External Quality Assessment: an effective tool for

Article in press - uncorrected proof Clin Chem Lab Med 2006;44(6):740–749 2006 by Walter de Gruyter • Berlin • New York. DOI 10.1515/CCLM.2006.133

2006/110

Review

External Quality Assessment: an effective tool for Clinical Governance in Laboratory Medicine

Laura Sciacovelli1,*, Sandra Secchiero1, Lorena Zardo1, Martina Zaninotto2 and Mario Plebani1,2

ratories; quality specifications; unacceptable performance.

1

Centre of Biomedical Research, Civil Hospital, Castelfranco Veneto (TV), Italy 2 Department of Laboratory Medicine, UniversityHospital, Padova, Italy

Abstract The implementation of Clinical Governance will require a redefinition of duties and accountability as a prerequisite to develop and achieve an overall improvement in clinical care through a culture of assessment and monitoring of quality. External Quality Assessment Schemes (EQAS) are the main tool enabling laboratories to measure the quality of their results; they must carefully assess and monitor all elements contributing to the formulation of laboratory information (results, reference ranges/decisional levels, interpretative comments and diagnostic algorithms). There are different ways to design and manage a Scheme and EQAS coordinators are mainly responsible for its effectiveness. The present paper reports, as an example, some experiences of the Centre of Biomedical Research (CRB), which manages EQAS according to high quality specifications and laboratories’ needs, that can reflect the Clinical Governance philosophy. Our findings show that EQAS are able to control all the above aspects and, if organisers are committed to fulfilling the responsibility and accountability principles, they will be of great value in quality assessment and in developing an External Quality Assurance Program (EQAP). This is an interlaboratory comparison designed and conducted to assure the following: evaluation of participants’ performance (by evaluating not only analytical performance, but also test interpretation, and advice for clinicians on laboratory requests and diagnosis); evaluation of method performance; and continuous education, training and help. The main aim of the activities of an EQAP in Laboratory Medicine is to sustain improvements in the quality of services provided by participating laboratories for the benefit of patients. Keywords: Clinical Governance; External Quality Assessment; interpretative comments; medical labo*Corresponding author: Laura Sciacovelli, Centro di Ricerca Biomedica, Via Ospedale 18, 31033 Castelfranco, Veneto (TV), Italy E-mail: [email protected]

Introduction Several papers have described the concept, scope, and advantages of the approach of Clinical Governance in the National Health Service (NHS) and in Laboratory Medicine. Clinical Governance strongly emphasises the principles of accountability for quality in healthcare, strengthening Certification/Accreditation models for which clinical laboratories have been the first and main users. Each organisation that is part of the healthcare service or is involved in the system must ensure that: quality improvement processes are in place and that they are integrated with the quality programme of the organisation as a whole; good practice ideas and innovations are systematically disseminated and applied; poor clinical performance is promptly recognised and dealt with to protect patients; and the quality of data collected to monitor clinical care is in itself of a high standard. In this context, a fundamental redefinition of the duties and accountability is required to comply with the aims and principles of Clinical Governance (1). External Quality Assessment Schemes (EQAS) are an integral part of the overall quality assurance systems of the clinical laboratory. One of the main functions of EQAS is to assess whether laboratories perform tests competently. How are EQAS involved in Clinical Governance? And should roles, responsibilities and tasks be changed? Continuous progress made in Laboratory Medicine has led to constant development of and changes in External Quality Assessment design. Currently, EQAS aim to improve laboratory performances through education, scientific recommendations and standardisation, while taking into account clinical needs and quality specifications. There are different ways to design and manage a Scheme and EQAS coordinators are mainly responsible for its effectiveness. The Accreditation Program for EQAS guarantees compliance with high quality specifications, in the organisational and technical areas, and in fulfilling users’ needs. However, not even the best EQAS information can provide advantages if laboratories do not use them to check and eventually modify internal processes and to implement improvement actions. EQAS organisers play a very important

