This procedure generates Levey-Jennings control charts on single variables. The Levey-Jennings control chart is a special case of the common Shewart Xbar . The Levey-Jennings chart was created in the s to answer questions about the quality and consistency of measurement systems in the. The Levey-Jennings chart usually has the days of the month plotted on the X-axis and the control observations plotted on the Y-axis. On the right is the Gaussian.
|Published (Last):||7 July 2018|
|PDF File Size:||19.91 Mb|
|ePub File Size:||9.3 Mb|
|Price:||Free* [*Free Regsitration Required]|
This lesson discusses one of the cornerstones of QC practice. We can no longer take for granted that everyone knows how to build a control chart, plot the control values, and interpret those results correctly.
Barry, co-author of Cost-Effective Quality Control: Managing the Quality and Productivity of Analytical Processesprovides a primer on how to construct, use, and interpret the Levey-Jennings chart.
This exercise is intended to show, in step-wise fashion, how to construct a Levey-Jennings control chart, plot control values, and interpret those results. This assumes you already have a selected appropriate control materials, b analyzed those materials to characterize method performance by collecting a minimum of 20 measurements over at least 10 days, c calculated the mean and standard deviation of those data, and d selected the number of control measurements to be used per run and e selected the control rules to be applied.
See QC – The Materials for more information about selecting appropriate control materials. See QC – The Calculations for detailed information about calculating the mean and standard deviation. See QC – The Planning Process for a description of the approach, tools, jenninggs technology available to select QC procedures on the basis of the quality required for a test and the performance observed for a method.
The materials were analyzed once per day for a period of twenty days. From these data, the means and standard deviations were calculated to be:. Each of the two control materials will be analyzed once chaft run, providing a total of two control measurements per run. Control status will be judged by either the 1 2s or 1 3s rule.
These rules are defined as follows:. The 1 2s rule is very commonly used today, and while it provides high error detection, the use of 2s control limits gives an expected high level of false rejections.
The 1 3s rule provides an alternative Ldvey procedure that has lower false rejections, but also lower error detection. In this exercise, you will see how to apply both QC procedures and also get a feel for the difference in their performance. Two sets of control limits will be needed to implement the rules described above.
For Control 2, you should have 2s control limits of and and 3s control limits of and This exercise shows how to construct control charts manually using standard graph paper. For this exercise, graph paper having 10×10 or 20×20 lines per inch works well. You will need two sheets, one for each chart of the two control materials. While it is possible to prepare both charts on a single sheet, this may reduce the readability of the control charts. If you do not have graph paper available at this time, print out the lower resolution grids below.
Click here to get a larger chart you can print out separately. Click here if you want to print a larger version of this chart separately. Once the control charts have been set up, you start plotting the new control values that are being collected as part of your routine lwvey.
The idea is that, leveg a stable testing process, the new control measurements should show the same distribution as the past control measurements. That means it will be somewhat unusual to see a control value that exceeds a 2s control limit and very rare to see a control value that exceeds a 3s control limit.
If the method is unstable and has some kind of problem, then there should be a higher chance of seeing control values that exceed the control limits. Therefore, when the control values fall within the expected distribution, you classify the run to be ” in-control, ” accept the results, and report patient test results. When the control values fall outside the expected distribution, you classify jenninhs run as ” out-of-control, ” reject the test values, and do not report patient test results.
Prepare appropriate control charts and interpret the results. Click here to view the answers to the exercise.
QC: The Levey-Jennings Control Chart – Westgard
Tools, Technologies and Training for Healthcare Laboratories. The Levey-Jennings Control Chart. Example application QC procedure s to be implemented Calculation of kennings limits Preparation of control charts Use of control charts Answers for this exercise Interpretation of example test results Please Note: This article is from the first edition.
An updated version is now available in Basic QC Practices, jenningw Edition You can link here to an online calculator which will calculate control limits for you. From these data, the means and standard deviations were calculated to be: These rules are jenninge as follows: In many laboratories, this rule is used to reject a run when a single control measurement exceeds a 2s control limit.
An analytical run is rejected when a single control measurement exceeds a 3s control limit.
Calculation of control limits Two sets of control limits will be needed to implement the rules described above.
Other information typically included on the chart are the name of the analytical system, the lot number of the control material, the current mean and standard deviation, and the time period covered by the chart. Scale and label x-axis. The horizontal uennings x-axis represents time and you will typically set the scale to accomodate 30 days per month or 30 runs per month. For this example, divide the x-axis into evenly sized increments and number them sequentially from 1 to Label the x-axis “Day.
The vertical or y-axis represents the observed control value and you need to set the scale to accomodate the lowest and highest results expected.
This can be rounded to to to fit the 10×10 or 20×20 grids of the graph paper. Mark off and identify appropriate concentrations on the y-axis. Label the y-axis “Control value. Draw lines for mean and control limits. What are the mean and control limit lines for Control 2? Use of Control Charts Once the control charts have been set up, you start plotting the new control values that are being collected as part of your routine work. For practice, the accompanying table provides some control results for our example cholesterol method.
Plot these results, one from Control 1 and one from Control 2, for each day.
Levey Jennings Control Chart
You can print the Levey-Jennings QC Practice Chrat below to obtain a worksheet that shows all these control results. For day 1, the value for Control 1 is and Control 2 is On the chart for Control 1, find the value of 1 on the x-axis and the value of on the y-axis, follow the gridlines to where they intersect, and place levye mark; it should fall on the mean line.
On the chart for Control 2, find the value of 1 on the x-axis and the value of on the y-axis, legey mark that point; it should fall a little below the mean line. In plotting control values, it is common practice to draw lines connecting the data points on the control chart to provide a stronger visual impression and make it easier to see patterns and shifts.
Apply the 1 2s and 1 3s control rules and make a decision whether you should accept or reject the run for each day. The control values for the first day are in-control and the patient results can be reported.
Continue plotting the 2 control values per day and interpreting those results. Circle those points that correspond to runs that should be rejected.
Patient results obtained in runs where the 1 3s rule is violated are most likely incorrect. This is a “false alarm” problem that is inherent with the use of 2s control limits with an N of 2. In spite of this serious limitation, many laboratories continue to use 2s control limits an just routinely repeat the run and the controls, or sometimes repeat only the controls by themselves. Note that if a control is out a second time, the actual control rule that is being used to reject a run is a 2 2s rule rather than the stated 1 2s rule.
Unfortunately, the 2 2s rule by itself is not very sensitive, therefore, it is better to use the 1 3s and 2 2s rules together cyart a multirule procedure to improve error detection while, at the same time, maintaining a low false rejection rate. We’ll provide more discussion of multirule QC procedures in a later lesson.