NCCLS document EP5-A2 (ISBN ). NCCLS, West Valley Road, Suite , Wayne, Pennsylvania USA, THE NCCLS . Evaluation of Precision Performance (EP5-A2). (This feature is only available in GenEx Enterprise). Introduction. The EP5 module in Genex implements the. Buy CLSI EP5 A2 Ed. 2 () Evaluation Of Precision Performance Of Quantitative Measurement Methods from SAI Global.

Author: Gardall Arakazahn
Country: Cambodia
Language: English (Spanish)
Genre: Video
Published (Last): 18 July 2011
Pages: 54
PDF File Size: 18.79 Mb
ePub File Size: 6.53 Mb
ISBN: 399-8-53395-794-4
Downloads: 5881
Price: Free* [*Free Regsitration Required]
Uploader: Guzilkree

The average of the two samples that form the run is marked with a red bar. EP5 also estimates the variance of the sum of q2 three errors, the so called within-laboratory precision.

We give below some of the main features of the EP5 guidelines, for a detailed description see the EP2-A2 document.

The experiment produces a data set with one or two runs per day and two replicas per run.

Page not available

Finally, the last two lines indicate whether the claimed precisions were consistent with the precision estimated from the measurements. EP5 estimates the repeatability, defined as the between-sample precision, i. Note that it is the absolute value of the variance that clzi be entered. The measurements continue for another 15 days. The runs should be separated by at least two hours, or if they take very long time to perform, one run per day is acceptable. This feature is only available in GenEx Enterprise.

The fields under manufacturers claimed precisions allow testing against a claimed precision. If two or more concentrations are used, they can be handled as separate data sets, and be analyzed independently of each other in Genex. Using the EP5 module in Genex. The between-runs and between-days clsk estimates are also given on lines 10 and Device Familiarization Period is a period to learn the operation of equipment and preparation of samples required.


If testing against a claim is not desired, these fields can be left blank.

Evaluation of Precision Performance (EP5-A2)

Genex assumes that the experiment has been conducted according to the EP5-A2 instructions, producing data during 20 days with one or two runs per day. Genex will estimate the precision in the data by the statistical methods required by the EP5-A2 document. Where X drs is the measurement on day d in run r on sample s. The purpose of EP5 is to evaluate the precision of clinical laboratory measurement methods. The eep5 and runs should be given as classification columns in the data input to Genex, as in the example.

Furthermore, EP5 allows testing of the estimated precisions against given precisions, e. The alternative measurement protocol with one run per day, only requires the day classification column, and would have had 40 rows. The EP5 module is found by selecting the Quality Ctrl tab, and clicking on the aa2 labeled EP5, shown below as the second button from the left. Quality controls should be made regularly, based on quality control samples that are measured along with the rest of the experiment.

The precision evaluation experiment is divided into three different periods. The analysis produces two windows, one scatter plot of the data and one data window with the computed precisions.

Furthermore, the protocol defines a procedure for dealing with outliers. The random error is modeled according to the hierarchical model.

Page not available

Line 11 shows the same quantities as line 9, but for the total, within-laboratory, precision. In EP5 there should be two runs on each day, where each run consists of two test samples.


After all fields have been set to appropriate values, press the run button to perform the analysis. The picture only shows the first part of the data set, in total there are 80 sample rows, since the measurements are carried out for 20 days. This period is intended for the experimenter to learn the experimental protocol, and also to put quality controls into place, and perform an initial evaluation of precision and other acceptance tests, e.

If there is only one run per day, the Runs: Furthermore, EP5 recommends that the experiment is performed at two different concentrations. This scatter plot displays the input data with the runs on the x-axis. Each sample is displayed as a blue marker.

This period lasts until data are obtained without operational difficulty for five days. We refer to the EP5-A2 document for details about the experiment protocol.

The control panel for EP5. The estimated between-samples variance the repeatabilityshown on line 9 is 2. If the claim is given as a percentage, it has to be translated to a variance by multiplying the average of the measured values. A22 the menus to the right under Columns to tell Genex which classification column that corresponds to days, and which column corresponds to runs. The average of all measurements is shown on line 6.

The EP5 guidelines are intended both for developers of such methods and for users of methods who want to verify the validity of given precision claims, or just measure their own precision.

In total, data should be generated for 20 days, the five last days of the Protocol Familiarization plus the 15 days of the Precision Evaluation Experiment. The SD min and SD max values form a confidence interval for the precision estimate, hence the confidence interval is [2.


The model assumes that rp5 the U d are identically distributed, and similarly that the V dr clsu identically distributed and that the W drs are identically distributed.