Harmonization of accreditation to ISO15189

Scope of this paper

By using the historical background of ISO15189 and accreditations to it, the authors describe challenges and opportunities that they see in the further development of the standard with regard to its leading principles. The analysis does not include a study whether application of, and accreditation to, the standard indeed improves laboratory quality.

Historical background of quality systems in medical laboratories

Since the 1970s, external proficiency testing programs were set up all over the world to increase analytical quality in laboratory medicine. This has led to better analytical performance as evidenced by the publications of their results over the years [1]. It became evident as well that most errors were not analytical failures but the results of preanalytical mistakes [2]. In the United States, this has led to the Clinical Laboratory Improvement Amendments (CLIA) accepted by the US Congress with demands that go beyond scoring sufficiently in analytical external quality assessment schemes [3]. The extensiveness of the demands was based on the risk attributed to the specific test. The College of American Pathologists added quality items to their program and set up a voluntary accreditation program based on the CLIA before the US Congress made accreditation mandatory. The CLIA system was modernized in 1988 to CLIA88, and since then many specifications have been added [4]. Also, other countries started thinking about quality management systems. In Australia, the Royal College of Pathologists of Australasia has developed a solid quality management system together with the National Association of Testing Authorities [5].

In Europe, countries developed different approaches. The Scandinavian countries accepted the ISO17025 quality management system for testing laboratories [6] and added specific annexes for the medical laboratory field. In France, the Guide for Good Laboratory Practice was developed [7]. The United Kingdom has developed a specific set of quality requirements for clinical laboratories in the form of the clinical pathology accreditation (CPA) program [8]. It was also used in other countries such as Ireland and the northern part of Italy. In the Netherlands, the “Coördinatie Commissie ter bevordering van de Kwaliteitsbeheersing van het Laboratoriumonderzoek op het gebied van de Gezondheidszorg” (CCKL) (the coordinating committee for quality improvement in laboratory diagnostics in health care) set up a standard for quality management in laboratory medicine in the 1980s and an accreditation system under the same name in the 1990s [9].

In 1993, the European Communities Confederation of Laboratory Medicine (EC4) was founded as a confederation for mutual projects of European societies for clinical chemistry and laboratory medicine. Directly after the formation of EC4, a Working Group on the Harmonization of Quality Systems and Accreditation was initiated. This group concluded that aspects of preexamination and postexamination, including advice on which test was appropriate, were an essential part of laboratory medicine and therefore should be covered in the quality management system for that field. This resulted in the development of the Essential Criteria and Additional Essential Criteria [10].

At about the same time, international discussion took place about developing a specific quality standard for medical laboratories at a global level. Medical laboratory professionals felt that ISO17025 lacked specific elements that distinguished laboratory medicine from testing laboratories. Laboratory service concerning preanalytical advice on test requesting, turnaround time and post-analytical interpretation in the context of medical questions were not addressed in ISO17025, and accreditation to this standard did not even allow for addition of such aspects. A technical committee, TC212, was installed by the International Standard Organization (ISO) with the Committee for Clinical and Laboratory Standards Institute (CLSI) taking care of the secretariat. Many members of the EC4 Working Group on Accreditation took part in this technical committee. They decided to use both the ISO9001 [11] standard concerned with quality management and the ISO17025 standard for testing laboratories as the fundamental tenets. Equal attention had to be paid to preexamination, examination and postexamination aspects. The essential criteria of the EC4, the CPA standard and the CCKL criteria played a prominent role in the development of ISO15189:2003: Medical laboratories Requirements for Quality and Competence [12]. The publication of this standard was severely delayed because of a discussion about its format. The preliminary intention was to follow the process cycle of the ISO9001 standard, but the ISO central secretariat insisted it was changed into the ISO17025 format [13]. Ironically, in the current revision of ISO17025, this standard will be rearranged to the more logical, process oriented order of ISO9001.

