Fundamental Concepts

In this chapter, quality is defined as conformance with the requirements of users or customers. More directly, quality refers to satisfaction of the needs and expectations of users or customers. The focus on users and customers is important, particularly in service industries such as healthcare. Users of healthcare laboratories are often nurses and physicians; their customers are patients and other parties who pay the bills.

Cost must be understood in the context of quality. If quality means conformance with requirements, then quality costs must be understood in terms of “costs of conformance” and “costs of nonconformance,” as illustrated in Figure 8-1. In industrial terms, costs of conformance are divided into prevention costs and appraisal costs. Costs of nonconformance consist of internal and external failure costs. For a laboratory testing process, calibration is a good example of a cost incurred to prevent problems. Likewise, quality control is a cost for appraising performance, a repeat run is an internal failure cost for poor analytical performance, and repeat requests for tests because of poor analytical quality are an external failure cost.

This understanding of quality and cost leads to a new perspective on the relationship between them. Improvements in quality can lead to reductions in cost. For example, with better analytical quality, a laboratory would be able to reduce waste; this, in turn, would reduce cost. The father of this fundamental concept was the late W. Edwards Deming, who developed and internationally promulgated the idea that quality improvement reduces waste and leads to improved productivity, which, in turn, reduces costs and provides a competitive advantage.⁴² As a result, the organization stays in business and is able to continue providing jobs for its employees.

Fundamental Principles

Quality improvement occurs when problems are eliminated permanently. Industrial experience has shown that 85% of all problems are process problems that are solvable only by managers; the remaining 15% are problems that require the action and improvement in performance of individual workers. Thus quality problems are primarily management problems because only management has the power to change work processes.

This emphasis on processes leads to a new view of the organization as a system of processes (Figure 8-2). For example, physicians might view a healthcare organization as a provider of processes for patient examination (A), patient testing (B), patient diagnosis (C), and patient treatment (D). Healthcare administrators might view the activities in terms of processes for admitting patients (A), tracking patient services (B), discharging patients (C), and billing for costs of service (D). Laboratory directors might understand their responsibilities in terms of processes for acquisition of specimens (A), processing of specimens (B), analysis of samples (C), and reporting of test results (D). Laboratory analysts might view their work as processes for acquiring samples (A), analyzing samples (B), performing quality control (C), and releasing patient test results (D). The total system for a healthcare organization involves the interaction of all of these processes and many others. Given the primary importance of these processes for accomplishing the work of the organization, TQM views the organization as a support structure rather than as a command structure. As a support structure, the most immediate processes required for delivery of services are those of frontline employees. The role of upper management is to support the frontline employees and to empower them to identify and solve problems in their own work processes.

The importance of empowerment is easily understood if a problem involves processes from two different departments. For example, if a problem occurs that involves the link between process A and process B in Figure 8-2, the traditional management structure requires that the problem be passed up from the line workers to a section manager or supervisor, a department director, and an organization administrator. The administrator then works back through an equal number of intermediaries in the other department. Direct involvement of line workers and their managers should provide more immediate resolution of the problem.

However, such problem solving requires a carefully structured process to ensure that root causes are identified and proposed solutions verified. Juran’s “project-by-project” quality improvement process⁵³ provides detailed guidelines that have been widely adopted and integrated into current team problem-solving methods.^11,12,97 These methods outline distinct steps for (1) carefully defining the problem, (2) establishing baseline measures of process performance, (3) identifying root causes of the problem, (4) identifying a remedy for the problem, (5) verifying that the remedy actually works, (6) “standardizing” or generalizing the solution for routine implementation of an improved process, and (7) establishing ongoing measures for monitoring and controlling the process.

The quality improvement project team provides a new flexible organization unit. A project team is a group of employees appointed by management to solve a specific problem that has been identified by management or staff. The team comprises members from any department and from any level of the organization and includes anyone whose presence is necessary to understand the problem and identify the solution. Management initiates the project, and the team is empowered and supported to identify the root cause and verify a solution; management then becomes involved in replanning the process (i.e., planning the implementation of changes in a laboratory process, defining and standardizing the improved process, and establishing appropriate measures for ongoing evaluation and control of the process).¹²⁹

