ANQAP Information

2012 Information

Contact ANQAP

Participant Information

Other Information

Participating Laboratory Information

Introduction

The Australian National Quality Assurance Program (ANQAP) is coordinated at the Victorian Department of Primary Industries on behalf of the Sub-Committee of Animal Health Laboratory Standards (SCAHLS). 

The major focus of the program is proficiency testing of assays used in quarantine, export certification and national disease control programs. 

Sample panels are forwarded to laboratories as per the Annual Test Timetable. ANQAP aims to provide a high quality service: 

·       Test samples are selected to be homogeneous and stable, to minimise result variability within and between laboratories.

·       Test samples will be, as far as is technically possible, matrix samples of a similar type to those routinely analysed by most participating laboratories.

·        The levels of target analytes in the samples distributed are selected to represent levels that would bee measured by most participating laboratories.

·         The results from all laboratories are collated and evaluated, and reports prepared.

·         Statistical analysis of the results will be undertaken where practicable.

Confidentiality

To protect the identity of participating laboratories, each laboratory will be allocated a confidential code number.  The identity and codes of all participating laboratories will be known only to ANQAP staff. Laboratories are advised in writing of their confidential code number at the beginning of the year’s program. This laboratory identification is changed annually.

Tests included in the program

Refer to the Annual Test Timetable

All participants will be notified in writing if ANQAP needs to alter the timetable.

Enrolment

 Registration

Enrolment paperwork is forwarded to interested laboratories before each new cycle. Late enrolments can be accepted however a laboratory may be too late to be included in some of the scheduled tests.  Invoices are payable within 30 days from the date of invoice.   Sample dispatch and/or reports may be delayed where payment is outstanding.

 Cancellation of participation

If a laboratory wishes to cancel participation in the program, a partial refund may be possible.  Notification of cancellation of participation must be in writing. ANQAP will seek feedback from a laboratory that withdraws from the program.

Scheduled Testing

 Sample dispatch

Refer to the Annual Test Timetable 2010. Laboratories will be advised in writing prior to ANQAP dispatching samples. For each test ANQAP will provide 6 freeze-dried samples. Each sample will be clearly labelled with the test name, sample number and the nature of the sample where required. The sample panels are accompanied by a specimen advice form. 

Samples are forwarded at ambient air temperature. This method of transport and the mode of packaging has been validated by ANQAP and will not harm freeze dried samples. However, ANQAP suggests samples are stored at –20^oC until required for testing. Instructions for storage requirements for the Aquatic samples are provided on the Specimen Advice Form.

Testing

Laboratories should test each sample in triplicate within the same test run and a single operator. Where triplicate testing requires a greater volume of sample than the 0.5mL provided by ANQAP the laboratory is asked to test in duplicate. Extraction for PCR tests should be done once for each sample with the extracted sample tested in triplicate within the same test run. Tests that require titration of samples must be taken out to read the endpoint for any positive samples identified. This is necessary for ANQAP to compare results, determine a consensus median and provide an accurate classification for each laboratory.

 Laboratories should treat the ANQAP samples as routine diagnostic specimens. They should be tested in-house and not sub-contracted to another laboratory. Results should be treated in confidence by participant laboratories.

  Reporting results

The current version of the Result Reporting Form QF 15 for Veterinary and QF 16 for Aquatic testing must be used. Results can be reported electronically and emailed to ANQAP or the form may be photocopied, filled in by hand and faxed to ANQAP.

 Results should be reported:

·               on time - late results will not be accepted unless ANQAP has been contacted before the due date with a reasonable explanation for the delay

·               in a legible hand or typed to avoid transcription errors

·              including details of the method used, the assay operator, the laboratory name, cut-off limits, raw and calculated results including means of replicate results where appropriate and interpretation of results (positive / negative).  Any sample or test problems should also be described.

 Results should be reported as they would be for routine diagnostic samples.

 Results not submitted in a reasonable time will be classified as ‘U’, unable to submit results.