Article in press - uncorrected proof Sciacovelli et al.: External Quality Assessment for Clinical Governance in Laboratory Medicine 741

role: they encourage, lead and advise laboratories to undertake appropriate actions and to manage wider problems that can have different causes and involve several actors. The ways and commitment with which EQAS coordinators manage problems arising within the EQAS, implement new schemes and decide on information to send laboratories are a testament to their competency and their participation in the Clinical Governance approach. The present paper reports, as an example, some experiences of the Centre of Biomedical Research (CRB), which manages EQAS according to high quality specifications and laboratories’ needs, that can reflect the Clinical Governance philosophy.

Users and goals In order to operate effectively, each organisation must first identify the users and their needs, and can then define missions and goals. The main task of the clinical laboratory is to provide an effective answer to clinical requests and to help clinicians in making decisions. The laboratory must be able to offer clinicians the most appropriate test possible in response to their request (diagnosis, monitoring, prognosis). For example, when a patient is admitted to hospital with atypical chest pain and acute myocardial infarction is suspected, troponin is the best test available to establish whether or not the patient has had a myocardial infarction. To evaluate the efficacy of treatment of diabetic patients, glycohaemoglobin (HbA1c) is the best test available to establish whether or not the patient has complied with the treatment protocol. The laboratory has to guarantee a satisfactory answer to questions posed; it must therefore: • Produce a correct result; • Define an appropriate reference range/decisional level; • Provide, whenever appropriate, suitable interpretative comments and diagnostic algorithms. In fact, the interpretation of these elements provides the information that assists the clinician in

making decisions regarding the care of an individual patient. The laboratory has a duty to verify the quality of the service in order to guarantee increasingly high quality standards, thus providing increasingly effective information; it must also be able to constantly measure the quality of its service, an essential component in the successful realisation of a culture supporting Clinical Governance (2, 3). Since EQA Programs are the main tool enabling laboratories to measure the quality of their results, they must carefully assess and monitor all elements contributing to the formulation of laboratory information. In particular: a) Results: • Assess and monitor the analytical performance of the laboratory; • Promote the improvement of overall performances; • Support the management of unsatisfactory performances. b) Reference ranges/decisional levels • Analyse the state-of-the-art; • Divulge scientific guidelines. c) Interpretative comments and diagnostic algorithms • Verify consensus among laboratories.

Laboratory results and analytical performance In order to encourage laboratories to handle EQA control samples in the same way as patient samples, the CRB requires EQA results to be returned on a medical report form and then provides the laboratories involved with a report containing an assessment of their results. The control sample is processed as a patient sample, from its receipt to the formulation of the report. The analytical performance of the laboratory is assessed against a consensus value, taking into account the total analytical error (TEa) assigned to each constituent according to specific criteria. Clear judgement is provided on analytical performance in

Figure 1 Assessment of analytical performance: example of EQAS reports.

Article in press - uncorrected proof 742

Sciacovelli et al.: External Quality Assessment for Clinical Governance in Laboratory Medicine

Figure 2 Performances obtained in the last cycles of EQAS for CK-MB mass: from 1999 to 2002, an increase in satisfactory (from 86.8 to 95.7%) and a decrease in unsatisfactory (from 13.2% to 4.3%) performance is observed. Acceptability limits have been reduced from 2003 to stimulate laboratories to achieve continuous improvement.

terms of excellent, good, acceptable, or unacceptable (4, 5). Figure 1 shows the part of report for the Biochemical Markers of Myocardial Damage Scheme and the Glycohaemoglobin Scheme containing this type of information. Participation in an EQAS does not guarantee laboratories the quality of their own results. The laboratory staff have a duty to analyse the EQA report, discuss the results and evaluate possible actions that might be undertaken to improve laboratory performances. The information supplied by EQAS is effective only if the laboratory takes appropriate actions to ensure improvement. EQAS organisers are responsible for continuous evaluation and monitoring of the trends for total performances to check the distribution for different classes, improvements and disimprovements, and to implement appropriate investigations. Figures 2 and 3 give an example of this, showing the percentage of performances obtained in recent EQAS cycles for CK-MB mass and glycohaemoglobin: there is an increase in satisfactory performances and a decrease in unsatisfactory performances. When an improvement in performances is observed and technological opportunities are available, EQAS organisers can evaluate whether it is possible to reduce the acceptability limits for analytical performances to stimulate laboratories to achieve continuous improvement, if clinical needs require it; otherwise, it is deemed unnecessary to make useless efforts to improve the performances.