Number of Errors Made in the Clinical Laboratory

A study of 363 incidents captured by a laboratory’s quality assurance program in a hospital enumerated the sources and impact of errors.⁹² Incidents included those in which (1) physicians’ orders for laboratory tests were missed or incorrectly interpreted; (2) patients were not properly prepared for testing or were incorrectly identified; (3) specimens were collected in the wrong containers or were mislabeled or mishandled; (4) the analysis was incorrect; (5) data were entered improperly; or (6) results were delayed, not available, or incomplete, or they conflicted with clinical expectations. Upon evaluating the data, the authors found no effect on patient care for 233 patients; 78 patients were not harmed but were subjected to an unnecessary procedure not associated with increased patient risk; and 25 patients were not harmed but were subjected to an additional risk of inappropriate care. Of the total number, preanalytical mistakes accounted for 218 (45.5%), analytical for 35 (7.3%), and postanalytical for 226 (47.2%). Nonlaboratory personnel were responsible for 28.6% of the mistakes. An average of 37.5 patients per 100,000 treated were placed at increased risk because of mistakes in the testing process.

Witte and colleagues investigated rates of error within the analytical component and found that widely discrepant values were rare, occurring in only 98 of 219,353 analyses.¹⁴⁵ When these results were converted into a standard metric of errors per million episodes, an error rate of 447 ppm was calculated.¹³ In another study, Plebani and Carraro identified 189 mistakes from a total of 40,490 analyses, with a relative frequency of 0.47% (4667 ppm). The distribution of mistakes was 68.2% preanalytical (3183 ppm), 13.3% analytical (620 ppm), and 18.5% postanalytical (863 ppm).⁸⁶ Most laboratory mistakes did not affect patients’ outcomes, but in 37 patients, laboratory mistakes were associated with additional inappropriate investigations, thus resulting in an unjustifiable increase in costs. In addition, laboratory mistakes were associated with inappropriate care or inappropriate modification of therapy in 12 patients. The authors concluded that “promotion of quality control and continuous improvement of the total testing process, including pre-analytical and post-analytical phases, seems to be a prerequisite for an effective laboratory service.”⁸⁶

In a study of common immunoassays, Ismail and colleagues found only 28 false results from 5310 patients (5273 ppm).⁵¹ However, as a result of incorrect immunoassay results attributable to interference, 1 patient had 15 consultations, 77 laboratory tests, and an unnecessary pituitary computed tomography scan. The authors stress (1) the necessity for good communication between clinician and laboratory personnel, (2) the importance of the clinical context, and (3) the necessity for use of multiple methods of identifying erroneous test results—a necessity for a rigorous and robust quality system. Heterophilic antibody blocking studies were most effective in identifying interference, but in 21% of patients with false results, dilution studies or alternative assays were necessary to identify the problem. In a similar study, Marks enlisted participation from 74 laboratories from a broad international spectrum of settings and found that 6% of analyses gave false-positive results and, as in the Ismail study, found that use of a heterophilic blocking reagent corrected approximately one third of these.⁷² Further evaluation of the data showed no consistent pattern for false results: errors were distributed across donors, laboratories, and systems of analysis. In reviewing the data from these last two studies, Leape suggested setting up a system that would ensure that every result was given a rigorous review before being reported.⁶⁴

Bonini et al conducted several MEDLINE studies of laboratory medical errors and found large heterogeneity in study design and quality and lack of a shared definition of laboratory error.¹⁷ However, even with these limitations, they concluded that most such errors occur in the preanalytical phase and suggested that these could be reduced by the implementation of a more rigorous method for error detection and classification and the adoption of proper technologies for error reduction. Thus current QA programs that monitor only the analytical phase of the total process have to be expanded to include both preanalytical (see Chapter 6) and postanalytical phases (www.westgard.com/essay34/assessed March 22, 2011). Through expanded monitoring, the total process would then be managed so as to reduce or eliminate all defects within the process.⁸⁶

Six Sigma Principles and Metrics

Six Sigma,41A,46A,48 is an evolution in quality management that is being widely implemented in business and industry in the new millennium.⁸⁹ Six Sigma metrics are being adopted as the universal measure of quality to be applied to their processes and the processes of their suppliers. The principles of Six Sigma go back to Motorola’s approach to TQM in the early 1990s and the performance goal that “6 sigma’s or 6 standard deviations of process variation should fit within the tolerance limits for the process,” hence, the name Six Sigma (http://mu.motorola.com/accessed March 22, 2011). For this development, Motorola won the Malcolm Baldridge Quality Award in 1988.

