Tools, Technologies and Training for Healthcare Laboratories

QC - The Planning Process

How do you design quality? Not by throwing darts at a Levey-Jennings charts. You have to choose methods, numbers of materials and controls, control rules, and more. You need a design (or planning) process for your QC. Here's the one that Dr. Westgard recommends.

PLEASE NOTE: An updated version of this lesson is now available in Basic QC Practices, 2nd Edition.

Many of you know about my Scandinavian heritage from my frequent references to one of my ancestors, Hagar the Horrible, who is a source of both inspiration and practical advice. In one of my favorite Hagar cartoons, Helga asks why girls mature faster than boys. This is a good example of a process problem that all of us have experienced as teenagers and many of us also as parents. Hagar's answer is "poor planning." That certainly rings true. And those of us who are parents of more than one child have been repeatedly frustrated by this process, even though we supposedly gain experience that should make it easier the next time. The difficulty is that we really aren't able to change the process and just have to live through it again and again. Everyone in the future will also have to endure this problem and live through a similar experience. What is really needed is to replan the process to eliminate the problem, but that's something that only top management has the power to do.

In many ways, this problem is analogous to the QC problems experienced in many laboratories. Bench level analysts often have to live with QC problems because the testing process has not been properly planned, particularly the selection of an appropriate statistical QC procedure. The people at the bench can't generally change the process because management retains that power. Many of you who are reading this are probably QC specialists, supervisors, managers, or directors, and you do have the power to change the process. You need to be able to select QC procedures that will assure the desired analytical or clinical quality is achieved, taking into account the imprecision, inaccuracy, and instability observed for your methods. That's what QC planning is all about.

What information is needed for QC planning?

QC planning requires information about the quality requirements for different tests, the imprecision and inaccuracy observed for a specific testing process, and the probabilities of rejection expected for the QC procedures of interest. Information on quality requirements will be found in recommendations in scientific literature, regulatory guidelines, and proficiency testing criteria. This website provides a discussion of different types of quality requirements, as well as summary tables of analytical quality requirements, clinical quality requirements, biological goals, and an extensive database of desirable specifications based on biologic variation.

Reliable estimates of method performance can be obtained initially by following established method evaluation protocols and later from on-going estimates of imprecision from measurements accumulated on control materials and estimates of bias from proficiency testing surveys. Performance characteristics of commonly used QC procedures are available in the clinical chemistry literature. QC simulation programs may be employed by laboratory analysts to estimate the power of new control rules, including patient data algorithms. Probability calculations may also be utilized by industrial statisticians and laboratory analysts with the necessary mathematical skills.

What are some practical QC planning approaches?

A systematic QC planning process is needed to consider all this critical performance information in an orderly manner. The key to practicality is being able to do QC planning in a matter of a few minutes, rather than a few hours or days. This means supporting data calculations, preparing graphical tools and charts, and making it easy to document the QC recommendations.

Based on my experience with QC planning over the last several years, both in the laboratory and in classrooms and workshops, there are four approaches that are practical today. These approaches take advantage of tools and technology that make it simpler and faster to perform QC planning. It is essential to recognize that new tools and technology will be needed if you are to perform QC planning quickly and efficiently. As with any new tool or technology, you need to initially spend some time understanding the principles and theory, satisfy yourself that the approach is scientifically sound, and then implement the process steps and operations. The planning process can be quick and easy to perform with the new tools and technology, even though the principles and theory may be complicated.

1. Computer supported QC planning process

The most quantitative, comprehensive, and flexible approach is to provide computer support for the QC planning process. For several years, I used electronic spreadsheets to prepare power function graphs and OPSpecs® charts, but this still took hours and sometimes days to assess QC performance, select new QC procedures, and document QC recommendations. To reduce the time required, specialized software ( our EZ Rules® 3 programs) has been developed to perform the calculations, prepare the necessary graphs and charts, and provide documentation in a matter of a few minutes. Use of QC Design software has been described in the literature [1,2] and elsewhere on this website. Computer support permits both analytical and clinical quality requirements to be considered and makes available the power functions for as many as 100 combinations of control rules and N's. The program's HELP facility can be used to provide quick access to information, such as the CLIA PT requirements. QC planning applications on this website illustrate the use of power function graphs, critical-error graphs, OPSpecs charts, and QC Validation Reports that have been prepared by the program. For a cholesterol example, see our QC application using Cholesterol with a clinical requirement.

