College of American Pathologists
Printable Version

  Feature Story


cap today

Creating a common language for labs

July 2003
Karen Southwick

You’d think that laboratorians from New Jersey to New Guinea
would speak the same language by now, at least when it comes to professional terms, but you’d be wrong.

The United States, European nations, and other countries, such as Japan, define the same basic words very differently in some cases, and the division is creating problems for vendors, regulatory and standards-setting agencies, journal editors, patients, and clinicians.

Although this is a longstanding problem, our increasingly global society has made it worse. Not only are vendors selling the same products in many countries, but patients travel internationally and expect to get the same diagnostic test results in Frankfurt as in Honolulu. Clinicians are practicing across borders; laboratories are becoming regional, national, and even international in scope.

It all adds up to a need to harmonize terms, such as sensitivity, specificity, accuracy, trueness, analyte, specimen, and error, so the meaning is consistent throughout the world.

“It’s a big problem growing bigger because the marketplace [for diagnostic testing] has evolved into a global marketplace,” says Steven Gutman, MD, director of the FDA’s Office of In Vitro Diagnostic Device Evaluation and Safety.

Dr. Gutman, a champion of harmonization, says the United States and Europe, in particular, took different paths in their approach to the use of common diagnostic terms. “One of the major differences is that our regulatory model is equivalency based, while the European model is standards based,” Dr. Gutman says.

That means the United States tends to be more reactive to established practice in setting regulatory terms, while European regulatory bodies and international/European standards groups such as ISO and CEN “are very much oriented toward precise definitions of analytical performance,” he explains.

For example, “accuracy” in the United States generally means agreement between the mean or best estimate of a quantity and its true value, and it is a measure of systematic but not random analytic error. In Europe, “accuracy” is a measure of both systematic and random analytic error, while “trueness” is a measure of systematic error and corresponds to the U.S. definition of accuracy.

The difference between analytical and diagnostic sensitivity and specificity “is a problem on both sides of the ocean,” Dr. Gutman says. “It’s a matter of trying to clarify which one you mean.” The term “analytical sensitivity,” for example, is considered by some to represent the smallest amount of a substance that can be measured in a sample and by others to represent the slope of the calibration function, while “diagnostic sensitivity” refers to the percentage of people with a disease who test positive for it with a given test.

Although the terminology gap is particularly taxing to vendors, because they must prepare separate package labels for the same product when it is sold in different countries, it can also create headaches for organizations such as the CAP, which accredits labs worldwide and provides them with proficiency tests, says Gerald Hoeltge, MD, head of transfusion medicine at the Cleveland Clinic Foundation and CAP representative to the NCCLS.

“It’s expensive for vendors. They’d like to make one set of instructions as a package insert in every country,” Dr. Hoeltge says. “It’s also a significant issue for the lab accreditation and Surveys programs for CAP because they deal with labs all over the world.”

If a CAP resource committee writes up participants’ results from a Survey in U.S. terms, participants in other countries could be confused, he explains. The checklist questions for the CAPLaboratory Accreditation Program are in English, bypassing the translation barrier, but the terms are U.S.-based. “If the wrong words are being used for someone in another country, it complicates and confuses,” Dr. Hoeltge says.

Take a laboratory in Germany that’s CAP-accredited but performs testing in an area where the CAP doesn’t have a proficiency test. “They will be submitting their data based upon a European proficiency testing or audit program,” he notes. “That puts the CAP in the position of interpreting another country’s survey summary. If everybody were using the same terms in the same way, life would be much simpler.”

Vendor groups and standards bodies on both sides of the Atlantic are trying to improve the situation by rewriting package labeling and standards, but the process will probably take a generation to complete because textbooks and training must also be revised.

NCCLS leads the way

Leading the effort is the NCCLS, which oversees laboratory standards in the United States and serves as administrator of ISO/TC 212, an international technical committee working on standardization in laboratory medicine. “If you’re going to have cost-effective services and products, you need globally applicable standards to minimize confusion,” says John V. Bergen, PhD, executive director of the NCCLS. When it comes to nomenclature, “the big outlier is the U.S.,” he adds.

Three harmonization programs are underway at the NCCLS.

First is the Harmonized Terminology Database, developed by the executive staff to give direction about nomenclature for the clinical laboratory and in vitro diagnostics to the volunteers who write NCCLS standards and to the editors and writers of professional journals. “It will let people know the most internationally accepted definitions of common terms,” Dr. Bergen says. “The goal is to see if, over time, the editors of professional journals could put statements in their instructions to authors encouraging them to use the database.”