Organisers are required to control the percentage of unacceptable performances in every EQA survey conducted. A high percentage of unacceptable performances may have different causes, which must be carefully identified. Analysis of the problem involves an investigative process to determine the root causes and to select the action most likely to eliminate the problem and prevent its recurrence. In the 2004 cycle of the EQA Scheme for Biochemical Markers of Myocardial Damage, a high number of unacceptable performances was observed for troponin I (cTnI) and troponin T (cTnT). This event involved all diagnostic systems used by participants. To investigate the causes of this unexpectedly high number of unacceptable performances, the Scheme’s coordinator promoted more effective co-operation between laboratories and the manufacturers of diagnostic systems. An investigation was conducted to identify the procedures used by laboratories for: • Handling of control materials (verification at reception, storage, preparation) before analysis (preanalytical phase); • Calibration procedures and Internal Quality Control (IQC) management (intra-analytical phase). In particular, with regard to pre-analytical phase, the laboratories were sent: a) A brief operating procedure describing the correct steps for handling control materials before analy-


Figure 3 Percentage of performances obtained in last EQAS cycles for glycohaemoglobin showing an increase in satisfactory (from 73.7% to 82.3%) and a decrease in unsatisfactory (from 26.3% to 17.7%) performances.

sis (underlining in particular the need to store control material at a temperature of y208C until analysis) (What must be done!). b) A questionnaire concerning the handling (receipt, storage, treatment) of EQA control materials before the analytical phase in relation to operating procedures, to verify the steps actually followed (What was done?). c) A questionnaire on the test calibration and IQC management.

In collaboration with diagnostic system specialists, the coordinator of the Scheme prepared and distributed a questionnaire on the intra-analytical phase; the aim was to verify the procedure followed for instrument calibration and the management of IQC. On evaluating the laboratories’ answers, it was found that, in some cases, procedures for the validation of calibration were incorrectly applied and management of IQC was generally poor. In particular, the calibration procedure used was often inadequate for:

Moreover, specialists for each diagnostic system visited each laboratory with poor performances to verify the maintenance and calibration of instruments. A total of 84% of laboratories completed and returned the questionnaire and the findings of the investigation were unexpected. Regarding the preanalytical phase, the features of the control materials of the Scheme called for particular attention, since the laboratories were sent liquid samples in a frozen state. Although the participating laboratories had been alerted to the problem and a clear operating procedure had been sent to them, evaluation of the questionnaire answers revealed that the participants’ approach to handling (receipt, storage, treatment) of EQA control materials was often inappropriate (Table 1).

• Storage of material, before and after use (temperature and conditions); • Conditions and time used to defrost the material (when material is frozen); • Procedure for mixing the material; • Analysis temperature; • Frequency of calibration (at a defined time or on changing reagent lot, when necessary); • Criteria for validation of calibration points. The analysis of laboratories’ answers on IQC management highlighted inappropriate practice in: • Use of IQC control material that did not provide useful information about what was going on in the testing process;



Table 1 EQAS for Biochemical Markers of Myocardial Damage: results of investigation into the handling and storage of EQA control materials. Procedure items

Laboratory response

Verifying state of EQA material on receipt

State of control material on receipt • Frozen • Defrosted • Not verified Storage temperature • From y20 to y808C • From y12 to y208C • From y2 to y108C • From q2 to q88C Time required for defrosting • From 3 to 30 min • From 36 to 60 min • From 75 to 360 min • From 8 to 20 h • 11 days • Not verified • No answer Time between defrosting and analysis • From 0 to 30 min • From 30 to 60 min • From 70 to 240 min • 25 h • 13 days • No answer

