Harmonisation of Laboratory Testing

The Clinical and Laboratory Standards Institute refers to a definition of ‘(method) harmonization’ as ‘the process of recognizing, understanding, and explaining differences while taking steps to achieve uniformity of results, or at a minimum, a means of conversion of results such that different groups can use the data obtained from assays interchangeably’.¹

In the following series of articles in Part A of this two-part series on the Harmonisation of Laboratory Testing, the scope of harmonisation addressed is more far-reaching than method harmonisation, and includes other aspects of pathology testing namely:

1. Standardisation of pathology units and terminology.

2. Harmonisation of report formats where there are patient safety issues, e.g. left to right versus right to left reporting.

3. Harmonised reference intervals (RIs) and decision limits.

4. Best practice evidence for test requesting.

The Electronic Health Record (EHR) and patient safety issues are driving the need for harmonised methodology, terminology and units of reporting in pathology. Analytical tests use different methods that may not have been ‘harmonised’ or which may have different units of reporting. Those who request pathology tests and receive pathology reports, the information systems developers and even laboratories appear to be unaware of these reporting differences, especially if the transfer of results from the laboratory to the report recipient does not show the units of reporting or the assay methods in use. Inevitably the assumption made by clinicians is that the different numbers can be compared. This has the potential for misinterpretation of results and adverse patient outcomes.²

However, pathology harmonisation is more than harmonised terminology, units of reporting, methodology and decision limits. It covers a wide range of topics from the ‘pre-pre-analytical’ phase (‘Right test at the Right time for the Right patient’) to the analytical aspects and reporting of critical tests, through to consumer education and the meaning of laboratory tests in lay terms (‘post-post-analytical’ phase). It is important to identify where harmonisation is potentially beneficial across all pathology disciplines and to involve the relevant stakeholders (Table). Both the pathology laboratory community and clinicians using the tests need to become involved in the harmonisation process.

Table.

Requirements to achieve pathology harmonisation.

Phase	Harmonisation Requirements	Vested Stakeholders
Pre-analytical	Standardise: - Laboratory and Pre-laboratory/External pre-analytical processes. Implement SOPs to reduce error and ensure patient safety. Recommend use of evidence-based guidelines for appropriate test selection. Plan for implementation and educational phases.	Patients (inpatient and outpatient); healthcare practitioners e.g. phlebotomist; laboratory personnel. ACSQHC; IFCC WG-LEPS. Clinicians; the pathology community; guideline organisations. Professional societies; the pathology community.
Analytical	Standardise pathology test names and units. Standardise test requesting and reporting for the EHR. Harmonise report formats where there are patient safety issues. Harmonise patient results through a standardisation and/or method harmonisation process. Plan for implementation and educational phases.	1–3. NEHTA Clinical Terminology and Information; Australian Government Department of Health & Ageing. 4. National metrology institute; diagnostic assay manufacturers; laboratory personnel; clinicians. 5. PAC, RCPA, other Professional Societies and Industry Associations.
Post-analytical	Harmonise reference intervals and/or use clinical decision limits where the between-method bias is within the allowable limits of performance and provides consistent patient result interpretation. Plan for implementation and educational phases. Requires long-term monitoring of between-method bias using commutable patient samples. Report critical patient values according to a national critical test list.	1–2. Professional Societies; the pathology community. 3. EQA schemes, e.g. RCPA QAP; diagnostic assay manufacturers. 4. Laboratory personnel; patients (inpatient and outpatient); clinicians; ACSQHC.
‘Post-post-analytical’	Educate consumers about the meaning of laboratory tests. Develop an on-going laboratory-clinical-systems provider working relationship for long-term sustainability of pathology harmonisation.	Patients; clinicians; consumer advocate groups. The pathology community; clinicians; systems providers.

ACSQHC, Australian Commission on Safety and Quality in Health Care; EQA, external quality assurance; EHR, electronic health record; IFCC WG-LEPS, International Federation of Clinical Chemistry and Laboratory Medicine Working Group on Laboratory Errors and Patient Safety; NEHTA, National E-Health Transition Authority; PAC, Pathology Associations Council⁷; RCPA QAP, Royal College of Pathologists of Australasia Quality Assurance Programs Pty. Ltd.; SOPs, standard operating procedures.

In the following papers several of the key issues that have important implications for pathology harmonisation are reviewed and discussed.

Mario Plebani, who leads the International Federation of Clinical Chemistry and Laboratory Medicine Working Group on Laboratory Errors and Patient Safety (IFCC WG-LEPS), describes the quality indicators (QIs) to detect pre-analytical errors in laboratory testing. The 16 QIs developed for the pre-analytical phase include those from outside the control of the laboratory, e.g. test requesting, patient and sample identification and sample collection (so-called ‘pre-pre-analytical phase’ QIs), to QIs that cover all steps of the laboratory ‘pre-analytical phase’, e.g. sample preparation and storage. Through an understanding of the risks to patients undergoing laboratory testing, standardised operating procedures can be developed to reduce laboratory error and improve patient safety.

Already there are global pathology harmonisation activities underway as described by Jonathan Berg in his article. In the UK in 2007, the Pathology Harmony initiative was begun. The project aims to optimise pathology services through the harmonisation of RIs, test names and reporting units, and to develop guidelines relating to pathology testing, e.g. tumour marker requesting. Berg describes how the Pathology Harmony working groups have developed harmonisation techniques to determine the suitability of an analyte to be harmonised.³

They include:

• Quantification of the degree of variation for an analyte (by laboratory survey)

• Examination of the scientific basis to determine if the evidence explains this variation or not (in which case harmonisation would proceed)

• Development of consensus recommendations

• Communication of the recommendations to the key stakeholders including the pathology community, relevant clinical groups, information technology (IT) providers, the UK Department of Health

• Implementation of the recommendations with distribution of appropriate support documentation to consumers.

