(tutorials:reporter-review-and-curation)=
# Review and curate results

The **REVIEW** stage is where the MedicalTechnologist inspects the analysis outputs, selects reportable findings, and prepares the case for reporting.

## Before you start

- Confirm that you have the **MedicalTechnologist** role. BioinformaticsScientist creates interpretations during ANALYSIS; MedicalTechnologist curates findings in REVIEW.
- The sample must be in the **REVIEW** state with completed or rerun results. If the sample is still in **ANALYSIS**, the BioinformaticsScientist must first complete the pipeline run.
- Understand the assay-specific reporting criteria used by your laboratory.

## Open a sample for review

1. Go to **Reporter → Management**.
2. In the **REVIEW** lane, find the sample you need to curate.
3. Open the review dialog for the sample.

## What to review

The review dialog ("View Results") opens a tabbed, assay-specific view. Typical tabs include:

| Tab | Purpose |
| --- | --- |
| **Variants** | The variant list, with per-variant classification chips (ACMG Pathogenic/Likely Pathogenic/etc. and, for oncology assays, VICC Oncogenicity), ClinVar links, sequencing depth, and variant allele fraction (VAF). Quick filters let you jump to truncating variants or to Pathogenic/Likely Pathogenic calls. |
| **Quality Control** | Run-level QC metrics such as mean depth of coverage and percentage of the targeted region covered above set thresholds (for example, coverage \> 20x and \> 60x). |
| **Files** | Files linked to the run. |
| **Panel** | Gene panels or regions relevant to the assay, where configured. |

"Interpretation" in SeqsLab is a record produced by the analysis pipeline and refined here; it carries the variant-level classification shown in the Variants tab plus a written **summary** field. The summary is not optional decoration: Reporter will not let you approve a sample into a report until its interpretation has a summary of meaningful length, so make sure you write a real clinical summary rather than leaving it blank or trivially short.

## Curate findings

1. Inspect each variant or result in the **Variants** tab, using the classification and ClinVar/VAF/depth columns to judge significance.
2. Check the **Quality Control** tab for coverage and depth issues that could affect confidence in a call.
3. Edit or confirm the interpretation summary text for the case.
4. Click **Preview** to generate a preview of how the curated data will render in the report, and review it before approving.
5. When curation is complete, click **Approve** to open the Approve Results dialog, confirm or edit the auto-generated Report ID, and submit. Reporter blocks approval if any selected sample has not been previewed or lacks a valid interpretation summary.

## Rollback to ANALYSIS

Send the sample back to **ANALYSIS** when:

- The data quality is insufficient.
- A variant or region must be recalled with different parameters.
- The wrong workflow or reference was used.
- Additional sequencing data is required.

In the review dialog, use **Revert** to send the sample back to ANALYSIS. Reporter warns that reverting discards all review information and updates made since the sample entered REVIEW, and this action is permanent. Confirm the target state before proceeding, and always document the reason for the rollback so the BioinformaticsScientist or MedicalTechnologist can take corrective action.

## Escalation

If the case requires input from a more senior reviewer or a different specialty, use your organization's escalation process. Do not advance a sample to **REPORT** until the required reviews are complete.

## Audit expectations

Each curation decision should be traceable to:

- The reviewer who made it.
- The timestamp.
- The evidence or source used.
- Any comments or reasoning recorded.

## Roles and handoffs

| Action | Typical role | Result |
| --- | --- | --- |
| Open and review results | MedicalTechnologist | Findings are inspected. |
| Add interpretations and comments | MedicalTechnologist | Curated record is created. |
| Mark as ready for report | MedicalTechnologist | Sample is prepared for LaboratorySupervisor. |
| Return to analysis | MedicalTechnologist | Sample returns to **ANALYSIS**. |

## Troubleshooting

| Problem | Possible cause | Action |
| --- | --- | --- |
| Results are missing | Run failed or outputs were not produced. | Return the sample to **ANALYSIS** and rerun. |
| Cannot save interpretation | Required fields missing or permissions issue. | Check mandatory fields and verify your role. |
| Expected variant not shown | Filter settings are too restrictive or pipeline settings differ. | Adjust filters or consult the bioinformatics team. |

## Related material

- {ref}`tutorials:reporter-sample-lifecycle` for state transitions.
- {ref}`tutorials:reporter-reporting` for the next stage.
- {ref}`security:healthcare-rbac` for role permissions.