Clin Chem Lab Med 2006;44(9):1156–1157  2006 by Walter de Gruyter • Berlin • New York. DOI 10.1515/CCLM.2006.195 2006/149 Article in press - uncorrected proof Letter to the Editor ESEAP: the national External Quality Assessment Scheme for clinical chemistry in Greece and Cyprus Othon Panagiotakis 1 , Effie Anagnostou- Cacaras 1 , Gerard Jullien 1 , Angelos Evangelopoulos 2 , Alexander Haliassos 3 and Demetrios Rizos 4, * 1 Biochemistry Department, ‘‘Evangelismos’’ General Hospital, Athens, Greece 2 Lab Organization and QC consulting department, Roche (Hellas) S.A., Athens, Greece 3 Greek External Quality Assessment Scheme, Athens, Greece 4 Hormone Laboratory, ‘‘Aretaieion’’ University Hospital, Athens, Greece Keywords: clinical chemistry; external quality assess- ment (EQA); quality control. External quality assessment schemes have been operating almost worldwide for several decades (1–6) and they constitute one of the basic tools for quality assurance of laboratory results. These schemes con- tribute to the evaluation of the overall quality level of participants, as well as each individual laboratory’s quality level, allowing the comparison of inter-labor- atory performance and analytical methods. Moreover, they are educational, as they provide valuable infor- mation to participants that facilitates self-assessment (7). In Greece, the first national external quality assessment scheme in clinical chemistry was not firmly established until June 1994. This scheme, known in Greece and Cyprus under the name ‘‘ESEAP’’ (initials of the Greek words for National External Quality Assessment Scheme), was designed and implemented by three scientists from the Bio- chemical Department of ‘‘Evangelismos’’ Hospital, the largest public hospital in Athens. At the very begin- ning, approximately 100 laboratories in public hospi- tals, diagnostic centers and private laboratories all over Greece participated in the scheme. Today, the number of participants has reached 280 laboratories, including almost all public hospital laboratories in Greece and 40 laboratories in the public and private sectors in Cyprus. ESEAP operates a bimonthly distribution of two lyo- philized controls (A and B) of different concentrations *Corresponding author: Demetrios Rizos, PhD, Lecturer of Clinical Chemistry, 2nd Department of Obstetrics & Gynecology, Medical School, University of Athens, Hormone Laboratory, ‘‘Aretaieion’’ University Hospital, 76, Vas. Sophias Ave., 115 28, Athens, Greece Phone/Fax: q30-210-7286229, E-mail: drizos@aretaieio.uoa.gr to participating laboratories. These controls are pur- chased from In Vitro Diagnostics companies (SERO AS, Billingstad, Norway for cycles 1–5 and Bio-Rad Laboratories S.r.l., Segrate, Italy for cycle 6) based on the range of analytes and their stability. Each pair of controls is arbitrarily chosen. Participating laborato- ries determine the concentration of 18 analytes in the distributed controls (albumin, calcium, glucose, potassium, creatinine, total protein, sodium, urea, uric acid, triglycerides, total bilirubin, cholesterol, alkaline phosphatase, g-glutamyltransferase, lactate dehydro- genase, creatinine kinase, aspartate aminotrans- ferase, alanine aminotransferase). The results are reported back to the organizers for data registration and statistical processing. Then evaluation reports are printed and sent back to the participants in a booklet entitled ‘‘Results Reports’’, which includes 18 color printed pages, one for every analyte. One full scheme cycle consists of 12 control distributions and it is com- pleted within 2 years (six distributions annually). There is no replicate distribution of the same controls. Sixteen out of 18 analytes are statistically processed based on overall results, regardless of the analytical methodology. For alkaline phosphatase (ALP) and lac- tate dehydrogenase (LDH), where clear dependence of the measured value on the assay method is known, (buffer 2-amino-2-methyl-1-propanol or diethanol- amine for ALP and reaction order lactate to pyruvate or pyruvate to lactate for LDH), statistical processing is performed separately for each analytical method- ology. After the elimination of outliers, the ‘‘consen- sus’’ mean (8), namely the mean from all individual results, is used as the target value. Each page of the ‘‘Results Reports’’ refers to a dif- ferent analyte and includes the following information: the number of the distribution, date, analyte name and laboratory code. The summary statistics for each control (A and B) include: the number of the labora- tories that define the target value, the target value, the SD and CV of the overall results, as well as the results returned by the respective laboratory for controls A and B and the corresponding deviations from the tar- get value, which are expressed as both percentage deviation and a standard deviation index. For inter- pretation of the results, three different charts are also included in each page of the ‘‘Results Report’’: a fre- quency distribution histogram, a Levey-Jennings chart and a Youden plot (9). Apart from the regular assessment included in the results reports returned bimonthly to laboratories after each distribution, there is an additional assess- ment every 2 years, after completing a full cycle of 12 distributions. This assessment is performed accord- Brought to you by | University of Iowa Libraries Authenticated Download Date | 5/30/15 9:59 PM