Six Sigma provides a more quantitative framework for evaluating process performance and more objective evidence for process improvement. The goal for process performance is illustrated in Figure 8-4, which shows the tolerance specifications or quality requirements for that measurement set at −6S and +6S. Any process can be evaluated in terms of a sigma metric that describes how many sigma’s fit within the tolerance limits. The power of the sigma metric comes from its role as a universal measure of process performance that facilitates benchmarking across industries.

Two methods can be used to assess process performance in terms of a sigma metric (Figure 8-5). One approach is to measure outcomes by inspection. The other approach is to measure variation and predict process performance. For processes in which poor outcomes can be counted as errors or defects, the defects are expressed as defects per million (DPM), then are converted to a sigma metric using a standard table available in any Six Sigma text.⁴⁸ This conversion from defects per million to sigma levels is an enumeration of the area under the error curve plus or minus the tolerance limits (±2 S = 308,500 DPM; ±3 S = 66,800 DPM; ±4 S = 4350 DPM; ±5 S = 230 DPM; ±6 S = 3.4 DPM). In practice, Six Sigma provides a general method by which to describe process outcomes on the sigma scale.

To illustrate this assessment, consider the rates of malfunction for cardiac pacemakers. Analysis of approved annual reports submitted by manufacturers to the Food and Drug Administration (FDA) between 1990 and 2002 revealed that 2.25 million pacemakers were implanted in the United States. Overall, 17,323 devices were explanted because of confirmed malfunction.⁷⁰ The defect rate then is estimated at 7699 DPM (17,323/2,250,000), or 0.77%, which corresponds to a sigma of 3.92 using a DPM-to-sigma conversion calculator (http:www.isixsigma.com/sixsigma/six_sigma_calculator.asp?m=basic/accessed March 22, 2011). For comparison or benchmarking purposes, airline baggage handling has been described as 4.15 sigma performance, and airline safety (0.43 deaths per million passenger miles) as better than Six Sigma performance. A defect rate of 0.033% would be considered excellent in any healthcare organization, where error rates from 1 to 5% are often considered acceptable.¹³ A 5.0% error rate corresponds to a 3.15 sigma performance, and a 1.0% error rate corresponds to 3.85 sigma. Six Sigma shows that the goal should be error rates of 0.1% (4.6 sigma) to 0.01% (5.2 sigma) and ultimately 0.001% (5.8 sigma).

The first application describing sigma metrics in a healthcare laboratory was published by Nevalainen et al⁷⁹ in the year 2000. This application focused on preanalytical and postanalytical processes. Order accuracy, for example, was observed to have an error rate of 1.8%, or 18,000 DPM, which corresponds to 3.6 sigma performance. Hematology specimen acceptability showed a 0.38% error rate, or 3800 DPM, which is a 4.15 sigma performance. The best performance observed was for the error rate in laboratory reports, which was only 0.0477%, or 477 DPM, or 4.80 sigma performance. The worst performance was therapeutic drug monitoring timing errors of 24.4%, or 244,000 DPM, which is 2.20 sigma performance.

Of the studies discussed in the previous section, it is possible to convert the error rates computed in DPM to sigma metrics. For example, for the Ross-Boone study,⁹² the computed DPM corresponds to a 3.3 sigma long-term performance. For the Plebani et al study⁸⁶ a DPM of 620 DPM corresponds to a 3.2 sigma long-term performance. In the Ismail et al study,⁵¹ a DPM of 5273 corresponds to a 2.6 sigma long-term performance. On average, this indicates about 3.0 sigma long-term performance.

The application of sigma metrics for assessing analytical performance depends on measuring process variation and determining process capability in sigma units.95,110,126 This approach makes use of the information on precision and accuracy that laboratories acquire initially during method validation studies and have available on a continuing basis from internal and external quality control. An important aspect of this method is that the capability, or predictive performance, of the process must be ensured by proper quality control; therefore the ease of assessment comes with the responsibility to design and implement QC procedures that will detect medically important errors.

To apply this method, the tolerance limits are taken from performance criteria for external quality assessment programs or regulatory requirements [such as the U.S. Clinical Laboratory Improvement Amendment (CLIA) criteria for acceptable performance in proficiency testing]; process variation and bias can be estimated from method validation experiments, peer-comparison data, proficiency testing results, and routine QC data. For laboratory measurements, it is straightforward to calculate the sigma performance of a method from the imprecision: SD or CV and inaccuracy (bias) observed for a method and the quality requirement (allowable total error, TE_a) for the test [Sigma = (TE_a − bias)/SD]. For a cholesterol test with an NCEP total error of 9%, method bias of 1.0%, and method CV of 2.0%, the sigma metric is 4.0 [(9.0 − 1.0)/2]. If the method had a CV of 3.0% and a bias of 3.0% (the maximum allowable figures according to NCEP guidelines), the sigma metric is 2.0. Sigma metrics from 6.0 to 3.0 represent the range from “best case” to “worst case.” Methods with Six Sigma performance are considered “world class”; methods with sigma performance less than 3 are not considered acceptable for production.

