BQC4ED Preview
BQC4ED Preview
BQC4ED Preview
FOURTH EDITION
Training in Statistical Quality Control
for Medical Laboratories
ISBN 1-886958-30-0
ISBN-13 978-1-886958-30-2
Published by Westgard QC, Inc.
7614 Gray Fox Trail
Madison, WI 53717
Phone 608-833-4718
It was 1998 when we published the first edition of this book, 2002 for the second
edition, and 2008 for the third edition. Since that time, it might be hoped that
the advances in technology have improved the Total Testing Process to the
point where a book on Basic QC Practices is no longer necessary. Many think
that analytical quality is a given and laboratories should focus on pre-analytic
and post-analytic processes, rather than analytic processes. Let us remind you
that there is nothing more dangerous to our patients than test results that are
wrong! Neither doctors nor patients can adequately inspect the products of our
testing processes to make sure they are okay. They depend on us to make sure
the numbers are correct. Analytical quality is perhaps even more critical today
when patient testing is performed by many different measurement procedures in
different laboratory settings by personnel with different laboratory skills.
Quality is like safety; our efforts must first attempt to prevent problems
from occurring, but if and when problems happen, it is essential they be detected
before they endanger the patient. In a perfect world of analytical testing, all
problems would be prevented. Unfortunately, we do NOT yet live in a perfect
world, there continue to be analytical problems with today’s measurement sys-
tems, and we still need the capability to detect those problems when they occur.
One of the most cost-effective tools for detection is Statistical Quality Control
(SQC). It is a powerful tool when properly designed and properly implemented,
however, SQC can itself be very problematic if poorly designed or improperly
implemented.
This book will help you understand the principles and practices so you can
do the right SQC right. The first right has to do with implementing the right
control rules and right numbers of control measurements in order to detect medi-
cally important errors. The second right has to do with knowing how to properly
implement SQC procedures, which involves many details about setting up control
charts, plotting control data, interpreting control results, recognizing when test
results can be reported, and deciding when it is necessary to trouble-shooting
problems and fix the testing process.
Acknowledgments
This book would not exist without the help of others. A number of colleagues
have helped with the writing over the years and are identified in the list of
authors and contributors. Their help has been critical for conveying the ideas and
practices that are necessary to make SQC practical in medical laboratories.
James O. Westgard
Sten A. Westgard
James O. Westgard
Madison Wisconsin
About the authors and contributors
• Glossary of terms
• New in this 4th edition: free access to an online short course in "West-
gard Rules" – which will allow access to an online lecture as well as
the opportunity to earn continuing education credits.
Westgard QC, Inc, Copyright © 2016
Table of Contents
Index .................................................................................................................................................
329
Westgard QC, Inc. Copyright © 2016
What is Quality?
Everyone talks about quality, but what exactly does “Quality” mean?
You can find many definitions in the literature, but we need one that
helps us understand laboratory quality in a practical way. Here are
some definitions that provide a good starting point:
• ANSI/ASQC A3-1978. Quality – the totality of features and
characteristics of a product or service that bear on its ability to
satisfy given needs [5].
• Juran – Quality is fitness for use [6].
• Crosby – Quality is conformance to requirements [7].
• Deming – Quality should be aimed at the needs of the customer
[8].
Page 9
Basic QC Practices, 4th Edition
These definitions are from industry and from some of the quality
gurus who guided industrial quality management in the last half of
the 20th century. Notice that all these definitions focus on the needs
or requirements of the customer, in our case the patient (consumer)
and the physician (user) who acts on behalf of the patient. Note also
that there may be many dimensions to quality, i.e., many features or
characteristics that are important. The following is a good synthesis
of these definitions and should provide us with a starting point for
understanding quality in a quantitative way:
• CDC 1986. The quality of a laboratory testing service depends
on providing the totality of features and characteristics that
conform to the stated or implied needs of users or customers
[9].
Note that this definition is 30 years old! Quality management
in medical laboratories has a long history and was not invented by
ISO or CLSI or CLIA. Those guidelines emerged from attempts to
summarize or codify practices that had already been developed in
laboratories and to transfer “good laboratory practices” more widely
throughout the field.
