An Objective Treatment of QC using LabWare. Robyn Lale Cairns Regional Council

Transcription

1 An Objective Treatment of QC using LabWare Robyn Lale Cairns Regional Council

2 Who We Are Laboratory Services at Cairns Regional Council. The laboratory has grown from 7 people in 1995 to having a current full time staff of 19 Laboratory operations comprises activities of sampling, chemical and biological i l analysis of mostly water and wastewater

3 Service Area

4 Our situation with LabWare Started in February 2011 Went LIVE Feb 13, 2012 Cyclone Yasi (Feb 2011)

5 QC Review whatever the means used to determine the acceptability of results or test data, it is important thatpre determined determined, objective acceptance criteria are used to make decisions. (NATA tech note # 23 p6)

6 Previous System Initially developed in 1995 based on a manual scoring system used in PNG QC Score 2 acceptable maximise the # of valid batches while maintaining i i levels l of precision ii and accuracy expected by the client Any score over 0 needs to be reviewed and justified by analyst or signatory

7 Previous System (cont ) Based around an analytical batch Batch type is pre defined including QC samples (NATA tech note # 23) Objective limits and acceptance criteria also dfi definedd Fail score and not done score defined

8 Previous System (cont ) QC material low level i.e. 20 times detection limit high level i.e. > 20 times detection limit Fail Score duplicates 25% 10% 1 1 sub samples 25% 10% 1 1 replicate test portions ( rd ) 25% 10% 1 1 replicate aliquots ( ra ) 25% 10% 1 1 previous sample (test portion) 25% 10% 1 1 previous sample (sub sample) 25% 10% 1 1 Matrix spike recovery ( mspk ) 75 % 125 % 90 % 110 % 1 1 Sample spike recovery ( spk ) 75 % 125 % 90 % 110 % 1 1 Not done score

9 Previous System (cont ) QC material Acceptable limits Fail Score Not done score Inhouse standard ( IHS ) Certified Reference Material ( CRM ) Mean and standard deviation (used to determine warning and action limits) are determined by the quantitation of a set of replicates (at least 10 points) Limits defined on the analysis certificate IHS have a range of checks performed based on a simplified range of Western Electric Rules (1956) (D. Brynn Hibbert, Quality Assurance for the Analytical Chemistry Laboratory, (New York: Oxford University Press), 2007, p118) One point lies outside a control limit Two consecutive points lie between a warning and its control limit Seven consecutive points onthesame side ofthecentre line Seven consecutive points are increasing or decreasing

10 Previous System (cont ) Create Analysis Batch Re-analysis of selected individual tests and/or samples or run confirmatory tests Run analysis and Log data yes Admin Officer to input any calculations and print Analyte suitable for reanalysis? no Problem Identification yes Score of 2 or less Analyst to check calculation input and calculate QC score Score of 4 or greater First time done? no Score of 3 Explain the reason for a score of 1 or 2 Non-conformance Evaluation And Quality Improvement (see LI0075) NATA Signatory assessment of reasonableness of dataset Assessment of overall data reasonableness for submitted sample group yes Accept Batch? no yes Accept Data? no Analyte suitable for reanalysis? yes Report Results Reject results for individual samples / whole batch Document Reasons for failing sample(s) no

11 Our Expectations with Labware vague wish, but not much expectation that Labware could be configured to give us exactly the same thing thought we would have to institute a completely new process of batch (QC) review we thought we would have to change our process to fit with the new LIMS system

12 Labware s Response During Scoping Phase: While all the concepts are well known and used at many laboratories, this is the first time I have seen where a formal evaluation of QC results have been performed and recorded. The configuration of these procedures in LabWare will be a stimulating exercise and though challenging, I expect a successful outcome (CRC Gap Analysis Study. LabWare LIMS)

13 Labware Implementation

14 Labware Tables

15 Database New Tables Area Structural Requirements Description of New Features C_BATCH_QC_SCORE_SET C_BATCH_QC_SCORERULE C_BATCH_QC_SCORE Database New Columns to BATCH_HDR_TEMPLATE.C_BATCH_QC_SCORE_SET standard tables BATCH.C_QC_SCORE_TOTAL BATCH.C_QC_SCORE_MAX_COMP Lists QC_SCORE_RULES Table Template None required Menu Routines BM_QC_SCORE_CALCULATE CALCULATE BM_QC_SCORE_DISPLAY Subroutines BM_QC_SCORE_CALCULATE BM_QC_SCORE_CALC_FAILS _ BM_QC_SCORE_CALC_WER BM_QC_SCORE_DISPLAY Function Privileges No new Function Privileges have been added.

16 Defining the QC Scoring Rules

17 Defining the QC Samples (this actually happens first)

18 Defining the QC Samples Set up PRODUCT Specs

19 Defining the QC Samples Set up a Standard/Reagent

20 Setting up the Batch Standards

21 Linking the scoring to the Batch setup

22 Analytical Process Workflow Create the batch (fixed standards are added automatically) Add non fixed QC samples (replicates and spikes) Perform analysis and enter results Calculate batch (determines % recovery and %RSD for spikes and replicates) Calculate the QC score and close the batch The score can be re calculated if necessary (e.g. forgot to enter result before calculating) l The score can be adjusted if necessary Accept/Reject the batch (with appropriate comments)

23 Summary of QC Score calculation process Check that all 'Fixed Standards' are present on the Batch and they are complete or authorised, where, for the QC Type, the 'Not Done Score' > 0. For each QC Sample on the Batch, apply the QC Rules specified for the same QC Type as that of the QC Sample. Rules are applied to each of the reportable results on a QC Sample. Each QC Rule failure for a result is written to the table C_BATCH_QC_SCORE The total of all (fail) scores is recorded against the Batch. The scores for each component are totalled separately and the maximum component total is recorded against the Batch. Evaluation Rules for reviewing batches are based on the Maximum Component Total.

24 QC Score Summary Only those QC materials that failed are displayed d on the summary

25 QC Review Process If score < 3, then the analyst (if also a signatory) can authorise the batch without input from a Scientist. If not, batch can t be authorised except by a signatory other than the analyst Our target of passed batches is 95% greatly increases the efficiency of the lab analyst feels greater ownership and responsibility for results it s their signature on the COA identifies who needs more training

26 Benefits with Labware Didn t have to change the basic philosophy of our QC process that had been endorsed by NATA Saved considerable effort (e.g. in process documentation) Didn t have to re train staff in QC evaluation procedures Didn t t haveto re convince NATA More robust application of our QC system Analyst can see at once if a QC material passed Batch QC data isn t re keyed into a separate QC computer application True multi user functionality

27 Benefits with Labware (cont ) More scope to view result history graphically

28 Challenges Different groupings of our old analyses into Labware tests and components A TtlQC Total Score AMax Component Score

29 NATA s reaction to the QC system Have had a NATA technical and quality system audit of our lab since Labware (both chemical and biological) They LOVE the QC system (even the microbiologists!!) They can easily see (at a click of the mouse) Everything required is included The results of objective evaluation QC result history

30 Future Enhancements Planned link the QC score into some form of competency based staff training scheme using Labware Currently staff are deemed competent when they have scored a 0 in at least 3 unsupervised batches

31 Wrap up The flexibility and configurability of Labware allowed us to retain a system that had worked well for us for many years

32 Questions?