Interpreter Reprocessing IFU039(02)

What is Interpreter Reprocessing?

This guide shows you how to create and manage Interpreter Reprocessing runs in the RARE portal. Interpreter Reprocessing lets you re-analyze stored results using new or updated batch (quantitative) or classifier (qualitative) configurations, generating a new set of results while preserving full traceability to the originals.

Interpreter Reprocessing is ideal when you need to:

• Apply new calibration curves or threshold logic to existing data.

• Validate the impact of configuration changes on historical results.

• Support regulatory or QC investigations without altering original records.

Interpreter Reprocessing Management

Accessing Interpreter Reprocessing

• Go to Data Analysis > Interpreter Reprocessing in the RARE web portal.

• You need Analyst or Admin permissions to create or manage reprocessing runs.

• What you see and can do may depend on your role and subscription.

Understanding the Interpreter Reprocessing List

The Interpreter Reprocessing list shows all previous and in-progress reprocessing runs, including:

• Run name and description

• Date created and operator

• Linked dataset

• Batch and classifier configurations used

• Status (e.g., Completed, In Progress, Failed)

Filtering and Sorting Interpreter Reprocessing Runs

Filter

• Filter runs by status, operator, dataset, or date range.

• Use the search bar to find runs by name or description.

Sort

• Sort runs by creation date, name, or status for quick access.

Actions Available

• View details: See configuration, results, and audit trail for any run.

• Export results: Download reprocessed results with full metadata.

Create a New Interpreter Reprocessing Run

Interpreter Reprocessing lets you apply new batch or classifier logic to existing datasets, generating a new set of results for review and comparison. This is especially useful for calibration updates, method validation, or retrospective QC checks—without ever altering your original data.

Step 1 - Describe the Run

Start by giving your reprocessing run a clear, descriptive name and a short explanation of its purpose. This helps with traceability and future audits.

• Enter a clear Run Name (e.g., "QC Panel Reprocess Mar 2026").

• Add a Description to explain the purpose or context.

Step 2 - Select Dataset

Pick the dataset you want to reprocess from the dropdown list. Only datasets that are eligible for interpreter reprocessing will appear.

• Choose the dataset you want to reprocess from the dropdown. If you’re testing a new configuration, consider starting with a small or representative dataset.

• The system will check that all results in your dataset are compatible with the batch/classifier you select in the next step.

Step 3 - Apply Batch (Quantitative) / Apply Classifier (Qualitative)

Select Batch Configuration

If you want to update quantitative values (e.g., concentrations), select a batch configuration. This applies the chosen calibration model to all results in your dataset.

• Pick a batch configuration to apply updated calibration or quantitative models.

• Only valid batch configs for the dataset’s assay type are shown.

• The system uses the selected model (e.g., 4PL, 5PL, Linear) to recalculate quantitative values.

Select Classifier Configuration

If you want to update qualitative interpretations (e.g., Positive/Negative), select a classifier configuration.

• Choose a classifier configuration to apply new qualitative thresholds or logic.

• Only compatible classifiers are listed.

• The system applies the new thresholds or logic to generate updated interpretation labels for each result.

Step 4 - Finalise and Save

• Review all your selections in the summary screen. Double-check the run name, dataset, batch/classifier, and any warnings or compatibility notes.

  • When you’re ready, click Save and Run.

  • The system will create a new set of results, each linked to its original, and log all actions for traceability.

After You’ve Done This

• The system creates a new result for each item in the dataset, inheriting key metadata (sample ID, ground truth, etc.) and linking back to the original result.

• All actions are logged with operator, timestamp, and configuration details for auditability.

• You can view, filter, and export the new results. The original results remain unchanged.

• Any updates to ground truth at the group level are automatically reflected in all linked reprocessed results.