Assessing Results against Precision Limits

Overview

Precision assessment is a component of quality assessment applied to duplicate and replicate QC samples, and pairs of blank, standard and spike QC samples.

Precision assessment occurs whenever a sample scheme analyte receives a result, or a result is cleared, and sets the sample scheme analyte's Precision Status.

Why is Precision Assessment Done?

Repeatability, intermediate repeatability and reproducibility are the fundamental approaches to expressing precision in a laboratory environment, that is, how possible it is to repeat a sample preparation (duplicated samples) or analytical preparation (replicated samples) to return the same results from analysis. There is no difference in respect to the mechanism of calculating precision for these three types as the differences lie only in when, where and how the data is gathered.

In CCLAS, the assessment of duplicate or replicate samples against the samples from which they were made, for reproducibility or repeatability is known as precision assessment, and sets the precision status for the analysis.

When a result is assessed against a reading from a paired sample and assessed against precision limits, the outcome sets the sample scheme analyte's Precision Status which is rolled into the sample scheme analyte's Composite QC Status. The result cannot be validated until any failure on this status is addressed.

Duplicate and Replicate QC Samples

Samples can be analysed in pairs, triplicates and so on, which can give a measure of the repeatability of the testing. These sample copies could be split from the original sample early in the sample preparation steps (in which they are called duplicates), or could be just be taken from the same container/packet at analysis time (in which case they are called replicates).

Pairs of samples are made from the same material and can be:

An unknown sample and a duplicate sample
An unknown sample and a replicate sample
A duplicate sample and a replicate sample
Two blank samples
Two standard samples
Two spike samples.

Reproducibility

Reproducibility is the measured outcome of a result recorded from a broader range of inputs. This is used to express precision where a sample is split at any point prior to analysis.

In CCLAS, this mode of measuring reproducibility is known as duplicate analysis.

A duplicate QC sample is created from a QC type that has an Analytical Type of Duplicate. A QC type can be linked to a QC Script Code where the script contains custom-based assessment rules that are run after the core precision assessment provided by CCLAS.

Variations in measuring reproducibility is managed by creating different QC types, each with an Analytical Type of Duplicate but a different assessment script.

Repeatability and Intermediate Precision

Repeatability is the measured outcome of a result recorded under identical conditions over a short interval of time (for example, a sample aliquot measured twice).

In CCLAS, this mode of measuring repeatability is known as replicate analysis.

Intermediate precision is similar to repeatability, the difference being that the repeat measurements may be made at a significant time interval (perhaps by a different analytical instrument) or, by a different operator. Twin stream sample analysis is a good example of this type of precision assessment.

A replicate QC sample is created from a QC type that has an Analytical Type of Replicate. A QC type can be linked to a QC Script Code where the script contains custom-based assessment rules that are run after the core precision assessment provided by CCLAS.

Variations in measuring repeatability is managed by creating different QC types, each with an Analytical Type of Replicate but a different assessment script.

Limits of Precision

A user with Laboratory Operator or QC Manager privileges can assess analytical results against precision limits. If a sample scheme analyte has defined precision limits for reproducibility and repeatability, then those limits are used for precision assessment.

Any sample scheme analyte that has a Data Type of Numeric, and linked to a parent sample in a duplicated way, that is, it is a duplicate of the parent sample, can be bound by precision limits determined from:

a statistical detection limit—The scheme version analyte's Duplicate Stat Det Limit is the statistical detection limit observed for duplicate samples.
a limiting repeatability—The scheme version analyte's Duplicate Limiting Repeatability is the limiting repeatability observed for duplicate samples.

Any sample scheme analyte that has a Data Type of Numeric, and linked to a parent sample in a replicated way, that is, it is a replicate of the parent sample, can be bound by precision limits determined from:

a statistical detection limit—The scheme version analyte's Replicate Stat Det Limit is the statistical detection limit observed for replicate samples.
a limiting repeatability—The scheme version analyte's Replicate Limiting Repeatability is the limiting repeatability observed for replicate samples.

When a result is received or cleared for a sample scheme analyte, precision assessment occurs. From this, the sample scheme analyte's Precision Status is returned as either Not Tested>, Not Required, Passed, Lower Warning, Upper Warning, Lower Failure or Upper Failure.

Process

Configuring Precision Assessment

Scheme version analytes are configured for QC assessment using the CCSCHM—Scheme application, which is also accessed using the Analysis Setup » Schemes menu option.