When an unsatisfactory result is reported to a participant laboratory, the cause should be investigated and corrective action taken where appropriate.

Retests

ANQAP will calculate the result for each sample and an acceptable variation range.  Laboratories reporting results outside the acceptable variation range (AVR) will be required to retest.  For tests that are classified on interpretation only, laboratories will be asked to retest if their reported interpretation differs from the consensus. A retest report and notification of dispatch of retest samples and an expected report date will be forwarded.  Retest results should be reported to ANQAP on the Result Reporting Form (QF15 and or QF 16).

ANQAP reports 

ANQAP will aim to issue a report to each participating laboratory within two weeks of the results due deadline. Separate reports will be prepared for each test and will indicate whether the testing was satisfactory or unsatisfactory.  The report also includes a summary table of results from all participating laboratories and where appropriate will also include test statistics (generally CFT, VNT and HI tests with sufficient participant numbers): split-level analysis will be done on two samples. 

ANQAP makes every effort to avoid transcription errors however participants are requested to check that their results are entered correctly. Should any errors be identified laboratories are asked to inform ANQAP as soon as possible. An amended report will be forwarded to the affected laboratory as soon as possible.  Participants are encouraged to contact ANQAP to discuss reports if further clarification is required or if they disagree with the report in any way.

 On submission of all results a median value will be calculated for each sample. For many of the tests monitored, an acceptable variation range (AVR) will be calculated from the median and results should fall within this range. This forms the basis on which laboratories are classified for each test.  Tests reported as positive or negative will not have an acceptable variation range and results will be assessed on correct interpretation.

 Results will be classified as:    

✓         Initial results submitted by this laboratory are satisfactory as they fall within the acceptable variation range from the  median.          

R         Initial results submitted by this laboratory are outside the acceptable variation range from the median and retesting is required. 

U         Initial results were not submitted by this laboratory.

Retest results are classified as: 

R✓     Retest results submitted by this laboratory are satisfactory as they fall within the acceptable variation range or match the consensus interpretation.           

Rm      Retest results submitted by this laboratory demonstrate minor variation as results fall just outside the acceptable variation range according to the defined criteria; these results are acceptable, however the minor variation is worthy of note.           

RU       Retest results are not submitted or retest results submitted by this laboratory are outside the acceptable variation range or disagree with the consensus interpretation  and are unsatisfactory.

Determination of the Acceptable Variation Range (AVR)

 CFT, MAT, HI and VNT

The acceptable range for the CFT, MAT, HI and VNT is a dilution factor either side of the median value. 

 Example:  A CFT median value of 3/32.The AVR is 3/16 to 3/64.

 Results falling within this range will be classified as:

Satisfactory (✓) 

Results falling outside the acceptable variation range are temporarily classified as:

Retest Required (R) 

Following retesting the result will be classified as:

Retest Result Satisfactory (R✓) - result within the AVR

Retest Result Demonstrating Minor Variation (Rm) - result within 3/8-3/128

Retest Result Unacceptable (RU) - result  ≤ 2/8 -  ≥ 4/128

 AGID and RBPT

The acceptable variation range for AGID and RBPT results is one grading on either side of the median value. The exception to this is where the Consensus Median (CM) value calculated for the sample is a low positive, such as 1+. The AVR cannot include a non-positive value, thus the acceptable range will be identified as1+ to 2+.

 Example: An AGID median value of 3+. The AVR is 2+ to >3+.

 Results falling within this range are classified as:

Satisfactory (✓)

 A negative result or a result of 1+ would be classified as:

Retest Required (R)

and the laboratory would be required to retest

 Following retesting, results would be classified as:

Retest Result Satisfactory (R✓) - result within the AVR

Retest Result Demonstrating Minor Variation (Rm) - for result within 1+ - >3+

Retest Result Unacceptable (RU) - for result outside the Rm

 ELISA and PCR

The ELISA's included in the program will follow different methods and will vary in the interpretation of positive and negative results. ANQAP will report and classify laboratories based on the consensus interpretation, which will be positive or negative for most ELISA’s. Some tests may also include ranges such as low / high positive, doubtful, suspicious and inconclusive ranges. In determining the consensus interpretation ANQAP will also take into account the previous testing history of the sample.