Keeping EQA materials at temperature Fy208C

Before analysis, defrosting EQA materials

• The number and concentration of materials were not sufficient to determine proper operation over the range of interest and at clinically relevant levels to reflect values encountered in patient specimens; • Frequency of control measurements; • Decision criteria and graphic display of control results. Moreover, the data reported by specialists during their verification in the laboratories were very interesting. The EQA material was analysed after maintenance and calibration were carried out by laboratory staff, and then after maintenance and calibration procedures were carried out by the diagnostic system specialists. Table 2 reports, for example, the results obtained for cTnI and myoglobin. Intra-laboratory variability decreased considerably when specialists car-

89% 7% 4% 83.6% 9.7% 5.2% 1.5% 64.2% 17.9% 9.7% 3.0% 0.7% 1.5% 3.0% 63.2% 22.6% 9.8% 0.7% 0.7% 3.0%

ried out the procedures for maintenance and calibration, thus demonstrating that accurate and correct application of the procedure assures better control of the analytical variables and provides more reliable results. To improve upon the management of the above aspects, brief instructions emphasising the phases (e.g., handling reagents and calibrators, equipment maintenance, calibration validation criteria and IQC) to be controlled were compiled and distributed to all participants in agreement with different manufacturers. In subsequent surveys (5th and 6th), the total number of unacceptable performances for cTnI was reduced from 38% to 12% and the total number of excellent performances increased from 28.3% to 57.3% (Figure 4).

Table 2 Results for EQA control materials obtained after maintenance and calibration procedures carried out by laboratory staff and diagnostic system specialists. Troponin I

Myoglobin

Maintenance and calibration

Maintenance and calibration

Laboratory staff

Specialists

Laboratory staff

Specialists

Median 75th percentile 25th percentile SDrob CV%

0.17 0.19 0.14 0.04 21.80

0.16 0.17 0.15 0.02 8.70

37.60 38.6 37.48 0.83 2.20

37.30 37.49 37.10 0.29 0.78

Mean SD CV%

0.17 0.03 17.46

0.16 0.01 6.85

37.58 1.35 3.59

37.46 0.69 1.84

SDrob (robust estimation of the standard deviation) s (75th percentiley25th percetile)/1.349.


Figure 4 Improvement in laboratory performance obtained in different control samples as a result of actions (procedures for pre- and intra-analytical phases and support from specialists of manufacturers’ diagnostic systems) undertaken in agreement with manufacturers, EQAS organisers and participants.

On the other hand, the number of unacceptable performances for cTnT decreased only when new control materials were used. The CRB, in fact, periodically evaluates all commercially available control materials (when fresh human serum is not available) and utilises those that best satisfy commutability, homogenous and stability criteria. For the 2005 cycle, new material was used in EQAS for Biochemical Markers of Myocardial Damage after reliability tests had been conducted according to scientific recommendations (6, 7). Using all diagnostic systems, we analysed all commercially available control materials randomly interspersed between 20 fresh patient samples (chosen analyte concentrations were approximately evenly distributed over the concentration range of the processed sample of interest). The absence of a reference method with presumed minimal error led us to use bivariate Passing-Bablok regression analysis for all analytes and diagnostic systems, choosing the 95% confidence intervals. For EQAS, we chose control material with inter-assay analytical behaviour that best simulated the behaviour of authentic patients samples. Figure 5 reports the results of this evaluation for troponin. cTnT performances improved markedly with the new control materials, which are lyophilised differently from previous material (unacceptable performances: from 52.4% to 17.7%; excellent performances: from 23.6% to 45.0%).

Reference ranges/decisional levels and interpretative comments Well-designed EQAS can support laboratories in establishing appropriate reference ranges and verifying the clinical significance of their results in comparison to those obtained in other laboratories.