Those conclusions can be readily understood by considering the amount of quality control that is necessary for measurement processes having different performance metrics. Figure 8-6 shows a power function graph that describes the probability of rejecting an analytical run on the y-axis versus the size of the systematic error that has to be detected on the x-axis. The bold vertical lines correspond to methods having 3, 4, and 5 sigma performance (left to right). The different lines or power curves correspond to the control rules and the number of control measurements given in the key at the right (top to bottom). These different QC procedures have different sensitivities or capabilities for detecting analytical errors. Practical goals are to achieve a probability of error detection of 0.90 (i.e., a 90% chance of detecting the critically sized systematic error), while keeping the probability of false rejection at 0.05 or less (i.e., 5% or lower chance of false alarms). This is easy to accomplish for processes with 5 to 6 sigma performance; it requires more careful selection and increased QC efforts for processes from 4 to 5 sigma, and it becomes very difficult and expensive for processes less than 4 sigma.

As demonstrated previously, the application of Six Sigma principles and metrics is very valuable for all phases of the laboratory testing process.41A,46B Because the core business of the laboratory is to produce accurate test results, it makes sense to first apply Six Sigma to the analytical processes. This also is the easiest application because there are tolerance limits in the form of acceptability criteria from peer-comparison and proficiency-testing programs, QC data available for estimating method precision, and peer data available for estimating method bias. Laboratories should next expand their efforts to the preanalytical and postanalytical processes, knowing that their core processes are producing the necessary analytical quality.

Commitment

Dedication to quality service must be central and reflect a team effort. Otherwise, quality goals are not likely to be achieved. Quality must be a major consideration in all management decisions because any single decision may compromise other plans and practices for attaining quality goals. A true commitment is required by laboratory directors, managers, and supervisors if the efforts of other laboratory personnel are to be successful.

Facilities and Resources

Laboratories must have the administrative support necessary to provide the quality of services desired. This means having (1) adequate space, (2) equipment, (3) materials, (4) supplies, (5) staffing, (6) supervision, and (7) budgetary resources. These resources provide the basis upon which quality services are developed and maintained.

Personnel Competency and Training*

People are critical components of a total quality system, and training and education are vital to the performance of personnel. A key factor in successful training and assessment of laboratory staff is the planning and implementation of targeted education programs. Given the shortage of clinical laboratorians in many countries, laboratories are providing in-depth training for both new and existing staff and are assessing their competency on an ongoing basis. In the United States, CLIA requires competency assessment as part of the laboratory quality system. In other countries, ISO 15189 requires competency assessment as part of the technical requirements for medical laboratories.

CLIA identifies the following six elements as required components of a laboratory competency assessment program: (1) direct observation of routine patient test performance; (2) monitoring of the recording and reporting of test results; (3) review of intermediate test results, QC records, proficiency testing results, and preventive maintenance records; (4) direct observation and performance of instrument maintenance and function checks; (5) assessment of test performance through testing previously analyzed specimens, internal blind testing samples, or external proficiency testing samples; and (6) assessment of problem-solving skills.⁸⁷

Assessment of competence in job tasks as required by CLIA must be conducted semiannually the first year of employment and annually thereafter, and upon implementation of new test methods before patient test results are reported. Guidelines to assist in the development and documentation of competency assessment are available from the Clinical and Laboratory Standards Institute (CLSI). The CLSI Guideline, Training and Competence Assessment, gives detailed instructions on how to develop and implement a training and competency assessment program that meets regulatory requirements, and provides examples of forms for documentation and record keeping.³⁸

Design of an in-service training program based on instructional systems design includes the following elements: (1) analysis, (2) design, (3) implementation, and (4) evaluation. It begins with a needs assessment or gap analysis to determine employee performance requirements, to identify deficiencies, and to evaluate existing education and training resources. It requires the development of measurable instructional objectives that are based on the specific skills and competencies required of the employee to perform the job or task, and selection of an appropriate teaching strategy. The in-service training program also considers how an instructional program will be delivered and a range of organizational factors that may impact the successful delivery of the instruction. These include (1) employee participation, (2) scheduling, (3) availability of subject matter experts to teach, (4) budget constraints, and (5) assessment of learning outcomes. It provides evaluation of the effectiveness of the instructional program.³⁹