The second is ensuring that, as new standards are developed and existing ones revised, the language in them is harmonized. This NCCLS revision process is handled through the Chairholders Council, the volunteer body that oversees standards projects, with the assistance of a special harmonization team drawn from industry, regulatory bodies, professional societies, and other interested groups.

“You can’t do this overnight,” Dr. Bergen cautions. The NCCLS oversees 150 published standards, each of which is supposed to be reviewed every three to seven years. “Within five years we should be through most of the documents,” he says.

Any time the NCCLS considers a new standard, “we talk to the team about the need for harmonization and encourage them to identify a member who will be the harmonization champion and look over the document to identify terminology issues,” Dr. Bergen says.

NCCLS president Donna Meyer, PhD, director of community health for Christus Health System, headquartered in Dallas, says that where appropriate in new documents, “we will put in synonyms for terms” so different groups of readers will understand them. She singles out the European concept of trueness as one instance where that would be appropriate; another is the use of the term analyte (measurand in Europe). “These will not be geared just to a U.S. audience,” she says. The synonyms probably will be used for a few years to ease the transition to harmonized terms.

The third harmonization program involves outreach to editors and writers of medical journals, textbooks, and other materials through a task force formed by the NCCLS board’s Global Programs Committee and Membership and Marketing Committee. “We’re developing a communication plan for talking with professional organizations and journal editors to get this [standards harmonization] into textbooks and other constituencies where tomorrow’s physicians learn,” Dr. Meyer says.

“Textbook and journal editors and writers are vital to the process,” she says. “It doesn’t do any good to put out standards documents that people don’t understand or use.” Dr. Meyer also wants to make sure the outreach includes academic communities and practicing labs.

In all of these efforts, the NCCLS will adopt ISO-accepted meanings for terms such as sensitivity, specificity, and accuracy, Dr. Bergen says. That’s because ISO is an international organization composed of member countries spanning the globe.

“That doesn’t mean there won’t still be national standards” for specific uses, he adds. For example, laboratory waste practices in the United States are “linked toU.S. regulatory requirements and require U.S.-focused guidelines to help labs,” Dr. Bergen says.

Dr. Meyer notes that the effort to harmonize will have to filter through “the communities we serve” before it becomes effective. Despite a big push to introduce the metric system in the United States, “How far have we gotten?” she asks, to illustrate the difficulties groups like the NCCLS face in attempting harmonization.

Vendors chime in

Vendors on both sides of the Atlantic, however, are pushing for harmonization. In the United States, Kay Setzer, manager of external standards for Abbott Laboratories, is chairing a working group—part of ISO/TC 212—that is developing global standards for labeling in vitro diagnostic products.

“Labeling has become a big problem because countries tend to put their own meanings in their regulations,” Setzer says. “If you make a product you want to sell globally, you have all these separate requirements to satisfy.” Even the designation for the product information differs—the United States calls it “labeling,” while Europe calls it “information supplied by the manufacturer.”

More than 45 countries have some regulations on in vitro diagnostic products, and “every year more are added to the list,” Setzer says. “Every year there’s more confusion on what needs to be put on the labeling.”

Setzer’s working group is trying to harmonize “exactly what is required on a label, a package insert, or the actual container—both the terms themselves and the parameters of what needs to be included,” she says. Industry representatives and laboratory professionals from the United States, Japan, Canada, Korea, and Europe participate.

As Setzer’s group finishes its standards—four are underway now—it circulates them for review within ISO/TC 212. The committee’s approximately 35 member countries “will have comments and additions,” she says. “Then ISO will vote on the final document to be accepted as an ISO standard.”

Setzer’s group is also working with the Global Harmonization Task Force (, which is composed of regulators (including Dr. Gutman) from the United States, Japan, Canada, Europe, and Australia. “They’re trying to set down a group of requirements for medical devices,” she says, including the best way to regulate labeling.

The hope is that the regulatory bodies within the countries will adopt the ISO standards developed by her working group. “We want everybody to follow those standards,” Setzer says. “If the regulatory bodies adopt them, that would be a great step forward.”

Like others involved in harmonization, Setzer warns that there are no quick fixes. She hopes ISO will adopt her working group standards by early 2005, if not before. “Then they will be available to regulatory bodies like the FDA,” she says. “It’s a long-term process.”

In Europe, Emil Voelkert, PhD, senior vice president of Roche Diagnostics, Mannheim, Germany, chairs the European Diagnostic Manufacturers Association (EDMA) working party on standardization. “Harmonization of terms is one aspect of our work,” he says. “In standards development, we see a definite trend away from national to regional and international standards.”