To be classified as Satisfactory, a laboratory will be expected to report a positive result for a positive sample and a negative result for a negative sample for ELISA and PCR tests.

Participants are expected to complete all calculations following the method used and report to ANQAP the interpretation, raw OD values and any further calculations (percent inhibition, ELISA Units and ELISA ratio, CT values for PCR.

 SAT

The acceptable variation range for the SAT is a dilution factor either side of the median.

Example:  An SAT median of 160 IU (2+ at 1/80). The AVR is 80 IU (2+ at 1/40) to 320 IU (2+ at 1/160). 

Results falling within this range will be classified as:

Satisfactory (✓)

 Results falling outside the acceptable variation range are temporarily classified as;

Retest Required (R) 

Following retesting the result will be classified as:

Retest Result Satisfactory (R✓) - result within the AVR

Retest Result Demonstrating Minor Variation (Rm) - result within range 40IU (2+@1/20) to 640IU (2+@1/320)

Retest Result Unacceptable (RU) - result <40IU (2+@1/20) to >640IU (2+@1/320) 

Report Statistics 

Where statistical analysis is appropriate this will be done on two samples (split-level analysis). The samples must have results with only minor differences from each other. ANQAPs statistical analyses identify laboratories by their confidential code number. 

Youden plot

A Youden plot shows the result of one sample as a function of the result of the other sample in a sample pair. The Youden plot gives an idea of the dominating sources of error in the results.

·               Laboratories with results in the upper left or lower right hand corner of the diagram have results dominated by random error and little systematic variation.

·               Laboratories with results in the upper right or lower left hand corner of the diagram have results dominated by systematic error.

·               An ellipse along the axis indicates results have been significantly affected by random variation for one of the samples.

Z-scores 

Z-scores are calculated using the robust summary statistics (median and normalised IQR). For each pair of results two Z-scores are obtained - a between laboratory Z-score and a within laboratory Z-score. The values are tabulated in a results table, and results are graphically represented in bar charts. Two charts are generated, one for the between laboratory Z score and another for the within laboratory Z score.

Results are plotted in order of magnitude. Lines at +3 and -3 are included so that outliers can be clearly identified.

·        A Z-score close to zero means that a result agrees well with the other laboratories.

·        A Z-score greater than 3 (Z >3 or Z <-3) identifies a result which demonstrates significant variation from other laboratories. These results are identified as outliers.

Z-scores also provide information on the type of variation. 

Between laboratory Z-scores are based on the sum of the results and describe the variation between all laboratory results (reproducibility). A between laboratory outlier indicates results that demonstrate significant variation from the other laboratories. 

·        Positive between laboratory Z-scores indicate results are above the median value. An outlier with a positive >3 Z-score indicates significant increased sensitivity. 

·        Negative between laboratory Z-scores indicate results are lower than the median value. An outlier with a negative <-3 indicates significant decreased sensitivity. 

Within laboratory variation is based on the difference between the sample pair and describes the variation within a laboratory (repeatability).  A within laboratory outlier indicates variation between the individual results submitted by that laboratory and low precision.

·        Positive within laboratory Z-scores indicates the difference between the laboratory’s sample pair is overestimated.

·        Negative within laboratory Z-scores indicates the difference between the laboratory’s sample pair is underestimated or has estimated the difference to be in the opposite direction to the median difference.

Contact Details 

Australian National Quality Assurance Program (ANQAP)

Department of Primary Industries

475 Mickleham Road,

Attwood, Victoria, 3049

Australia. 

Phone:  61  3  9217 4360

Fax :     61  3  9217 4371 

e-mail: anqap.quality@dpi.vic.gov.au