Latin-American Biological Dosimetry Network (LBDNET).Evaluation through triage and conventional scoring criteria. Development of a new approach for statistical data analysis

M. DIGIORGIO; M. VALLERGA; A. RADL; M. TAJA; A. SEOANE; J. DE LUCA; M. STUCK OLIVEIRA; P. VALDIVIA; O. GARCÍA LIMA; A. LAMADRID; J. GONZÁLEZ MESA; I. ROMERO AGUILERA; T. MANDINA CARDOSO; C. GUERRERO CARVAJA; C. ARCEO MALDONADO; M. ESPINOZA; W. MARTÍNEZ LÓPEZ; M. DI TOMASSO; F. BARQUINERO; L. ROY; D. LLOYD; C. LINDHOLM; H. ROMM

Buenos Aires. 19 al 24 de octubre de 2008

Congreso; 12th International Congress of the International Radiation Protection Association; 2008

International Radiation Protection Association (IRPA)

iological Dosimetry is a necessary support for National Radiation Protection programs and Emergency Response Schemes. A Latin-American Biological Dosimetry Network (LBDNET) has been constituted by the biological dosimetry laboratories from: Argentina, Brazil, Chile, Cuba, México, Perú, and Uruguay (1AEA Regional Project RLA9/054, 2007). The biological dosimetry laboratory of´ Argentina organized an international biological dosimetry intercomparison for the analysis of some relevant parameters involved in dose assessment, to reinforce the response capability in accidental situation requiring the activation of mutual assistance mechanisms and thus, consulting the bases of the  LBDNET organization. General Objectives To evaluate the inter-laboratory reproducibility. To identify potential difficulties and to promote the necessary modifications for the collaborative task. Specific Objectives To initiate, with this exercise, periodic intereomparisons among LBDNET participants. To develop technical competence of associated laboratories inside the country. INTERCOMPARISON PROGRAM Participant laboratories from the LBDNET, Spain (UAB), France (IRSN), United Kingdom (UPA), Finland (STUK) and Germany (BIS). ACTIVITIES: Are at present in progress and should be concluded in June 2008. 1. Human blood samples were irradiated in viiro with a 60Co source (Regional Reference Center for dosimetry of the Atomic Energy Commission), two points of doses and a control. Cultures and slide preparations were performed according lo standard methods. 2.     Samples were distributed to the laboratories. 3. Analyzing the samples: In progress in each participant laboratory. The following data are required: -Number of dicentrics observed in 50 cells/30 dicentrics (Triage scoring criteria), 100 cells (as an intermediate step) and 500 cells/100 dicentrics (conventional scoring criteria). In the case of two scorers: 250 + 250 cells each; reporting raw data, frequency and standard deviation associated lo the observations. -Dose estimates with 95% confidence limits, reporting the calibration curve applied (coefficients and its standard deviations). -Distribution of aberrations in the analyzed cells, relative variance, Papworth u test Index. 4, Statistical analysis: The presence of individual laboratories or values that appear to be inconsistent with all other laboratories may change the estimates and thus, decisions have to be taken. To discard or correct inconsistent values, two approaches will be used (ISO 5725-2 /5): -Numerical outlier tests (Cochran and Grubbs tests): To discard data that give rise to a test statistic that exceeds the critical value of the test at the 1% significance level. -Robust methods for data analysis: To yield robust values of the average and standard deviation of the data. 5. Estimating the mean value and the standard deviation inter-laboratory: To estimate the parameters once outliers are discarded or corrected 6. Determining laboratory’s performance: To calculate z-score parameter from the laboratory results, the reference value and the estimated standard deviation. To determine u-score parameter, which evaluation includes both participant measurements and reference value uncertainties. 7. Reporting results: To allow to irnplement corrective actions to reinforce service capabilities of the laboratories. PRELIMINARY CONCLUSIONS: Proficiency tests are essential tools for the quality assurance of the laboratories as they constitute an objective evaluation of its performance, from both human and technical point of view. Actions for a steady quality assurance and quality control, such as this Intercomparison exercise, will be the technical support of the LBDNET. Additionally, it will contribute in the accrediting process (ISO 19238:2004 and 1SO/1EC 17025:2005 requirements). KEYWORDS: Biological Dosimetry; Intercomparison Exercise; Triage; Statistical Data Analysis.