Validating Results in a Workbook Session

Overview

The collection of sample results and the validation of those results across the whole sample is a crucial process of the laboratory. Reporting results to the client and then subsequently informing them that the results are checked and there were mistakes creates a huge 'distrust' in the perception of the client. Systems and mechanisms need to be in place to minimise the risk of reporting an inaccurate or incorrect result. Some of these 'systems' include:

  • Testing QC samples distributed among the client samples.
  • Splitting samples for repeat testing at a number os stages in the analysis process.
  • Repeating some samples at random.
  • Performing some tests on complimentary schemes that cross-check each other.
  • Having levels of authority (and skill) to review the results before they are validated.
  • Having levels of validation (sample scheme analyte result, sample and job level) that different users can be authorised.

A sample may need to be analysed by multiple tests across many analytical disciplines. Each analytical discipline is usually aligned with a laboratory section, for example, wet chemistry, x-ray, fire assay. The section head and the operators of the various instruments in that section are responsible for the results of all tests performed in their section.

For the sake of maximum confidence in accuracy and correct reporting, it is advisable that the supervisor has designed and instigated cascading review and access authority levels. This ensures that review and checking, as well as being mandatory, can also be restricted to those with the necessary skills and experience before validation.

One of the actions for a QC failure is to acknowledge that it has failed the checking processes, but the reason for its failure is known (or thought to be known). In this case, the QC failure is ignored, in which case, a reason for the action might be expected, but then the results can be validated as the QC failure is effectively removed (but without removing the sample or the result).

After results are entered, they need to be released and validated for the analysis considered complete. Quality control assessments are carried out on the unknown and QC samples, because if their results are within expected limits, the operator can be assured that the instrument was calibrated correctly, the sample was prepared correctly during the sample preparation stage, and the aliquot was prepared correctly during the analytical preparation stage. Release and validation requirements for results of sample analysis is dependent upon the testing that is conducted or the type of laboratory.

Results of sample analysis need to be validated before the result can be included in a final certificate report. Validating results means that the supervisor or section head is happy to release the results from their laboratory section. A result cannot be validated unless it has first been released.

This process is usually carried out by the operator, that is, the user who captured the results from the instrument or entered the results into the workbook.

Some workbook operators have enough privilege to validate results. If this is the case, result validation can be performed by the operator in the CCWKOP—Open Workbook application following result entry. Otherwise, result validation is performed by the supervisor or section head using the CCVALN—Job and Sample Validation application, even if they do not have enough privilege or visibility to validate the entire sample.

A second-level checking can be performed (by a different person typically with a higher authority), which involves various checking and eye-balling tasks. They validate the results when they are happy with them. If there are QC failures, then the results cannot be validated and need to be checked, repeated, or the QC failure ignored with good reason, before they can be validated.

Process

Results can be validated only where the sample scheme analyte's Workflow Status is Released and Composite QC Status is Passed.

Validate results in workbook cells

Some analytes in some tests are presumed to be alright without validation, so do not need to undergo a validation process. In this cases, the analytes are configured for automatic validation.

Quality control samples are placed in the workbook so that the system performs various QC assessments on their results, the outcome of which indicates to the quality assurance manager that all the results in the workbook are good.

Where a sample scheme analyte has a composite QC failure, then the result cannot be validated until the failure is resolved. Validation indicates that no QC failures exist for the result.

One of the actions for a QC failure is to acknowledge that it has failed the checking processes, but the reason for its failure is known (or thought to be known). In this case the QC failure is ‘ignored’, in which case a reason for the action might be expected, but then the results can be validated, as the QC failure is effectively removed (but without removing the sample or the result).

If this second-level validation mechanism is not required, then analytes can be flagged to be validated automatically.

User Group Limitations on Validating Results

To limit what tests an operator has access to in the CCWKOP—Open Workbook application or in the CCVALN—Job and Sample Validation application, a scheme can be linked to a section, and the section can be linked to an access user group, release user group and validate user group. Refer to Limiting User Visibility of Schemes and Functionality by Section.

A user can validate a result manually where the scheme is not linked to a section, or where either:

  • The scheme is not linked to a section.
  • The scheme is linked to a section and the section is not linked to a user group for result validation.
  • The scheme is linked to a section and the section is linked to a user group for result validation and the user is a member of the user group.

Validating Results in Laboratory Batch Jobs

The cascade engine is responsible for cascading workflow status up job trees, however, the cascade varies when a result is captured in context of a laboratory batch job.

Typically, samples are registered, and batched if required, then results are captured, released and possibly validated in the CCWKOP—Open Workbook application.

When a sample scheme analyte's result is captured in a laboratory batch job, that result is subject to the QC assessments of that laboratory batch job. Therefore the result is not considered to be 'good' until the entire laboratory batch job is validated.

Where a production sample scheme has some or all of its sample scheme analytes included in one or more laboratory batch jobs, until each related laboratory batch job is validated, the highest Workflow Status that the production sample scheme can attain is Released. Where each sample scheme analyte in a production sample scheme is validated, it is only when all of the related laboratory batch jobs are validated that the production sample scheme's Workflow Status is set to Validated by the cascade engine.

The configuration and performance of result release and validation tasks must be completed in a specific order to ensure all settings and processing of results occurs effectively. These processes are defined in this section, and consist of:

Configuring Result Validation Options

Configuring result validation:

  • Configure analytes (that is, a result for that analyte) to be validated automatically or manually.
  • Configure a section for a scheme.
  • Configure an access user group for a section so users can open the scheme in workbook.
  • Configure a validate results user group for a section so users can validate results.

This configuration can be performed in the CCSCHM—Scheme and CCSECT—Section applications. Refer to Maintaining Schemes and Maintaining Sections.

QC History Observation Update

Any update to a QC sample scheme analyte's Workflow Status is persisted to the associated QC history observation.