This definition of quality acknowledges the “totality of features
and characteristics” that makes quality multi-dimensional. Some-
times when we argue about quality, we are actually arguing about
different characteristics, e.g., a physician says they are upset about
the poor quality of a test. We argue that the test was in-control and
meets the requirements for allowable error. But the physician is
actually upset about turnaround time, a very different characteristic
of quality. If the laboratory requirement for turnaround time was
set at 1 hour and the report was not provided until 2 hours, then the
physician has a right to complain that the laboratory did not provide
the quality of service expected. That also shows the importance of
defining how good the test or service must be.
Conformance to stated or implied needs, i.e., goals, objectives,
or requirements, is the key to making quality a quantitative and
measurable characteristic. For turnaround time, the laboratory
needs to establish its requirements for reporting routine, priority,
and emergency test results. Discussions with the users, in this
Page 10
Westgard QC, Inc. Copyright © 2016
case physicians and nurses, should help identify the clinical needs.
Because the measure is in units of time, i.e., minutes or hours, that
are understood by both the customer and the laboratory, we can
communicate, define, and measure quality in terms of time.
Analytical quality is not as easy to understand. How should
the requirement be stated? Should it be in the form of allowable
bias, allowable imprecision, or allowable total error? And how do
you come up with a number? In this case, the physician-user and
patient-consumer do not have the technical understanding to discuss
the performance characteristics of a measurement procedure, there-
fore, laboratory scientists must be responsible for understanding the
clinical needs and interpreting those needs in terms of analytical
performance goals. That’s our responsibility for “implied needs”
in the definition of quality. The laboratory must take responsibil-
ity because the physician and patient cannot define quality in the
technical terms that are used in the laboratory.
Page 11
Basic QC Practices, 4th Edition
Page 12
Westgard QC, Inc. Copyright © 2016
Re-engineering
Six Sigma Design
Lean
Risk
Q Planning CLSI, ISO
Q QQ
Management Guidelines
Laboratory
Improvement Processes
Q Q
Goals
Inspection
Six Sigma Accreditation
Assessment Control
Quality Indicators
Figure 1-1. Total Quality Management (TQM) Process Framework and
related programs for Quality Improvement.
This model clearly reflects the PDCA cycle, where the “plan”
is at the top, “do” describes the laboratory processes for getting the
work done, “check” includes both quality control and quality assess-
ment, and “act” reflects the actions needed to correct problems and
improve the process. In describing this process, we’ll start with the
“do” because the definition of work processes and standardization of
how the work gets done is usually the first step in managing quality.
Page 13
Basic QC Practices, 4th Edition
Page 14
Westgard QC, Inc. Copyright © 2016
Page 15
Basic QC Practices, 4th Edition
Page 16
Westgard QC, Inc. Copyright © 2016
Page 17
Basic QC Practices, 4th Edition
Page 18
Westgard QC, Inc. Copyright © 2016
Page 19
Basic QC Practices, 4th Edition
Page 20
Westgard QC, Inc. Copyright © 2016
References
1. ISO 15189. Medical laboratories – Particular requirements for quality
and competence. ISO, Geneva, 2012.
8. Deming WE. Out of the Crisis. Cambridge MA;MIT Center for Advanced
Engineering Study, 1986.
10. 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.
11. Westgard JO. Six Sigma Quality Design and Control: Desirable precision
and requisite QC for laboratory measurement processes, 2nd ed. Madison
WI:Westgard QC, 2006.
12. Westgard JO, Westgard SA. Basic Quality Management Systems. Westgard
QC, Inc. Madison WI 2014.
Page 21
Basic QC Practices, 4th Edition
Page 22
Westgard QC, Inc. Copyright © 2016
Histogram
8
5
Frequency
0
234
239
245
249
254
259
264
More
Values
Page 15
Basic QC Practices, 4th Edition
Page 16
Westgard QC, Inc. Copyright © 2016
More
234
239
245
249
254
259
264
Values
270
265
264
260
255
254
250
245
245
240
235
234
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
0 2 4 6 8
Page 17
Basic QC Practices, 4th Edition
Page 18
Westgard QC, Inc. Copyright © 2016
Page 19
Basic QC Practices, 4th Edition
Page 20
Westgard QC, Inc. Copyright © 2016
Page 21
Basic QC Practices, 4th Edition
Page 22
Westgard QC, Inc. Copyright © 2016
Page 23
Basic QC Practices, 4th Edition
References
1. Shewhart WA. Economic Control of Quality of Manufactured
Product. New York; D. Van Hostrand Company, Inc., 1931.
