Printer Friendly

Spreadsheet templates for calculating precision according to NCCLS guidelines.

THE CLINICAL Laboratory Improvement Amendments of 1988 (CLIA '88) mandate that laboratories establish and verify certain performance characteristics of new methods introduced into their laboratories after Sept. 1, 1992, prior to reporting patient test results. Precision is listed among those performance characteristics that must be validated (493.1213).

The National Committee for Clinical Laboratory Standards (NCCLS) has issued a tentative guideline called, "Evaluation of Precision Performance of Clinical Chemistry Devices," usually referred to simply as EP5.|1~ The "Forward" of this document describes a need to standardize protocols for product evaluation using appropriate experimental and statistical procedures. EP5 provides a paradigm for estimating both with-in-run and total precision using sound experimental design and statistical methods.

* Template features. The spreadsheet templates described here perform automatically all calculations described in EP5, including looking up and interpolating data from the statistical table of "Critical Values of Chi-Square." Logical functions are embedded in various data sheets. For example, functions to test for outliers are embedded in data sheet no. 2. Functions to evaluate and either accept or reject the device manufacturer's precision claims are embedded in data sheet no. 4. Macros are used to automate certain features, among them printing the four data sheets that emulate those found in EP5. Figure 2 shows the macros embedded in the Lotus (Lotus Development Corp., Cambridge, Mass.) template. A table describing the symbols used in the templates is located in cells A148..I166.

Before attempting to use the spreadsheet templates, the user should gain familiarity with EP5-T2 and follow the protocol outlined therein to obtain the raw data. The raw data may be recorded on a photocopy of a blank data sheet no. 1 found on page 20 of EP5-T2.

The template was originally written using Lotus 1-2-3, version 2.01. Cosmetic enhancements were added using WYSIWYG (Lotus) add-ins in versions 2.3 and 2.4. The Quattro Pro (Borland International, Inc., Scotts Valley, Calif.) macros were written in version 1.0 but were also tested in version 3.0 and found to be entirely compatible.

* The procedure. After obtaining the raw data, enter the spreadsheet session in the usual manner. Retrieve the template file suitable to the software package and version that you are using, whether it be for Lotus 1-2-3 having the WYSIWYG add-in options; Lotus 1-2-3 versions not having add-in features; or for Quattro Pro.

Figure 3 lists the user editable cells. In data sheet no. 1, the user types descriptive information in cells D6, B7, B8, G6, G7, H1, and F8 as indicated in Figure 3. When inputting "Concentration (nominal)" into cell D6, if a numeric value is used (as opposed to designations such as low, medium, high, normal, abnormal), type it as a label by preceding the number with an apostrophe. Typing the descriptive information into data sheet no. 1 automatically updates the information elsewhere in the spreadsheet.

Cells B15..B34 are formatted for the long international data format. To retain this feature, the user should input the date using the @DATE(YY,MM,DD) function, or simply edit the year, month, day values already stored in the cells. The dates may be erased (/reb15..b34), formatted (/rfd) to other styles, or typed as a label without adversely affecting the template.

Type the raw data into cells C14..F34. The \r macro will format the raw data to a fixed number of decimal places specified by the user. Invoke this macro by pressing <r> while holding down the <Alt> key. When the macro pauses at the prompt, "Enter number of decimal places (0..15):2," type the number corresponding to the number of decimal places to be displayed, and press <Enter>.

The table, "Critical Values of Chi-Square," found on page 15 of the EP5-T2 guideline, is embedded in the template. Formulas in cells F121 and F139 automatically search for TABULAR DATA OMITTED TABULAR DATA OMITTED and interpolate the correct value for Critical Chi-square for the Degrees of Freedom R (within-run) and T (total), respectively. The templates are set to use 95% critical value. The user may change to 99% critical value by editing cell R8 from "1" to "2".

The template will perform an outlier test on each run. For the template to accurately evaluate the runs for outliers, the user must run a preliminary precision, single-run batch experiment as recommended by the EP5-T2 guideline (page 5, paragraphs 3.6 and 3.7) and put the standard deviation (SD) from that run in cell I40. The data obtained from this preliminary precision test may be typed into cells Y6..Y45. The number of observations, mean, SD, and coefficient of variation will be calculated and displayed in cells Y47..Y50. The user may either type an SD into cell I40 or invoke the \y macro, which will stuff the SD in cell Y49 in cell I40. The flags are displayed in data sheet no. 2: cells D50..D69 for run 1 and cells F50..F69 for run 2. The flag will be either "ok" or "outlier." If the flag indicates an outlier, EP5-T2 recommends that the data pair be rejected and the run repeated.

Lastly, the user must edit cells F115 and F133 in data sheet no. 4. These cells are for the reagent or device manufacturer's claimed within-run and total precision expressed as standard deviation. This information is normally obtained from reagent package inserts or instrument operating manuals. (If the manufacturer provides precision data as the coefficient of variation, be sure to convert it to standard deviation.) Cells F123 and F125 for within-run precision and cells F141 and F143 for total precision display the acceptance or rejection of the manufacturer's claim. If "ACCEPT" is displayed, the precision of the method was better than or equal to that claimed by the manufacturer. If "REJECT" is displayed, the precision claimed by the manufacturer was not achieved.

To print the data sheets, set the printer pitch to 17 or 20 characters per inch, and invoke the \p macro (<Alt><p>). Figure 1 shows an example of the printed data sheets. To print a table of symbols used in the template and their descriptions, invoke the \s macro.

The templates may be obtained at no charge by sending a DOS formatted diskette and a self-addressed mailer with return postage affixed to: Carole Smith, Laboratory Service 113, VA Medical Center, 1310 24th Ave., South, Nashville, TN 37212-2637.


1. National Committee for Clinical Laboratory Standards. Evaluation of Precision Performance of Clinical Chemistry Devices. 2nd ed. Tentative Guideline EP5-T2. Villanova, Pa: NCCLS; 1992.

Smith is supervisor of clinical chemistry, Dr. Sephel is director of clinical chemistry, and Dr. Woodward is chief of laboratory service at the VA Medical Center in Nashville, Tenn.
COPYRIGHT 1993 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1993 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:National Committee for Clinical Laboratory Standards
Author:Smith, Mary Carole
Publication:Medical Laboratory Observer
Date:Oct 1, 1993
Previous Article:Handling romance in the lab.
Next Article:High tech gone awry.

Related Articles
A program to prepare Levey-Jennings charts with integrated software.
Use of a spreadsheet for method evaluation statistics.
Taking the 'byte' out of method evaluation number crunching.
Spreadsheet templates for calculating linearity according to EP6-P.
Templates for generating Levey-Jennings and Westgard QC charts.
Training verification for lab personnel: a new guide.
New quality guidelines for laboratories.
Specimen stability.
Lipase method problems.

Terms of use | Copyright © 2017 Farlex, Inc. | Feedback | For webmasters