Defini&on of EQAS allowable limits for HBA1C: a candidate method

Transcription

1 Defini&on of EQAS allowable limits for HBA1C: a candidate method A. VASSAULT 1, S. ALBAREDE 2, P. JOLY 3, JP SIEST 4, M. FONFREDE 1, JC EYNARD 3, J. de GRAEVE 2, B. POGGI 3 1- ASQUALAB, 8 rue Maria Helena Vieira Da Silva PARISh<p:// 2- CTCB, 33 route de Bayonne, TOULOUSE h<p:// 3- PROBIOQUAL, 7 Rue Antoine Lumière LYONh<p:// 4- BIOLOGIE PROSPECTIVE, 3 Route de l aviayon- BP VILLERS LES NANCY- h<p:// As members of FEDERATION of EQAS ORGANISERS (FAEEQ) A collaborafve study. EQALM SYMPOSIUM DUBLIN 2017

2 PLAN 1. Introduc&on 2. Objec&ves 3. Materials and methods 4. Results and discussion: According to the method According to the concentra&on level According to the material used 5. Proposals 6. Conclusion

3 INTRODUCTION, SCOPE and PURPOSE: EQAS are implemented to evaluate the reliability of the results provided by medical laboratories and, if necessary, helping to improve the prac;ces. Assessment of results provided by the par;cipants laboratories to EQAS programs, depends on the value defined to be allowable for their intended use. HBA 1C has been chosen as a model because of: The value of this test for the follow up of diabe;c pa;ents The importance of diabetes disease in the world The high level of standardisa;on for this test: reference method, reference material Criteria to evaluate the performances of HBA 1C measurements are not harmonised in France, ranging from 6 to 8% according to the organiser. Total error based on biological varia;on is +/- 3%.

4 Objec&ves To harmonise criteria used by EQAS to evaluate the quality of the results for HBA1c. To evaluate the behaviour of different QC samples. To evaluate the impact of the method used to evaluate the impact of the determina;on of allowable limits (AL) on the classifica;on of the par;cipants results.

5 MATERIALS AND METHODS: Data from 4 FRENCH EQAS non profit organisa;ons including the results provided during 2015, 2016 and 2017 by a thousand laboratories for 1250 sites data were collected and analysed using the same model according to: - The level of concentra;on, - The nature of QC samples and - The method used. The number of acceptable results was determined for different allowable limits values ranging from 2 to 20 %. Quality control (QC) samples were lyophilized haemolysed human blood from different commercial origin (AALTO, BEATRIX, BIOLABO, DIASERVE, EUROTROL, POLYMED, RANDOX..). A fresh human blood sample was also provided. Results are also evaluated according to the methods used (HPLC, immunological, enzyma;c...).

6 Preliminary statement Switch to IFCC units is not yet accepted in France. So, all the data presented are expressed using percentage as units. For samples with assigned values using a reference method, general means do not demonstrate difference up to 1% as compared to the reference value So, the results are compared to the general mean. A few data for enzyma;c method, affinity chromatography and for DCA 2000 were collected.

7 Methods used by the par&cipants total: 1250 lab sites 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 16% 24% 23% 22% 30% 27% 22% 14% 34% 29% 52% 49% 53% 42% 46% BP ASQUALAB CTCB PBQ ALL Enzyma&c Immunology Capillary electrophoresis HPLC

8 Allowable limits actually proposed ASQUALAB BIOLOGIE PROSPECTIVE % +/-8 Target value + peer group +/-7 Target value + peer group CTCB +/-8 Target value + peer group PBQ +/-6 Target value + peer group QUALAB PROPOSAL BP Allowable total error RICOS NGSP +/- 15% +/- 12% +/- 3% +/- 6%

9 Results and discussion Allowable limits - varia&on factors to evaluate: Concentra&on levels Analy&cal method Nature of QC materials

10 HbA1c EQAS - ASQUALAB 2017 (Aalto) - HPLC Method 100% 90% % of acceptable results 80% 70% 60% 50% 40% 30% 20% 10% HbA1c = 5,8% HbA1c = 6,6% HbA1c = 7,3% HbA1c = 8,1% HbA1c = 8,7% HbA1c = 9,4% 0% Allowable limits (%)