2. Levey S, Jennings ER. The use of control charts in the clinical
laboratory. Am J Clin Pathol 1950;20:1059-66.
3. Henry RJ, Segalove M. The running of standards in clinical
chemistry and the use of the control chart. J Clin Pathol
1952;27:493-501.
4. Westgard JO, Groth T, Aronsson T, Falk H, deVerdier C-H.
Performance characteristics of rules for internal quality control:
probability for false rejection and error detection. Clin Chem
1977;23:1857-67.
5. Westgard JO, Barry PL, Hunt MR, Groth T. A multi-rule
Shewhart chart for quality control in clinical chemistry. Clin
Chem 1981;27:493-501.
6. Westgard JO, Barry PL. Cost-Effective Quality Control:
Managing the quality and productivity of analytical processes.
Washington DC:AACC Press, 1986.
7. CLSI C24-A3. Statistical Quality Control for Quantitative
Measurement Procedures. Clinical Laboratory Standards
Institute, Wayne, PA 2006.
8. CLSI EP23A. Laboratory Quality Control Based on Risk
Management. Clinical Laboratory Standardis Institute, Wayne,
PA 2011.
Page 24
Basic QC Practices, 4th Edition
No No No No No
13s 22s R4s 41s 8X
Figure 11-6. Westgard Sigma Rules for 2 levels of controls. Note Sigma-scale at
the bottom of the diagram. To apply, determine Sigma-metric, locate on the Sigma-
scale, identify rules above and to the left, find N and R above the sigma value.
. On first glance, this figure looks just like the usual “Westgard
Rules.” But there is no 2 SD warning rule here. That is an important
distinction, but the most important change is at the bottom of the
diagram where there is a Sigma-scale. That scale provides guid-
ance for which rules should be applied based on the sigma quality
determined in your laboratory.
Page 148
Westgard QC, Inc., Copyright © 2016
Page 153
Westgard QC, Inc., Copyright © 2016
Page 161
Westgard QC, Inc., Copyright © 2016
Page 171
Basic QC Practices, 4th Edition
Page 184
Westgard QC, Inc., Copyright © 2016
Page 191
Westgard QC, Inc., Copyright © 2016
Risk Management
The purpose is to identify specific susceptibilities, or failure modes,
that may cause errors in test results. Manufacturers attempt to
identify those failure modes in the development of their test systems,
eliminate them through design improvements and scheduled
preventive maintenance, and integrate procedural controls or
functional checks to detect failures and alert the operators to
problems. Such manufacturer controls should always be included in
any QC Plan, but additional controls may still be needed because of
the operating conditions for a test system in your laboratory. The
need for these other controls is the focus of risk-based QC Plans and
the application of Risk Management concepts and principles to the
Total Testing Process in the medical laboratory.
Risk Management is widely used in industry, including the
manufacturers of diagnostic test systems that are intended for use in
medical laboratories. The fundamental guidance for manufacturers
of medical devices is found in ISO 14971 [2] and supplemented by
CLSI EP18A2 [3], the latter being aimed at both device manufac-
turers and medical laboratories and is more readily available and
readable by laboratory analysts. The heart of Risk Management
Page 207
Westgard QC, Inc., Copyright © 2016
Mitigating risks
Mitigation is the term that is used to describe an action to
eliminate, lower, or prevent risk. The strategies for reducing risk
depend on the particular risk factor that is being targeted. For
example, manufacturers design their test systems to ensure safety
characteristics (e.g., precision, bias, reportable range, etc.) are
appropriate for the intended clinical use, plus provide information
for safe use of test results (specimen types, reference ranges,
possible interferences, sample requirements, known limitations,
etc.). Manufacturers reduce occurrence by making improvements
that eliminate failure modes during the design and production of
new test systems and by recommending preventive and corrective
actions in their instructions for use. In situations where occurrence
of failures cannot be entirely eliminated or prevented, manufacturers
add controls to detect and alert the operators of potentially harmful
conditions.