While the harmonisation processes described above seem to be quite straightforward, no doubt a huge amount of work including countless meetings with stakeholders and considerable data extraction is required to underpin the consensus recommendations.

In the area of method harmonisation there is considerable debate about how best to achieve comparability of measurement for heterogeneous proteins and whether it should be via the harmonisation or the standardisation process. Ideally the harmonisation of patient results is based on standardisation and calibration traceability to the SI unit via a primary reference measurement procedure (RMP) and primary reference material (RM), as is the case for several chemistry analytes, e.g. cholesterol, glucose, creatinine. The method harmonisation process as compared with the standardisation process may be biased and possible only in a method-dependent manner with no long-term anchor of trueness to a reference measurement system available.⁴ Nevertheless, if routine measurement procedures have the same analytical specificity for a measurand, and the secondary reference materials or manufacturer’s calibrators to be used for value transfer are commutable between measurement procedures, then harmonisation is feasible.

Mary Lou Gantzer and Greg Miller in their article describe the importance of method harmonisation as a pragmatic procedure for achieving the closer comparability of patient results, in particular for high priority measurands where a decision limit or common RI is used and measurement inaccuracy can cause clinical misclassification. The authors refer to ‘(method) harmonization’ ‘to mean any process for achieving comparable results among measurement procedures for an individual measurand when no primary RMP exists’. Calibration traceability may be to a secondary RM or to a manufacturer’s internal working calibrator, and requires that these materials are both commutable in routine measurement procedures and adequately specific for the defined measurand.⁵

James Davidson contributes to the method harmonisation discussion through the example of growth hormone (GH) measurement in Australasia. Although assays have calibration traceability to the international conventional calibrator for GH, WHO 98/574, there remain large differences in patient results between assays. A numerical adjustment would eliminate large systematic assay differences and produce harmonisation of results. Clinically this is highly desirable and, depending on the exact process used for realignment, the clinical cut-off values for GH deficiency and acromegaly may also need to be re-evaluated. This approach to GH harmonisation requires an ongoing monitoring of assays to detect any systematic drift in assays by the evaluation of patient samples that are preevaluated for commutability.

In the next section on post-analytical reporting of patient results, Michael Legg and Christiaan Swanepoel describe the urgent need for standardisation of pathology units of reporting and terminology. Current reporting practices are heterogeneous and may lead to an increased risk of interpretation errors, possibly endangering clinical safety. It is these factors which are driving the requirement for standardisation of IT structures and terminology. Legg and Swanepoel go on to explain the difficulties of introducing electronic reporting of pathology results and the basic requirements for a standardised system interoperability approach if reporting is to be harmonised and the health informatics data are to be reliable. Currently there is an important contribution being led by the profession called the Pathology Units and Terminology Standardisation (PUTS) Quality Use of Pathology Project. The PUTS Working Groups are assessing units, terminology and which tests to combine for display as a graph or cumulative report (i.e. code harmonisation), and will provide the structure for future pathology test requesting and reporting.

Robert Flatman in his review of coding for pathology tests describes the strengths and weaknesses of computer automation in e-health and the key issues for sharing of EHRs. He describes the need for functional interoperability, semantic interoperability and standardisation of report elements between laboratories to ensure results can be distinguished by different terminology codes. In Australasia, New Zealand leads the way thanks to its National Health Index patient ID and use of EHRs. This practical assessment of the pros and cons of e-health is an important reminder that laboratories need to work more closely with their IT.

Another key to harmonised pathology reporting is the use of common RIs or decision limits. There is an urgent need to progress the implementation of common RIs and decision limits.⁶ Without this framework, the task of integrating results from different laboratories into any future national e-health framework will be made significantly harder. In the New Zealand approach to harmonised RIs, Maxine Reed describes how the three major New Zealand Quality Assurance Groups (QAGs) develop common RIs and standardise terminology, units and Logical Observation Identifiers Names and Codes (LOINC) coding when reporting laboratory test results. This practical model developed by the Auckland Regional QAG more than 35 years ago provides a mechanism by which RIs can be harmonised across laboratories in a region. Reed notes however, ‘that while harmonisation (of reporting of laboratory test results) is a noble principle, in practice it is difficult to achieve, and that the achievement of ‘National or Australasian Harmonisation’ may still be some way off’. Importantly laboratories need to work closely with IT staff, General Practitioner patient management system vendors and regulatory bodies that access patient databases to ensure optimal use and reporting of laboratory results and to avoid misinterpretation of cumulative results.

Conclusion

Those of us working in pathology laboratories are well aware that the harmonisation of laboratory testing is long overdue. The harmonisation quality initiative should raise the profile of pathology but does require a multidisciplinary approach to make progress. Close interaction with all stakeholders including the pathology laboratory community, diagnostic industry, clinicians, professional societies, IT providers, consumer advocate groups and the government is essential for harmonisation projects to be successful and to achieve improved clinical effectiveness of critical tests and greater patient safety. This will require a planned communication and marketing strategy in order to be able to roll out the relevant changes, educate clinicians, and to gain acceptance of these processes by all stakeholders.

The AACB extends an invitation to all pathology laboratories in Australia and New Zealand, and beyond, to participate in the ‘Harmonisation of Laboratory Testing’.