With the acceptance of the ISO15189 standard, the need of countries for their own quality management system for laboratory medicine and for an accreditation organization to assess and accredit such system no longer existed. They could approach the National Accreditation Body (NAB) in their country to include ISO15189 in their scope. Under the umbrella of the International Laboratory Accreditation Cooperation (ILAC) in each continent, an international accreditation body organizes mutual recognition of accreditation standards by national accreditation organizations. In each country, only one NAB is allowed under a mutual agreement. In Europe, this is regulated by the European cooperation for Accreditation (EA). ISO15189 is covered by the Health Care Committee (HCC) of EA, in which apart from nearly all European Accreditation Bodies, also the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) and the European Diagnostics Manufacturers Association (now MedTech) are represented. Mutual recognition of accreditation requires harmonized requirements and an assessment procedure with comparable focus, attention and scope. This was, and still is, discussed and regulated in the EA HCC.

Apart from advocating the use of ISO15189 as most suitable for medical laboratories, the extensiveness of the scope was an important item. Quite often, the ISO17025 was used to accredit only a small number of the tests that a laboratory offered. If these tests are part of a much larger test portfolio of a laboratory, accreditation of only part of the portfolio is undesirable in the medical field. When a laboratory claims to be accredited, the patients and physicians expect all the tests and services to be covered by that system and of the required quality. For that reason, the EFLM recommended accreditation of the complete array of tests offered by that specific laboratory [14]. Based on this discussion, EA has issued the EA 4/17 policy on the scope of accreditation [15]. Although an NAB cannot legally refuse to accredit a small portion of the tests offered by a laboratory, they will try to convince such laboratories to seek accreditation for the complete portfolio. This is in line with the way the accreditation of medical laboratories was set up in the UK by CPA and in the Netherlands by CCKL. It is in line with the situation in Australia and the United States.

Because ISO15189 does not give special attention for point-of-care testing, ISO 22870:2006 “Point-of-Care-Testing (POCT) Requirements for Quality and Competence” was published [16]. After a minor revision in 2007, ISO15189 was completely revised in 2012 with more focus on risk management, and it clarified the differences between validation and verification [17]. Apart from this, the vocabulary was revised in order to prevent unintended prescription and to avoid the need of an explanatory guideline. Since then, ISO15189 has become mandatory in France and for some specific tests in molecular biology and prenatal screening in some European countries. Adoption of ISO15189 is low in countries where a national accreditation program based on a different standard is mandatory (e.g. the United States). In most of the countries that have started with ISO17025, ISO15189 has taken over for laboratory medicine, except for Switzerland. An overview of accreditation trends and current status of ISO15189 in Europe is given in the publications by Huisman [18] and Boursier et al. [19].

Leading principles in ISO15189

Initially, ISO15189 was developed as a standard to serve as a fundamental tenet for the quality management systems in medical laboratories [13]. During its development, the specification on what has to be organized has got more specific, but the writers have deliberately chosen not to specify how to meet the criteria. The intention was to guide laboratories to build a quality system that identifies all aspects in the total testing process from test request to the laboratory, through preanalysis and analysis to the report and advice to the requester. All the steps in the process have to be regulated to certify constant quality as required for adequate patient care. The standard is mandatory on which procedures have to be in place and what aspects have to be covered by these procedures, but deliberately not on the specific way how these aspects have to be covered. In the 2012 version, the addition of prospective risk analysis has resulted in an extra tool for laboratories that gives direction to the question how to meet the criteria. Secondary to the first goal was the intention to assess the success of implementation of the standard in the laboratories leading to accreditation. An inspirational editorial on the philosophy of ISO15189 and its development by Plebani et al. [20] also refers to other literature that illustrates that both the standard and the accreditation process need to be subject to continuous improvement.

Harmonization is challenged constantly

During the development of the ISO15189 standard by TC212, there were many occasions when stakeholders requested more instruction and prescription on how to meet the requirements in the standard. Although some may see this as helpful, most scientific societies in laboratory medicine recognize too much prescription as a threat to the intentions of the standard. Laboratory professionals have the obligation to design their procedure in a way that fits in with the local circumstances and intended use. The standard only describes which procedures need to be in place and what elements those procedures should cover, but the design of that procedure should be subject to risk management and constant evaluation for effectiveness and user experience.