ISO 15189, for example, provides a global basis for accrediting labs. “Standards in the ISO world are horizontal, meaning they provide a general framework without going too much into specifics,” he says.

Dr. Voelkert agrees with Setzer that industry needs a common basis for marketing identical products. “Quality should have the same facets anywhere,” he adds. “And for the users and patients, the harmonization of concepts is providing the basis for better compatibility of results over regions.”

The EDMA working party monitors and provides input to a number of standardization projects, including those in ISO/TC 212; ISO/TC 210, which deals with general medical devices; and CEN/TC 140, which deals with in vitro diagnostic medical devices.

The EDMA also is monitoring efforts by such groups as the Global Harmonization Task Force and Setzer’s group and is providing input from industry with the European perspective.

Another important international endeavor, says Dr. Voelkert, is the Joint Committee on Traceability in Laboratory Medicine (JCTLM), which aims to establish internationally comparable reference systems. This represents “another facet of harmonization and will provide the basis for comparable results over region and time,” he says. For example, “the results for glucose or cholesterol in the United States would be the same as in Europe, Canada, Australia, Japan, and all other countries.”

Dr. Voelkert believes all these efforts, including those of the NCCLS, “are on the right track.” As far as harmonization of terms is concerned, there are, in reality, only 15 to 20 critical terms “where we need to get acceptance, particularly in the English-speaking world.” Of course, “once you devise a solution, you have to implement it,” he says, but he believes that will happen, albeit slowly. “I am optimistic,” he adds.

Acceptance will be gradual
Part of the process of accepting new standards or harmonizing terms will fall to medical journals. Kenneth McClatchey, MD, DDS, editor of the Archives of Pathology & Laboratory Medicine, says he’s prepared to support the NCCLS’ efforts, including using the new database.

Harmonization was a problem “even before I was a pathologist,” says Dr. McClatchey, professor of pathology at Loyola University Medical Center, Chicago. “NCCLS is the driving force because it administers ISO/TC 212,” he says. ISO’s new international standard on clinical lab management, 15189, should propel labs toward harmonization, he adds.

Dr. McClatchey plans to refer authors to the NCCLS database “to make sure the terminology they’re using is in sync with international standards.” Even with articles in the United States, he says, “we’re consistently dealing with terminology that one person uses one way and another person a different way.”

The CAP, which has a close relationship with the NCCLS, should take a leadership role in the harmonization effort, asserts Dr. McClatchey. “We in the U.S. will have to deal with international standardization. The globe is shrinking,” he says. “The lab business is an international business as much as manufacturing [of diagnostics] is an international business. We need an international understanding of terms and methodologies.”

Professional organizations, as well as government and industry groups, need to back the NCCLS, he adds. The CAP has an inherent interest because harmonization “will benefit our external quality assessment programs.”

David Bruns, MD, editor of Clinical Chemistry and professor of pathology at the University of Virginia, Charlottesville, agrees that standardizing terms is important in journal literature.

For instance, the term sensitivity (meaning analytical sensitivity) can be confusing because “a lot of people think it means the detection limit when really it is the slope of the calibration curve,” he says. “This is an example of where we have a Tower of Babel.”

Consequently, Dr. Bruns has banished the word sensitivity from his journal. Instead, he asks writers to use “detection limit,” which is “almost always what they mean—how much of the substance you need to get a measurable signal.”

Dr. Bruns approaches “analytical specificity” the same way. “The author needs to spell out exactly what was done,” he says. “If you have an assay for insulin, if it’s specific, it’s not going to react with other similar hormones.” Here again, he prefers that authors use more precise terms, like “cross-reactivity.”

As for the NCCLS database, Dr. Bruns believes it could be useful, “but we’d have to see how it comes out.” He hopes the NCCLS will open the database to comment before deciding on a final version. “You have to involve editors in the process if you want them to use it,” he says.

The database is a “living” work in progress, responds the NCCLS’ Dr. Bergen, and comments intended to improve the interface’s utility and user friendliness will always be welcome. “We can also review and process input on specific terms and definitions,” he says. But because the purpose of the database is to promote the use of internationally standardized terminology, says Dr. Bergen, “such [outside] input would need to be referred to and formally considered within the international standards system.”

In Dr. Hoeltge’s view, it will be a generation before everyone is on the same page in laboratory and diagnostic lexicon. That’s because the transition won’t occur entirely until “pathology residents and young clinical scientists become tomorrow’s leaders.” For now, Dr. Hoeltge communicates with his clinicians at the Cleveland Clinic in ways they understand. “If you didn’t use a term in training,” he says, “it’s hard to change.”

Karen Southwick is a writer in San Francisco.