2. Manual QC planning with a workbook of OPSpecs charts

A manual process can be supported using a collection of preprinted OPSpecs charts (OPSpecs Manual) [3] that cover a wide range of analytical quality requirements (0.5% to 50%), commonly used control rules (12s, 12.5s, 13s, 13.5s, 13s/22s/R4s/41s), and commonly used numbers of control measurements (2,3,4,6). This manual QC planning process is limited to analytical quality requirements that are stated in the form of allowable total errors, such as the CLIA proficiency testing criteria for acceptable performance. To use the OPSpecs Manual, you first define the quality required for the test of interest and then lookup the corresponding OPSpecs charts in the manual. You plot your observed imprecision and inaccuracy as the "operating point" and then inspect the charts to select QC procedures whose operating limits (for the imprecision and accuracy that are allowable) are above your operating point. Both 90% AQA and 50% AQA OPSpecs charts are provided to plan QC procedures that will achieve 90% error detection (generally preferred) or 50% error detection (okay for very stable methods that have few problems). Application of this approach has been illustrated in a recent paper [4] that provides an error budget perspective and explanation of the quality- planning models and the OPSpecs chart. For a cholesterol example, see our QC application using Cholesterol with an analytical requirement.

3. Manual QC planning with normalized OPSpecs® charts

Standardized or normalized OPSpecs charts can be prepared that apply to any total error requirement and a limited menu of control rules and limited numbers of control measurements. Manual calculations are necessary to present method imprecision and inaccuracy as a percent of the total error requirement. The normalized operating point is then plotted and its location used to determine which control rules and N's are appropriate. Normalized charts are available in the literature [5] and incorporate either 90% or 50% error detection for the 13.5s, 13s, 12.5s, 12s, and 13s/22s/R4s/41s control rules with N's of 2 and 4. A normalized OPSpecs calculator is available on this website, as well as a Lesson on Normalized OPSpecs Charts.

4. Manual QC planning with QC Selection Grids

Another approach is to simplify the QC planning process with the aid of QC Selection Grids (QCSG) [6]. One QCSG has been constructed for single-rule QC procedures that can be implemented on Levey-Jennings control charts. Another provides adaptations of the Westgard multi-rule procedure. These are tables for looking up QC recommendations based on the expected process capability and the process stability of a method. A quantitative estimate of process capability is obtained by manually calculating the size of the systematic error that is medically important; a qualitative estimate of process stability is assessed from the expected frequency of problems or errors, classifying method stability as excellent, moderate, or poor. After the candidate QC procedures have been selected from the tables, the actual QC performance must be assessed by interpolation of power function graphs to make a final selection.

Which approach should you use?

The easiest approach to get started with is manual QC planning using the OPSpecs Manual. The approach can be learned by following the example applications that are included in the manual, from information available in the laboratory management literature [4], or by participating in one of our workshops or training programs.

Computer support is essential to use clinical quality requirements. You may be able to implement the quality-planning models if you are skilled in using electronic spreadsheets, or you can consider specialized software such as the QC Validator program that implements both the analytical and clinical models. Validator manual features plus an Automatic QC Selection feature that is initiated with the click of a button. Operation of the program can be learned from the tutorials that are supplied with the program.

Normalized OPSpecs charts and QC Selection Grids require manual calculations, thus they are not as simple and reliable as the OPSpecs® Manual and the EZ Rules 3 program. However, they are documented in the scientific literature [5,6] and are readily available alternatives that should be practical in the initial learning phase. Once the importance and usefulness of QC planning has been established, you will want to progress to the OPSpecs Manual or the EZ Rules 3 program.

Education and training in QC Planning

Extensive materials are available at


  1. Westgard JO, Stein B. Automated selection of statistical quality control procedures to assure meeting clinical or analytical quality requirements. Clin Chem 1997;43:400-403.
  2. Westgard JO, Stein B, Westgard SA, Kennedy R. QC Validator 2.0: a computer program for automatic selection of statistical QC procedures for applications in healthcare laboratories. Comput Methods Programs Biomed 1997;53:175-186.
  3. Westgard JO. OPSpecs Manual - Expanded Edition. Ogunquit, ME, Westgard QC, 1996, distributed by Dade International and also available through the AACC, ASCLS, and CLMA.
  4. Westgard JO. Error budgets for quality management: Practical tools for planning and assuring the analytical quality of laboratory testing processes. Clin Lab Manag Review 1996;10:377-403.
  5. Westgard JO. Charts of operational process specifications ("OPSpecs Charts") for assessing the precision, accuracy, and quality control needed to satisfy proficiency testing performance criteria. Clin Chem 1992;38:1226- 1233.
  6. Westgard JO, Quam EF, Barry PL. Selection grids for planning quality control procedures. Clin Lab Sci 1990;3:273-280.