Within TC212, others then argued that the way to meet the standard is different for the different disciplines in laboratory medicine. That argument was then used by others to recommend different versions of ISO15189 for every field in laboratory medicine. After significant debate, it was decided that within a certain discipline, local circumstances might require different solutions and that the standard should not be too prescriptive about how to meet its requirements. The third edition, issued in 2012, introduced prospective risk analysis as the tool to use when determining how laboratory policies could be optimized to meet local demands and circumstances. The 15189:2012 requires laboratories not only to perform risk analysis on all their procedures in the design phase but also to actively investigate whether current procedures are effective as a standard part of their system for internal audits and management review. This will lead to an active cycle of continuous improvement. In the end, quality systems in different laboratories will consist of the procedures for the same issues in every laboratory and of procedures that consist of the same elements in every laboratory, but that may differ in the exact content of those elements of the procedures.

Although this prevented the standard from being too prescriptive and also prevented the standard from being separately specified for every (sub)field of laboratory medicine, this did not stop the call for specification of the standard for specific fields of laboratory medicine or even for specific techniques that now are called “new” but which may be considered routine testing even before an updated standard is published. This constant call for more granularity and specification for certain techniques mostly comes from accreditation bodies and not from the specialist in laboratory medicine that participate in TC212. The accreditation bodies have an interest in the standard that is primarily focused on assessment. They are constantly looking for maximal clarity on how to assess and think that they are served best with checklist audits. The laboratory professionals on TC212 are constantly reminding other stakeholders that the primary goal of the standard is to present fundamental tenets for a risk-based quality management system applicable to all fields of laboratory medicine and that accreditation of such a system is secondary to that. Of course there must be place to harmonize or even standardize performance of certain techniques or practices, but this can be done in guidelines of (inter)national scientific societies or when necessary in ISO standards that have a normative reference to ISO15189. Technical Committee 212 is currently developing such standards, for example, for sample collection and isolation of nucleic acids. For emerging techniques, such as next-generation sequencing [21], international initiatives are formulating guidelines. Once established and officially recognized by scientific societies, such guidelines can have a mandatory role when assessing such activities in an accreditation process, thereby eliminating the need to describe such technical aspects in the ISO15189 standard or an annex to it.

Risk management is the intelligent alternative to prescription

By describing the harmonization of technical aspects in guidelines and separate standards, the ISO15189 standard is able to keep its position as the connecting standard for all fields in laboratory medicine. This is a strength and not a weakness as it urges laboratories focus on the general principles of quality management and technical competence. In this sense, it is comparable to ISO17025 which is applied in many different fields of testing and calibration laboratories.

The consequent lack of granularity gives room for professional autonomy. This autonomy, however, comes with the responsibility to document the choices a laboratory makes to fulfill the requirements in a specific situation or for specific technique. It is not only a right but also an obligation to use the “freedom” the standard gives on the exact content of procedures in a way that can be explained to and appreciated by the assessors.

As a consequence, assessment of conformity requires assessors who are knowledgeable in both the standard and the field of laboratory medicine that they assess. “Intended use” and risk management are key in ISO15189, and adherence to these aspects can only be judged correctly by professional peers who are educated in both ISO15189 and the standards and guidelines that are applicable to their part of laboratory medicine.

The lack of granularity not only necessitates professional assessors but also protects against a checklist approach in the assessment process. The standard requires a rationale from the laboratory professionals about the choices in their laboratory that cannot be assessed by a checklist but need an assessment approach with interest on local considerations and choices. The strength of ISO15189 is illustrated by the fact that it is even used outside the field of medical laboratories, e.g. biometric function tests (the Netherlands, Scandinavia), image/echo X-ray and MRI (Scandinavia, Canada). This demonstrates the usability of the basic structure as a generic standard that can be used in any (sub)field of laboratory medicine, or even in related fields of diagnostics outside the laboratory.