Laboratories have fewer options for mitigating risks. Figure
17-1 outlines the laboratory strategies:
• First, for occurrence, the laboratory must confirm that safety
characteristics are appropriate by performing validation
studies. To accomplish this, the laboratory should define
the quality required for the intended clinical use, evaluate
imprecision and bias, and calculate a Sigma-metric to assess
quality on the Sigma scale. Note that a laboratory cannot
modify a manufacturer’s method or instructions for use unless
Page 221
Westgard QC, Inc., Copyright © 2016
Page 233
Westgard QC, Inc., Copyright © 2016
Validate Safety
Characteristics
Determine Sigma Quality
Audit QC Practices
Analyze
2 levels of
controls/day?
Individualized QC Plan
Total QC Plan
(1) Risk Assessment
Right-Sized SQC
(2) QC Plan
Pre- & post-analytic controls
(3) QA Plan
Page 245
Westgard QC, Inc., Copyright © 2016
Page 253
Westgard QC, Inc., Copyright © 2016
Page 271
Westgard QC, Inc., Copyright © 2016
Page 291
Westgard QC, Inc., Copyright © 2016
Common Complaints
Page 309
Westgard QC, Inc., Copyright © 2016
Index CLSI 169
CLSI C24-A3 22, 104–105, 156, 181, 285
Assay Control Materials 124
Symbols cumulative mean and SD 125
Establish Mean for New Lot 128
1:2s control rule definition 20, 47, 49 Establish SD for new Lot 128
1:3s control rule definition 20, 47 imprecision 118–119
2:2s control rule definition 21, 48 interlaboratory QC programs 253
2of3:2s control rule definition 50 CLSI EP18A2 207, 282
3:1s control rule definition 50 CLSI EP23A 176, 177, 180, 208, 213, 222, 223
4:1s control rule definition 48 , 226, 233, 247, 282
6:x control rule definition 50 CLSI EP26A 241, 243
7:T control rule definition 51 CLSI GP26 9
8:x control rule definition 49 CLSI guidance 175
9:x control rule definition 51 CLSI HS1-A2 17, 272, 274
10:x control rule definition 49 CLSI “source of errors matrix” 292
ΔSEcrit 142 Coefficient of variation, CV 122
cumulative 126
A Coefficient of variation index, CVI 131, 257
COLA 9
Across materials 60 Collection and processing of specimens 199
Across run 61, 73 Commutability 226
Allowable bias 11 Continuous mode 158–159
Allowable imprecision 11 Control charts 16, 19, 32–41
Allowable total error 11, 140. 284 Control concentration level 99
Analytical QC Plan 283 Control limits 19, 27, 32, 117, 123, 181, 314
Analytical Run 22, 156 cumulative 127
Analytic Controls 200 Control materials 25–26, 123
Analyzer Acceptability 200 Assayed versus unassayed 108
Assay control limits 117 Assigned values 124
Assayed values 124 concentration levels 108
Automation 297–299 lot changes 127–128
Matrix 104–106
B Pre-treatment steps 106
Stability 107
Bad control habits 78–79 Vendor considerations 109–110
Batch mode 157 Vial-to-vial variability 107–108
Bottle values 123–124, 314 Control mechanisms 224
Control repeats 78, 96
C Control results 41–42
Control rules 20, 100
CAP 9, 179, 180, 240 Controls (see also control materials) 153–160
Cartridge stability 229 Control stability 99
Changing control lots 127 Corrective action 281
Charts of Operating Specifications ( see also OP- Critical Difference 241
Specs charts) 321 Customer complaints 201
CLIA 9, 21, 117, 191, 197, 204, 208, 229, 233 CVI 257, 258
, 246, 259, 273
run length 154–155
Page 329
Basic QC Practices, 4th Edition
Page 330
Westgard QC, Inc., Copyright © 2016
Page 331
Basic QC Practices, 4th Edition
Page 332
Westgard QC, Inc., Copyright © 2016
within material 60
within run 60
within-run 60
Y
Youden plot 257
Youden Plot 264, 266
Z
z-score 129
Page 333