11 HbA1c EQAS - Biologie Prospec&ve 2016 (Beatrix) - HPLC Method 100% 90% % of acceptable results 80% 70% 60% 50% 40% 30% 20% 10% HbA1c = 5,14% HbA1c = 5,56% HbA1c = 6% HbA1c = 6,39% HbA1c = 6,88% HbA1c = 7,3% HbA1c = 7,7% HbA1c = 8,12% HbA1c = 8,5% HbA1c = 8,87% HbA1c = 9,3% 0% Allowable limits (%) HbA1c = 9,76%

12 EEQ HbA1c PROBIOQUAL Evolu&on of the results according to the acceptable limits (%)- different kind of QC samples 100% 90% 80% % of acceptable results 70% 60% 50% 40% 30% 20% 10% 0% 0,0 1,0 2,0 3,0 4,0 5,0 6,0 7,0 8,0 9,0 10,0 11,0 12,0 13,0 14,0 15,0 16,0 17,0 18,0 19,0 20,0 HbA1c = 5,63% HbA1c = 5,66% HbA1c = 5,66% HbA1c = 5,9% HbA1c = 6,16% HbA1c = 6,54% HbA1c = 6,59% HbA1c = 7,78% HbA1c = 10,28% HbA1c = 10,59% HbA1c = 11,63%

13 PROBIOQUAL EQAS 2016 (same sample) 100% 90% 80% % of acceptable results 70% 60% 50% 40% 30% 20% 10% 0% Allowable limits (%) Capillary electrophoresis HPLC Enzyma;c Immunology

14 EQAS (2016 data) - Evalua&on of par&cipants results according to allowable limits (+/- 5 to +/- 20%) % of allowable results 100% 98% 96% 94% 92% 90% 88% 86% 84% 82% 80% 78% 76% 74% 72% 70% 5,0 6,0 7,0 8,0 9,0 10,0 11,0 12,0 13,0 14,0 15,0 16,0 17,0 18,0 19,0 20,0 Allowable limits (%) Human lyophilized erythrocytes Human whole total Human lyophilized hemolysate Lyophilized blood

15 Unsa&sfied results AL: +/-6 % HPLC Capillary Electr. Lyophilised 1 1% 1% 15% Lyophilised 2 8% 3% 40% Lyophilised 3 8% 10% 29% HUMAN WHOLE BLOOD 1% 0% 8% Imunologic methods AL: +/-8 % HPLC Capillary Electr. Lyophilised 1 0% 0% 1% Lyophilised 2 1% 0% 30% Lyophilised 3 1% 1% 10% HUMAN WHOLE BLOOD 1% 0% 8% Imunologic methods

16 Acceptable Limits for 90% and 95% of the par&cipants LYO 1 LYO 2 LYO 3 LYO 4 General mean HBA1c (%) 90% of sa&sfactory results 6,60 6,42 7,08 6,40 6,5% 7,5% 7,5% 7,0% 95% of sa&sfactory results 9,5% 9,5% 10 % 9,0%

17 RESULTS AND DISCUSSION: The number of acceptable results does not depend on the concentra;on level but greatly on the control materials used. For AL at 6%, results obtained for all data provided with fresh human blood demonstrate higher performance: 97% of the results are acceptable instead of 80-86% when using lyophilised haemolysed blood.

18 CONCLUSION: The consequence of the defini;on of allowable limits was evaluated for different methods, different QC matrices and different concentra;on levels. Between methods varia;on observed discrepancies depends on the QC matrix material. No influence of the concentra;on level was observed. The impact of QC materials behaviour (commutability) used as compared to fresh human blood is shown to be one of the major factors of varia;on. Furthermore, the quality of the results observed depends on the analy;cal method used by the par;cipants. The method used for this study could be used as a model for harmoniza;on of the allowable limits taking into account the state of the art.

19 Proposal Acceptance criteria used in European external quality assessment schemes (EQAS) for HBA1C concentra;ons showed a high level of varia&on (+/- 6 to 15%). The percentage of unsa&sfactory results obtained by the different organizers for the data studied varied widely according to. For this reason, we propose to implement a working group from the European Organisa;on for External Quality Assurance Providers in Laboratory Medicine (EQALM) with the aim of establishing appropriate acceptance limits (AL) allowing harmoniza;on between the evalua;on procedures of European EQAS organizers.

20 THANK YOU FOR ATTENTION