What I will talk about

Transcription

1 EQA of POCT methods: How to interpret results? Anne Stavelin Labquality Days, Helsinki 12 February Norwegian quality improvement of primary care laboratories What I will talk about What is special with quality control for POCT What is the difference between EQA for hospitals and for POCT How to interpret the EQA result Different EQA categories Examples 1

2 Quality Control of POCT Challenges Easy to use - everybody can do it Often used by people with little or no lab experience The instrument will give an error message if something is wrong No mistakes can be done Expensive to perform controls What is the advantage of performing controls can you prove it? Quality Control of POCT Built-in-controls/on-board QC systems Integrated control systems, both in the instrument (electronic controls) and in each test strip/cuvette Such controls are insufficient (Collective opinion paper, CCLM 2011;49: ) Internal quality control (IQC) Commercial control materials No published evidence that this is useful (Holt et al. IQC of POCT: Where is the evidence? Ann Clin Biochem 2015: epub October 20) External quality assessment (EQA) Schemes provided by an EQA organizer 2

3 Basic theory of EQA The aim is to detect analytical errors (and make corrective actions) to monitor the analytical quality to document the analytical quality Assessment of each single laboratory Assessment of the methods Is participation in EQA useful? Urine albumin POCT scheme from 1998 to 2012 Very good performance Poor performance Bukve T et al. Clin Chem Lab Med 2015; 53(1):

4 What to look for in an EQA? Sample design and frequency Analytical goals Easy to read reports Education Scientific support EQA of POCT - challenges Large number of participants Little or no lab experience Direct communication with the users (clinicians, nurses, patients) Different control materials to different POCT devices cannot compare results between instruments Lack of control material to some instruments 4

5 Example EQA for POCT INR 30 European countries were invited, 28 responded YES: 9 countries (12 EQA organizers) No: 19 countries Wide variation in scheme design! Stavelin et al. External quality assessment of point-of-care International Normalized Ratio (INR) testing in Europe. Clin Chem Lab Med 2012; 50: Example EQA for POCT INR Number of participants (2011) NEQAS (UK): 2662 MQ (Switzerland): 2264 CSCQ (Switzerland): 2232 Control materials Instrument-specific materials Lyophilized plasma, lyophilized whole blood, fresh frozen plasma, artificial liquid material, capillary whole blood (patient slit sample) No material available for one instr. special EQA approaches Performance specifications (analytical goals) Varied from 12% to 30% Not a POCT specific challenge only 5

6 How to interpret the EQA results? Key points 1. The control material Fresh material, stabilized material, lyophilized material, other? Commutable material? 2. The target value Reference target value? Overall mean (or median) target value? Method-specific target value? 3. The number of replicates One? (most common) Two? More? EQA categories 6

7 EQA categories EQA Commutable Reference Replicate Assessment of the analytical quality of Assessment of the single laboratory category material target value measurements methods 1 Yes Yes Yes 2 Yes Yes No 3 Yes No Yes 4 Yes No No 5 No No Yes 6 No No No Modified after Miller et al. Clin Chem 2011 Terminology Theodorsson. Bioanalysis 2012; 4:

8 EQA categories EQA Commutable Reference Replicate Assessment of the analytical quality of Assessment of the single laboratory category material target value measurements methods 1 Yes Yes Yes Trueness Harmonization of methods Trueness Comparison with other methods Precision (repeatability) Precision (reproducibility) Precision (repeatability) Comparison within own method 2 Yes Yes No Trueness Harmonization of methods Precision (reproducibility) Accuracy Comparison with other methods Comparison within own method 3 Yes No Yes Harmonization of methods Precision (repeatability) Precision (reproducibility) Comparison with other methods Precision (repeatability) Comparison within own method 4 Yes No No Harmonization of methods Comparison with other methods Precision (reproducibility) Comparison within own method 5 No No Yes Precision (repeatability) Precision (repeatability) Precision (reproducibility) Comparison within own method 6 No No No Precision (reproducibility) Comparison within own method Modified after Miller et al. Clin Chem 2011 Example Category 1 (HbA1c) Noklus EQA survey Nov

9 Example Category 1 (HbA1c) EQA survey Nov 2015 Reference target value 5.21 % HbA1c Number Repeatability CV% Reproducibility CV% Median Afinion 5.25 % HbA1c Median DCA 5.10 % HbA1c Example Category 5 (INR) EQA survey Oct 2015 CoaguChek Simple Simon Thrombotrack Median (target) Number Repeatability CV% Reproducibility CV% CoaguChek 2.30 INR Simple Simon 3.95 INR Thrombotrack 2.50 INR

10 New EQA model for POCT New EQA model for POCT Native patient samples approx. 100 samples for each POCT instrument analyzed by general practitioners assessment of systematic deviation (bias) between the POCT instruments and a designated hospital instrument Non-commutable control material assessment of single participant performance from a methodspecific target value Combined assessment 10

11 New EQA model - examples Ctr 1 Ctr 2 Survey 1 Survey 2 Survey 3 Survey 4 The POCT method: Too high values The participant: OK New EQA model - examples Ctr 1 Ctr 2 Survey 1 Survey 2 Survey 3 Survey 4 The POCT method: OK The participant: Too high values 11

12 New EQA model - examples Ctr 1 Ctr 2 Survey 1 Survey 2 Survey 3 Survey 4 The POCT method: OK The participant: Both high and low values The total testing process Examples correct identification of the patient appropriate test selection obtaining a satisfactory specimen analyze and record the results promptly and correctly interpret the result accurately taking appropriate action documenting all procedures and maintaining accurate records Anamnesis, clinical findings Pre-pre analytical phase Pre-analytical phase Patient Clinician Analytical phase Diagnosing, monitoring, treatment Post-post analytical phase Post-analytical phase 12

13 NOKLUS a POCT organization 2855 participants, voluntarily participation 1725 GPs offices (99,8%) 848 (96 %) nursing homes 31 military installations 48 oil platforms 120 others 88 Norwegian Hospitals 60 % of the participants are visited each year 350 courses with 5254 participants 34 EQA schemes Take home message There are many challenges in performing quality control of POCT How to interpret the EQA result is dependent on The control material The target value (and acceptability limits) The number of replicates A novel EQA model is developed in order to improve schemes in category 5 and 6 The total testing process should be controlled, not only the analytical phase 13