A workforce deficiency sometimes found with laboratory personnel is lack of academic education in basic quality control practices and statistical methods of analysis. Without formal training in statistical quality control for the clinical laboratory, employees are often taught what to do, but may not understand the “why,” which can impact their ability to troubleshoot and problem-solve. This background understanding is difficult to teach in short, “in-service” laboratory sessions and generally is better provided in a more structured program. Implementing an in-house education program in basic quality control practices that is dependent on the availability of subject matter experts to provide the instruction may not be feasible. A cost-effective, alternative approach is to access experts through existing Web-based training programs and complement this with local problem-oriented examples.

In the face of increased pressures to reduce operating costs, including expenses associated with attendance at and travel to conferences, Internet education programs provide an effective, cost-efficient way to implement in-service training. Web-based training programs in quality control concepts are available through both professional organizations and private companies. As an example, the Mayo Clinic identified that a gap in academic education in basic quality control concepts existed based on the diversity of its workforce and the varying academic backgrounds required for its highly specialized laboratories. To provide the desired level of academic education to its employees in a way that could be readily accomplished, employees were enrolled in the course Basic QC Practices available on the Web through Westgard QC Inc. The e-learning content of the curriculum includes the following modules: (1) statistical quality control, (2) construction and interpretation of Levey-Jennings control charts, (3) electronic checks and sources of error, (4) CLIA ’88 regulations for QC, (5) control materials and limitations of QC, (6) multirule and multilevel interpretation of QC data, (7) false rejection and error detection, (8) troubleshooting, (9) regulatory guidelines, (10) QC documentation and record keeping, and (11) external quality assessment programs. In addition, the online curriculum was customized to employees’ specific needs by adding staff lectures, six 2-hour laboratory sessions, assignments tailored to the clinical laboratory practice at Mayo, and pretesting and post-testing to assess competency.

Implementation of this in-house training program in basic quality control by the Mayo Clinic followed the educational model designed for the Clinical Laboratory Science Program. The didactic component is provided in an e-learning platform and is underscored by transactional distance theory,⁷⁶ whereby the three modalities of learner interaction with content, instructor, and fellow learners were integrated into the online module. Each lesson plan also includes a supplemental laboratory module taught by traditional methods of interaction between instructor and learner, which is closely anchored in the context of the work performed. Finally, the curricular model implements the reverse lecture-homework paradigm, whereby learners complete the Web-supported didactic modules asynchronously as “homework” assignments (Figure 8-7, A) and complete the laboratory lessons in the classroom (work setting) under the guidance and direction of the instructor/supervisor (Figure 8-7, B).

The in-service training program in quality control concepts follows the CLSI academic curricular model and applies the reverse lecture-homework paradigm (see Figure 8-7). To implement the program, employees complete 14 online lessons asynchronously, as “homework,” and participate in 6 scheduled laboratory sessions taught in the traditional format, over a period of 3 weeks. The electronic curriculum includes employee interaction with the written quality control course content (“learner-content”), threaded discussions (“learner-learner”), and email (“learner-instructor”). The online material is applied directly during “hands-on” instructor-facilitated laboratory sessions that are problem-based and consist of a combination of case studies, laboratory activities, and discussions.

Implementing an e-learning platform allows the curricular model to expand the number of students over time toward an improved economy of scale. For academic programs, this approach allows for a potential increase in class size with minimal additional expenses. For in-service training, an electronic curriculum provides the opportunity to share training across different physical sites in a healthcare delivery system. This sharing will eliminate the costs associated with duplication of effort, reduce the operating cost to cover additional employees, and decrease the startup costs for new academic and in-service programs at additional sites.

Technical Procedures

Technical procedures necessary for laboratory services include the following:

Problem-Solving Mechanism

Although it is a particularly critical element in a quality assurance program, the necessity for a mechanism for problem solving is often underemphasized. Such a mechanism provides the link between identification of a problem and implementation of a solution to the problem. It is a feedback loop that responds to an error signal by making adjustments to reduce the size of the error or to prevent its recurrence. For problems limited to individual methods or instrument systems, delegation of responsibility for the systems may provide the corrective mechanism. Specialized troubleshooting skills need to be developed and improved and preventive maintenance programs instituted. For problems that occur more generally, the in-service training program can be an important part of the mechanism but often requires additional input from a QC specialist or supervisor to initiate the use of this mechanism and to help define its objectives. A different approach to problem solving is the use of quality teams that meet regularly to analyze problems and identify solutions.⁹⁷ By involving personnel, quality teams heighten interest and commitment to quality and provide a creative feedback mechanism.