Create the analytical scheme. Refer to Maintaining Schemes.
Create the analyte to hold an analytical result by Maintaining Scheme Version Analytes, and during this:
1. Configure the analyte's Rep Statistical Detection Limit, Rep Limiting Repeatability, Dup Statistical Detection Limit and Dup Limiting Repeatability.
2. assessment.
3. Select the analyte's QC Active flag.
4. Select the analyte's Saved to QC History flag.
Publish the scheme version. Refer to Configuring Versions of a Scheme.
Where the scheme has org-scope:
1. Enable the org-scope scheme for use in the laboratory. Refer to Using Organisational Schemes vs Laboratory Schemes.
2. To specify the use of org-scope or lab-scope properties on the scheme version during accuracy assessment, Set the WORKBOOK_OVERRIDE_SCHEME_VERSION_LIST Preference.
3. To specify the use of org-scope or lab-scope properties on the scheme version analyte during accuracy assessment, Set the WORKBOOK_OVERRIDE_SCHEME_VERSION_ANALYTE_LIST Preference.

Effect of Scheme Scope upon Precision Assessment

Range tables are defined using the the CCRNGT—Range Table application, which is also accessed using the Analysis Setup » Range Tables menu option. Refer to Maintaining Range Tables.

QC types are defined using the the CCQCTP—QC Type application, which is also accessed using the Analysis Setup » QC Types menu option. Refer to Maintaining QC Types and Assessments.

QC masks are defined using the the CCQCMK—QC Mask application, which is also accessed using the Analysis Setup » QC Masks menu option. Refer to Maintaining Tray Container Types and Maintaining QC Masks.

Laboratories are defined using the CCLBRT—Laboratory application, which is also accessed using the System Setup » Laboratories menu option. Refer to Maintaining Laboratories.

The method used to determine how the difference is calculated between pairs of samples is set by preference.

Set the USE_REP_AVERAGE_FOR_PERCENT_DIFFERENCE Preference

The method used to determine which detection limit is used for precision assessment is set by preference.

Set the RPD_CALC_DL_TYPE Preference

Configuring Samples for Precision Assessment

The following conditions must be met for precision assessment to occur, to set the sample scheme analyte's Precision Status when a result is entered:

The sample is either a:
- Duplicate QC Sample—The sample's Primary Analytical Type is Duplicate.
- Replicate QC Sample—The sample's Primary Analytical Type is Replicate.
- Blank Duplicate QC Sample—The sample's Primary Analytical Type is Blank and Secondary Analytical Type is Duplicate.
- Blank Replicate QC Sample—The sample's Primary Analytical Type is Blank and Secondary Analytical Type is Replicate.
- Standard Duplicate QC Sample—The sample's Primary Analytical Type is The sample's Primary Analytical Type is Standard and Secondary Analytical Type is Duplicate.
- Standard Replicate QC Sample—The sample's Primary Analytical Type is Standard and Secondary Analytical Type is Replicate.
- Spike Duplicate QC Sample—The sample's Primary Analytical Type is Spike and Secondary Analytical Type is Duplicate.
- Spike Replicate QC Sample—The sample's Primary Analytical Type is Spike and Secondary Analytical Type is Replicate.
The sample is linked to a Duplicate Original Sample Code .
The scheme version analyte's Is QC Active is selected.
The scheme version analyte has a Dup Statistical Detection Limit, Dup Limiting Repeatability and Reproducibility, Reproducibility Range Table Code for the precision assessment of pairs of samples that have a duplicate relationship
A scheme version analyte has a Rep Statistical Detection Limit, Rep Limiting Repeatability and Repeatability, Repeatability Range Table Code for the precision assessment of pairs of samples that have a replicate relationship.

Refer to Creating QC Samples using QC Masks.

Where a sample scheme analyte is created or its result is cleared, then the sample scheme analyte's Precision Status is set to Not Tested.

When does Precision Assessment occur, and what does it set?

After a result is entered for a sample scheme analyte in the CCWKOP—Open Workbook application, and the system has performed precision assessment to set the sample scheme analyte's Precision Status and contribute to set the sample scheme analyte's Composite QC Status, the outcome of that assessment can be viewed in the lower panel. The outcomes of QC assessments can also be viewed in the CCVALN—Job and Sample Validation application.