Reference ranges/decisional levels Reference ranges and decisional levels are important benchmarks for the clinical interpretation of laboratory test results. Laboratory scientists are responsible for determining these values, and setting such ranges and levels is one of the most critical issues in this context. This complex task calls for a high degree of clinical knowledge that cannot be reduced to simple statistical processing of data in the laboratory. Reference ranges used by different laboratories are seldom identical, and this often confuses clinicians and patients. The concept of transferability of reference ranges is discussed: the main question for the laboratory specialist is whether the laboratory can use the limits proposed either by the manufacturers or reported in the scientific literature, or whether the information from scientific recommendations can be adapted for the test population and for the analytical system in an individual laboratory.



Figure 5 Troponin: evaluation of matrix effects according to the NCCLS EP14A protocol in patient samples and control materials with two analytical systems. The same graph is reported on different concentration scales (0–20, 0–5 and 0–2 mg/L). Abscissa: results obtained using Method XXX; ordinate: results obtained using Method YYY.

When there is standardisation among analytical systems and quality specifications are complied with suggestions proposed by scientific organisations, the results reported by different laboratories must be comparable and reference ranges have to be identical to each other, or, at least similar. EQAS organisers can help laboratories to control this aspect. The CRB provides laboratories with a graph reporting the reference ranges/decisional levels used by participants; laboratories can thus compare their own reference ranges/decisional levels with those employed by other laboratories using the same or a different analytical system. This extremely important aspect cannot be ignored by EQAS, and organisers must strongly encourage laboratories to carefully verify the reliability of their values (8, 9).

In the case of HbA1c, for example, the American Diabetes Association (ADA) recommendation defines the HbA1c value for glycaemic control as -7%, referenced to a non-diabetic range of 4.0–6.0% using a DCCTbased assay (10). It is interesting to observe that, in spite of consensus among laboratory results and satisfactory quality specifications for intra-laboratory (-3%) and inter-laboratory (-5%) variability, only 75% of laboratories participating in the CRB EQAS for HbA1c complied with this suggestion (Figure 6). The role of EQAS organisers is to disseminate scientific recommendations and to encourage laboratories to improve upon consensus by continuously verifying the values used. Co-operation between laboratories and EQAS organisers, using the brainstorming technique, can help in

Figure 6 EQA Scheme for HbA1C: graph reporting the reference ranges used by participants. A total number of 75% of laboratories used a range of 4.0–6.0%.


Table 3 Interpretative comments added to the results for two patients (samples A and B). Sample A

Sample B

Comment

Number Comment

Number

Liver and kidney disorders with inflammatory neuromuscular complications Liver and kidney disorders Acute muscular disorders with renal complications Heart disease or acute myocardial infarction Myocardial necrosis or autoimmune myosites Myopathy with renal insufficiency

1

Metabolic disease related to eating disorder

1

1 1 1 1 1

1 1 1 1 1

Comments only describing results No information

1 2

Kidney disease Liver steatosis, renal failure Dyslipoproteinaemia type III Renal failure and dyslipidaemia Metabolic disease with impaired glucose tolerance, dyslipidaemia No information

identifying the causes of differences, resolving them and individuating the right criteria to use. A group has now been set up to work on a ‘‘Survey on criteria used by laboratories to define reference ranges’’ promoted by the CRB to improve this aspect. Interpretative comments Interpretative comments, an important part of the information in a laboratory report, are particularly useful in giving the correct interpretation of new and complex tests, providing guidance on appropriate test ordering, and adding value to the report itself. EQAS for the evaluation of interpretative comments may help in education and continuing professional development (8). The CRB conducted experimental surveys on clinical cases for which all participants in a specific Scheme are involved, and the results were highly satisfactory (9). The following reports the results of a survey involving a few laboratories with excellent performances. Different samples of human origin, with possible disease affecting a specific organ, were distributed to nine clinical laboratories in the Veneto Region (Italy). We asked laboratories to test 11 samples and add an interpretative comment to their results and suggest further investigations. Some key phrases were identified in each comment according to the degree of abnormality found in the samples; these phrases were allocated to key groups,