The comprehensive nature of quality assurance programs and their missions, goals, and activities have been discussed in greater detail by Eilers.⁴⁴ Detailed outlines of the elements of cost management for quality assurance are available,^45,54 as are detailed recommendations by professional organizations, such as the College of American Pathologists (CAP),¹ the CLSI,^33,35–37 and the International Federation of Clinical Chemistry,^22–27 and books devoted to quality assurance practices in clinical laboratories.^29,120

Systems Analysis

The operation of the clinical laboratory consists of a series of processes, each of which has potential sources of error. Table 8-1 shows the processes that take place from the time of the physician’s initial request for a test to the time of final interpretation of the test result. This systems analysis identifies the critical processes for a typical laboratory; however, each laboratory situation is somewhat different, and additional processes and additional sources of error may be identified. It is important for each laboratory to perform a systems analysis of its own laboratory testing system to identify those areas where errors are likely to occur.

Once the processes have been documented, those that are most susceptible to error should be identified and should receive the most attention. Often, processes that lead to the greatest number of complaints, such as lost specimens or delayed results, are judged to be most important, even though other steps, such as appropriateness of test selection and the acceptability of a specimen, may be of greater importance for optimal medical care. Guidelines describing procedures for specimen handling are available from organizations such as the CLSI. Documents put forth by accrediting agencies, such as CAP,¹ Centers for Disease Control and Prevention, and state regulatory agencies, are also helpful in this regard.

Types of Preanalytical Variables

It is difficult to establish effective methods for monitoring and controlling preanalytical variables because many of these variables are outside of traditional laboratory areas (see Chapter 6). Monitoring of preanalytical variables requires the coordinated effort of many individuals and hospital departments, each of which must recognize the importance of these efforts in maintaining a high quality of service. Accomplishing such monitoring may require support from outside the laboratory, particularly from the institution’s clinical practice committee or some similar authority. Variables to consider are discussed in the following section.

Choice of Analytical Method

Selection and evaluation of analytical methods are discussed in Chapter 2. It is important to recognize, however, that the initial evaluation of a method often takes place in a setting somewhat more idealized than the production setting. Therefore, it is desirable to have a startup period in the service laboratory before test results are to be reported. This period allows time to (1) discover any additional problems, (2) develop maintenance programs that alleviate those problems, and (3) train a sufficient number of analysts to support the routine service operation.

Reference Materials and Methods

Reference materials and methods used are major factors in determining the quality and reliability of the analytical values produced by a clinical laboratory. Uriano¹¹³ and Tietz¹¹¹ have defined a hierarchy of analytical methods and reference materials that shows the relationship between them. For example, definitive methods are those of highest quality that are used to validate reference methods and primary reference materials. Primary reference materials are those of the highest quality that are used in the (1) development and validation of reference methods, (2) calibration of definitive and reference methods, and (3) production of secondary reference materials. Reference methods are used to validate field methods. Secondary reference materials are used to provide working calibrators for field methods and to assign values to control materials. Control materials are used only to monitor field methods.

Definitions of Relevance to Analytical Quality

Many organizations have developed definitions for terms of analytical relevance to the clinical laboratory. Depending on the organization that developed them and the circumstances and culture under which they were developed, definitions for the same term are often slightly different. As part of its global harmonization effort, the CLSI has developed the Harmonized Terminology Database, which is a compilation of internationally accepted terminology (http://www.clsi.org/Content/NavigationMenu/Resources/HarmonizedTerminologyDatabase/Harmonized_Terminolo.htm/accessed March 22, 2011). To use it, one clicks on the Website and then simply enters the term of interest, and definitions for that term are listed. Standard, internationally preferred terms and definitions are highlighted in blue. Terms and definitions in black are acceptable; but only if the standard internationally preferred terms and definitions are unacceptable for a certain context. Terms highlighted in red are not acceptable in the international standards community.

In general, the international standards community uses the definitions listed in the Vocabulary of Basic and General Terms in Metrology (VIM) document produced by the International Organization for Standardization (ISO). Several of these definitions that are of particular interest to clinical laboratorians are listed in Table 8-2.