Laboratory follow-up workflow

The lab workflow converts an evidence gap into controlled work and returns the observed result to the decision system. Each transition produces a distinct artifact so scientific advice cannot quietly become execution authority and a completed assay cannot automatically become decision-grade evidence.

flowchart LR
    A[Program, evidence gaps, and promoted decision] --> B[Advisory assay plan]
    B --> C[Design validation and dependency graph]
    C --> D[Batch, capacity, material, and burden planning]
    D --> E{Operational readiness}
    E -->|blocked| F[Refusal or review queue]
    E -->|ready| G[Executable plan and handoff]
    G --> H[Observed assay outcome]
    H --> I[Acceptance, reliability, and failure triage]
    I --> J{Promotion ready?}
    J -->|yes| K[Knowledge evidence and belief delta]
    J -->|no| L[Rerun, redesign, or hold]
    K --> M[Planned-versus-observed reconciliation]
    L --> M

Plan work that can answer the question

1. Start from a governed program, evidence bundle, and explicit decision or evidence gap. Build an advisory plan and retain the rationale for every recommended assay.
2. Validate contrasts, controls, replication, blocking, pairing, sample identity, and expected power. Add randomization, fractionation, multiplex, QC, plate, carryover, and protocol details where the assay requires them.
3. Build assay dependencies and reject unknown, self-referential, or cyclic edges. Separate gate assays from supporting assays.
4. Rank information gain against cost, material, capacity, turnaround, and operational burden. Deferred assays remain visible in the schedule report.

Authorize an executable handoff

1. Select one batch and declare available sample kinds and operational resources. Run the readiness surfaces; do not treat inventory as review clearance.
2. Confirm required controls, sample lineage, protocol attachments, failure caveats, instrument method, and preflight checks.
3. Build the executable plan, risk assessment, and canonical artifact envelope. Verify the envelope immediately before export.
4. If a target LIMS cannot represent a field, publish the mapping and loss report with the export. Never hide flattened notes, omitted constraints, or changed identifiers.
5. Refuse the handoff when evidence, controls, authority, or feasibility is insufficient. Refusal is an operational safety result.

Capture and return the outcome

1. Record assay result state, QC state, replicate values, dispersion, normalization, censoring, interpretation confidence, protocol deviations, and failure class.
2. Evaluate the declared acceptance rule and distinguish technical, reproducibility, biological, and inconclusive outcomes.
3. Build batch failure triage, reliability assessment, readiness matrix, and rerun plan before evidence promotion.
4. Promote only eligible outcomes under a named policy. Preserve blocked outcomes and the reason they were excluded.
5. Reconcile requested versus observed value and send evidence records, claim belief deltas, missing requested assays, and operational lessons back to knowledge and intelligence without rewriting the original plan.