3

and the consensus between laboratories in adding a brief interpretative comment to the patients’ results was verified. A general consensus among participants was observed for results obtained, while there were differences in the reference ranges, which did not, however, appear to affect the interpretation of clinical results. The comments expressed revealed a diversity of opinion, in particular regarding two samples, with different laboratories expressing different comments (Table 3). In contrast, a general consensus was observed for the other samples. Regarding further investigations suggested by participants, in response to the finding of phlogosis, renal failure and dehydration, most laboratories (70%, 88% and 75%, respectively) did not suggest that further investigations should be conducted. Specific suggestions were made to conduct further investigations, however, when cardiac, liver and metabolic diseases were found (Table 4). The results clearly demonstrate that EQAS organisers cannot ignore the verification of comments provided by laboratory specialists in their reports. EQAS regarding interpretative comments need to be further developed and fine-tuned, and EQAS organisers need to commit to implementing surveys to assess the competency of laboratory specialists and the criteria they used, and to help increase consensus between laboratories.

Table 4 Further investigations suggested by laboratories on the basis of interpretation of results. Interpretative comment

Further investigations suggested

Laboratory answers

Heart disease

Biochemical markers of myocardial damage No information Investigation of glucose tolerance Complete blood count, lipidic profile No information Specific autoantibodies CA 19-9 Complete blood count, ESR, CRP, urinalysis, PT, serum protein; IgG, IgM, IgA, AFP, hepatitis B and C markers Hepatitis markers, CEA, AFP No information

50% 50% 34% 33% 33% 9% 9% 9%

Metabolic disease

Liver disease

9% 63%

ESR, erythrocyte sedimenation rate; CRP, C-reactive protein; PT, prothrombin time; AFP, a-foetoprotein; CEA, carcinoembryonic antigen.



Continuing professional development

Considerations and conclusions

The EQAS generate a large amount of data, which can be used by scheme organisers to assist in the education of participants and potential participants. It is fundamental that EQAS organisers accept this educational role and manage meetings and training courses in which practical information on trouble shooting and on the interpretation of EQA (and IQC) data is provided, thus enabling participants to identify poor performances due to imprecision or to bias and to recognise, for example, any need to verify calibrators, methods, analysers or reagents. An interactive training course, involving organisers and participants, is likely to be more effective in improving quality. The CRB EQAS coordinators, in assuming this role and on the basis of the data obtained in the questionnaires (pre-analytical and intra-analytical phases), organised a course to train laboratories to:

The implementation of Clinical Governance will require a redefinition of duties and accountability as a prerequisite to develop and achieve an overall improvement in clinical care through a culture of assessment and monitoring of quality. Laboratory medicine specialists are probably one step ahead of many of our clinical colleagues, because we have been monitoring the quality of our laboratory results for several years using IQC and EQAS. Currently, the mission of laboratory information is to provide an answer to the clinical question based on the whole result, involving reference range/decisional levels, interpretative comments and diagnostic algorithms, and can indicate possible action to be taken by clinicians on the patient that produces an outcome. In this context, EQAS play a primary role in the assessment and monitoring of all elements that contribute to the formulation of laboratory information; they cannot simply provide an elaboration of EQA results as a mere picture of the state-of-the-art. EQAS organisers are required not only to send out samples for analysis, but also to devise a clinical context within which the results are to be interpreted. This approach not only brings into play the ability of a laboratory to produce accurate results, but also calls for the results to be interpreted in the context of a clinical question. Our findings show that EQAS are able to control all the above aspects and, if organisers are committed to fulfilling the responsibility and accountability principles, they will be of great value in quality assessment and in developing an External Quality Assurance Program (EQAP). The main aim of the activities of an EQAP in Laboratory Medicine is to sustain improvements for the quality of the services provided by participating laboratories for the benefit of patients. This is an inter-laboratory comparison designed and conducted to assure the following: the evaluation of participants’ performance (by evaluating not only analytical performance, but also test interpretation and advice for clinicians on laboratory requests and diagnosis); the evaluation of method performance; and continuous education, training and help. It is now our duty to put the Clinical Governance strategies into place, and to carry out an EQAP appropriate for achieving the best possible outcome for patients through clinical excellence.