Harmonization of the quality systems based on ISO15189 between countries

Because the way that the quality management system is configured depends on the field of laboratory medicine, specific guidelines are needed. The international scientific societies in a particular scientific field have a role in the harmonization of the “how” within their field. CLSI is editing many specific guidelines, but also the International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) and EFLM have to play a prominent role when it comes to the field of clinical chemistry. The national scientific societies have a responsibility to contribute to the development of international guidelines and translate these into national guidelines that form the basis of local choices in the quality management system built on ISO15189. These guidelines have to be developed in close collaboration with the accreditation bodies to make sure that they do not conflict with the standard, and they are officially recognized by (the members of) the scientific societies in order to have a mandatory status when used by the assessors. In some cases, laboratories may publish their way to fulfill the requirement as a useful example to others [22, 23]. At some stage, a scientific society may consider such an approach “best practice”. It is the responsibility of the scientific societies to make clear at what stage a best practice is granted “guideline status” and thereby becomes obligatory in the eyes of the assessors. An active dialog between scientific societies and the NABs is essential for this process in order to prevent unnecessary discussions between assessors and laboratories.

Harmonization of assessment within countries

Acceptance of ISO15189 as the standard for accreditation of medical laboratories is realized in most of countries worldwide and certainly in Europe. The way that assessment is performed by the NAB has to comply to ISO17011:2004 conformity assessment: “General requirements for accreditation bodies accrediting conformity of assessment bodies” [24]. Apart from this standard, different ILAC guidelines exist, as can be seen on the ILAC website (http://ilac.org). This includes choice of auditors, their training and assessment. The audit process needs assessors who are both professionals in the standard as in their own field of laboratory medicine; hence, their assessment skills are dual. Different parties may be needed and responsible to train the assessors in these two aspects. Specific attention to training of the “team leader” is needed. The NABs are responsible for training of the assessors to be professional when it comes to the standard. During the assessment, the team leader has to judge whether nonconformities are just incidents or examples of a structural non-conformity. The development of tools that can assist the assessors in weighing-related individual non-conformities could contribute to improve the objectivity of the assessments and the appreciation of such objectivity. The national scientific societies are responsible for training the assessors in the scientific guidelines and how to assess whether a laboratory is compliant. The choice of assessors is the responsibility of the NAB, but it is important that they contact the societies about their training. Cooperation between the NAB and the national scientific societies is essential.

Harmonization of assessment between countries

For mutual lateral agreement between countries, the NABs are assessed themselves by persons from other accreditation bodies with knowledge about ISO15189. This warrants some standardization. Harmonization of training on the assessment of the standard at the world level is the responsibility of ILAC and in each continent by the national ILAC daughter, which in Europe is the EA. Harmonization of the training on how to assess the scientific aspect that determines how some requirements have to be applied is a responsibility of the international federations of the scientific societies like the IFCC in clinical chemistry. Cooperation between international medical laboratory organizations, such as EFLM and IFCC, and international accreditation organizations, such as EA and ILAC bodies, is essential. In Europe, this task is accomplished through the HCC of the EA. There, discussion on flexible scope and POCT are resolved in the form of EA publications that are mandatory for NABs in their assessment procedure. Such documents decrease the risk for large differences between assessments in different countries.

Also, the EA HCC meeting provides a mechanism for harmonization between countries. All members can send in questions on how to interpret the intentions of the standard, and these are addressed accordingly. Answers to such questions are published as part of the minutes of these meetings and also published in a “frequently asked questions” format.

Moreover, constant attention has to be paid to the “calibration of the assessments”. All NABs expend considerable effort to this by way of training, discussion of points of controversy and attention to solving complaints. Present discussion points are the inclusion of all phlebotomy locations or only a representative sample during an accreditation cycle. Quite a difference still exists in the number of assessors and the duration of an assessment between the countries in Europe. On a global scale, differences are probably even larger. Considering that international mutual recognition of accreditation results in the equivalent accreditation status in all countries, differences in the (severity of) the assessment process seem undesired.

Future opportunities

The major opportunity lies in the impact of the ISO15189 standard on other regulations.

Both in national and in international legislation, ISO15189 has a growing impact. In some countries, ISO15189 accreditation is mandatory in order to be able to offer a medical laboratory service. This warrants an equal playing field for the laboratories and prevents them from only paying attention to revenue and not to the required quality.

Other parties like governments and insurance companies are starting to recognize ISO15189 accreditation as a requirement for reimbursement. Because this became a policy in Romania, for instance, it has led to accreditation of the majority of their laboratories.

