciad observations 90 actual observa8ions 200 Regression information:

'ntercept=-0.0682 (-1.57, 9.43) Slope=1.01 (0.998, 1.01) " R squared = iG0~o Simu lt aneous lest ef Intercept and Stope:

p -l a Reje ct Ide l Line Hypothesis Serum (control) .

Claims (23)

1. A method of determining whether an evaluation testing method is clinically equivalent to a reference testing method, comprising the steps of:
determining a level of variance in results obtained using a reference method, determining a bias between an evaluation testing method and the reference method, comparing the bias relative to the level of variance in the reference method, and based on the comparison, generating a report indicating whether the evaluation testing method is clinically equivalent to the reference method.

2. The method as claimed in claim 1, wherein said determining a level of variance in a reference method step comprises:
computing measures of variability using at least two sets of reference test data associated with the reference method.

3. The method as claimed in claim 1, wherein said determining a level of variance in a reference method step comprises receiving user input representing the level of variance in the reference method.

4. The method as claimed in claim 1, wherein said report comprises a modified mean difference plot.

5. The method as claimed in claim 1, wherein said report comprises a linear regression analysis and a scatter diagram.

6. The method as claimed in claim 1, wherein said report comprises a variability chart having a first axis representing accuracy and a second axis representing precision.

7. The method as claimed in claim 1, wherein said report comprises a confidence interval for bias and a plot of the confidence interval.

8. The method as claimed in claim 1, wherein said report comprises a conclusion as to whether the evaluation testing method is clinically equivalent to the reference method.

9. The method as claimed in claim 1, further comprising the step of providing a graphical user interface for identifying data associated with the reference method and the evaluation testing method.

10. A system for determining whether an evaluation testing method is clinically equivalent to a reference testing method comprising:
a memory adapted to store data related to a reference testing method and an evaluation testing method, the data comprising test results;

a processor adapted to determine from the data a level of variance in the reference method, and a level of variance in the evaluation testing method;
the processor further adapted to compare the level of variance in the reference method to the level of variance in the evaluation testing method; and the processor adapted to generate a report based on the comparison indicating whether the evaluation testing method is clinically equivalent to the reference method.

11. The system as claimed in claim 10, wherein the data comprises at least two sets of reference data associated with the reference method.

12. The system as claimed in claim 10, wherein said system is adapted to receive user input representing the level of variance in the reference method.

13. The system as claimed in claim 10, wherein the report comprises a modified mean difference plot.

14. The system as claimed in claim 10, wherein the report comprises a variability chart having a first axis representing accuracy and a second axis representing precision.

15. The system as claimed in claim 10, wherein the report comprises a conclusion as to whether the evaluation testing method is clinically equivalent to the reference method.

16. The system as claimed in claim 10, further comprising a graphical user interface adapted to identify data in a table associated with the reference method and the evaluation testing method, respectively.

17. A computer readable medium of instructions adapted to control a system to determine whether an evaluation testing method is clinically equivalent to a reference testing method, comprising:
a first set of instructions adapted to control the system to determine the level of variance in a reference method;
a second set of instructions adapted to control the system to determine the level of variance in an evaluation testing method;
a third set of instructions adapted to control the system to compare the level of variance in the evaluation testing method to the level of variance in the reference method; and a fourth set of instructions adapted to control the system to generate a report based on the comparison, the report indicating whether the evaluation testing method is clinically equivalent to the reference method.

18. The computer readable medium of instructions of claim 17, wherein the third set of instructions is further adapted to compare at least two sets of reference test data associated with the reference method to determine the level of variance in the reference method.

19. The computer readable medium of instructions of claim 17, wherein the third set of instructions is adapted to receive user input representing the level of variance in the reference method.

20. The computer readable medium of instructions of claim 17, wherein the fourth set of instructions is further adapted to generate a report including a modified mean difference plot.

21. The computer readable medium of instructions of claim 17, wherein the fourth set of instructions is further adapted to generate a report including a variability chart having a first axis representing accuracy and a second axis representing precision.

22. The computer readable medium of instructions of claim 17, wherein the fourth set of instructions is further adapted to generate a report including a conclusion as to whether the evaluation testing method is clinically equivalent to the reference method.

23. The computer readable medium of instructions of claim 17, further including a fifth set of instructions adapted to control the system to provide a graphical user interface adapted to identify data associated with the reference method and the evaluation testing method.