• • • •

Develop a specific quality control strategy; Identify the causes of problems; Manage out-of-control situations; Understand the inter-relation between IQC and EQA.

The aim of the course was to teach laboratories to correctly interpret data reported in the EQA report, investigate the causes of unacceptable results and implement changes to prevent their recurrence. Diagnostic system specialists were also involved in this teamwork, using their technical expertise to provide support for the investigative process. The course was much appreciated because, in particular, it put theory into practice. Particular item The choice of tests to be included in different Schemes is another aspect of the educational approach that is worthy of attention. EQAS must be provided with relevant laboratory investigations in the field covered (e.g., within the particular medical laboratory speciality or sub-speciality). High priority should be given to programmes for new types of investigations. The laboratories are involved, in particular, in assessing the appropriateness of test requests, and in promoting the use of novel, more effective tests, discouraging the use of obsolete ones. Analogously, the EQAS organisers are responsible for providing Schemes in which tests of high informative significance are assessed. For example, the EQA Scheme for Markers of Cardiac Damage must offer tests such as troponins, CK-MB mass, myoglobin, homocysteine, brain natriuretic peptide (BNP), NT-proBNP, and not tests such as total CK, CK-MB catalytic, isoforms CK, alanine aminotransferase (ALT), aspartate aminotransferase (AST), lactate dehydrogenase (LDH), which are considered obsolete for cardiac damage. The educational approach of an EQA Program should take into account this aspect.

Acknowledgements

The authors would like to thank the specialists of in vitro diagnostic companies (Abbott, Dade Behring, Beckman, Roche Diagnostics) for their technical assistance in the analysis of data from an intra-analytical questionnaire for cardiac marker testing.


References 6. 1. Donaldson LJ. Clinical Governance: a statutory duty for quality improvement. J Epidemiol Community Health 1998;52:73–4. 2. Kettelhut MM, Chiodini PL, Edwards H, Moody A. External quality assessment schemes raise standards: evidence from UKNEQAS parasitology subschemes. J Clin Pathol 2003;56:927–32. 3. Halligan A, Donaldson L. Implementing clinical governance: turning vision into reality. Br Med J 2001; 322:1413–7. 4. Sciacovelli L, Secchiero S, Zardo L, Plebani M. External Quality Assessment Schemes: need for recognised requirements. Clin Chim Acta 2001;309:183–99. 5. Zaninotto M, Sciacovelli L, Pagani F, Panteghini M, Plebani M. External Quality Assessment for biochemical

7.

8.

9.

10.

markers of myocardial damage: an Italian experience. Clin Chem Lab Med 2004;42:1434–41. NCCLS. Evaluation of matrix effects – Approved Guideline EP14-A. Wayne, PA: NCCLS, 2004. Passing H, Bablok W. A new biometrical procedure for testing the equality of measurements from two different analytical methods. J Clin Chem Clin Biochem 1983; 21:709–20. Sciacovelli L, Zardo L, Secchiero S, Zaninotto M, Plebani M. Interpretative comments and reference ranges in EQA Programs as a tool for improving laboratory appropriateness and effectiveness. Clin Chim Acta 2003,333: 209–19. Lim EM, Sikaris KA, Gill J, Calleja J, Hickman PE, Beilby J, et al. Quality assessment of interpretative commenting in clinical chemistry. Clin Chem 2004;50:632–7. Standard of medical care in diabetes. Diabetes Care 2004:27(Suppl 1);S15–35.