Also, the new in vitro diagnostics (IVD) Regulation that was accepted by the European Parliament in 2016 recognizes the ISO15189 standard [25]. It limits the use of in-house developed tests without CE marking to ISO15189 accredited laboratories in order to guarantee a proper validation procedure for such tests. This impact of ISO15189 on other standards and legislation is only possible when it is preserved as the common standard for all fields in laboratory medicine without decay into subfields and ultimately into increasingly specific sub-sub fields.

Future challenges and threats

The major threat towards the harmonization of ISO15189, both in its basis for quality management systems and in its use as the standard in accreditation, is the ongoing call for greater granularity. Scientific communities or specific laboratory disciplines may want their “own” version of the standard that is more aligned with their particular needs. Introduction of new techniques to the field may come with the call to accommodate assessment of particular aspects of these techniques into the standard and accreditation against it. National accreditation bodies that have difficulty in selecting the proper assessors for a particular laboratory may want to differentiate between different subgroups of laboratory fields by introducing different standard subtypes. The call for field-specific versions of ISO15189 ultimately is a threat to the position that ISO15189 has acquired in other medical fields, legislation and national and international sociopolitical systems of health care reimbursement. The specific needs of a discipline should be fulfilled by guidelines, which should become as international as possible.

In particular, if accreditation becomes mandatory, it is important that the assessment process is in line with the intentions of those who have written the standard. This requires the right attitude and training of the assessors. A too literal assessment to the letter of the text may lose sight of the intention. Therefore, it is essential that laboratory professionals and accreditation bodies are in constant dialog on a national and international level to preserve the intention of the standard in the assessment and accreditation process. On the international level, this can be achieved in both the technical committees of the “Comité Européen de Normalization” (CEN), (European Committee for Standardization) and ISO and in the EA HCC.

Opportunities for future versions of ISO15189 and its use in accreditation

In the introduction of the next version of the standard, it may be wise to recount the achievements of ISO15189 and the relationship between that achievement and the choice not to differentiate into several standards for every (sub)type of laboratory medicine or even technique. This will help prevent having the same redundant discussion repeatedly.

The POCT standard ISO22870 of 2006 was amended to fit in with the third edition of ISO15189, which resulted in the 2016 edition. In future ISO15189 versions, it would be better to incorporate the specific needs of POCT, but without technical detail, in line with our recommendation to keep the ISO15189 standard general and not too specific or too prescriptive.

To avoid the need for a guidance document, the text of the third edition was written with special attention for clear language. However, insufficient attention was paid to the difference between mandatory items (marked as shall) and items to be considered or recommended to be implemented (marked as should). Because an accreditation body has to look into all “shall” aspects, the present standard is more prescriptive than intended by the authors. In the light of the revision process, TC212 will circulate a questionnaire to the participating standardization organizations that will ask for feedback on the present edition of ISO15189. It is important that the scientific societies participate in this questionnaire through the national standardization organizations. They should address those points in the standard that are in need of better description. When they do not, only the voice of the NABs will be heard.

When national standard bodies recognize that different fields of laboratory medicine require different solutions to improve the standard, they may recommend scientific guidelines for the particular.

Harmonization of the accreditation between different fields in laboratory medicine may also be helped by the harmonization of the scope of accreditation. ILAC and EA regulations state that for every accreditation, a scope has to describe for which activities the laboratory is granted accreditation. A scope can be either fixed or flexible. A fixed scope consists of a listing of all individual services for which accreditation is granted, whereas a flexible scope consists of coherent groups of obviously related procedure within the same medical field, technical field and material. Both EA and EFLM promote the use of the flexible scope [26], but harmonization of the specificity of flexible scopes is needed to promote transparency and comparability of the accreditation status of different laboratories [27].

Conclusions

ISO15189 has found its way to become the harmonized standard for quality management systems in laboratory medicine and for its accreditation. This role is endangered by a constant call from both scientific societies and accreditation bodies for more specificity, for instance, by specific versions of the standard for particular fields or techniques. To preserve the harmonized and uniting role of the standard, scientific societies have to invest in guidelines and technical standards that are used in parallel with the standard. Scientific societies will have to participate in cooperation with the NABs and the national standardization organization structures to create future versions with greater impact and user satisfaction than the current version.