Keep not-detected labels distinct from zero

Record what the laboratory actually reported

Keep the source wording visible so the workbook does not turn a reporting limit or qualitative result into a numeric observation it cannot support.

Before paying for technical-document extraction, inspect how the source meaning survives in the output. ParseMill’s public material-register example gives you a synthetic PDF and downloadable evidence to compare. Bring your own document type and acceptance checks to a scope discussion; the self-service invoice plan is a separate offer.

This guide concerns transcription and handoff. The statistical or technical interpretation of non-detects depends on the method, sample and intended use. Route that interpretation to a qualified reviewer instead of embedding an undocumented substitution in the raw register.

Use a result-state worksheet

These invented rows are not automatically interchangeable. The explicit 0.00 remains source text even if it looks related to the non-detect rows. The last row needs the report’s legend before you assign it a broader result state.

EPA: Handling data near detection limits identifies unjustified treatment of non-detects as zero and omission of limit information as problems in its risk-assessment context. The same preservation discipline helps keep an extraction from making that interpretation silently.

Keep the evidence together

1. Capture the result label exactly, including symbols and footnotes. Locate the limit and verify that it belongs to the same sample and analyte.
2. Store limit type, value and unit separately. If the source does not state one of these, mark it unstated rather than borrowing from another row.
3. Leave the measured numeric result unavailable when the source supplies only a non-detect state. Keep any analysis substitution in a labeled derived field.
4. Review the export for blanks or zeros introduced by type conversion. Confirm that the non-detect records remain present in the dataset.

Microsoft: M language values distinguishes null from numeric zero. A null numeric field with an explicit non-detect state preserves more information than a replacement zero.

Avoid overclaiming the result

Do not translate not detected into absent, safe, compliant or approved. Those conclusions require additional context and decision criteria. A missing limit is also a review issue: it prevents a recipient from understanding the sensitivity associated with the reported state.

Limits may vary between samples or analytical runs. Do not propagate one limit down an entire column unless the source explicitly establishes that scope. Keep changes in method or dilution with the affected observations when they are reported.

Verify the handoff

For this fixture, every non-detect label remains visible and its available limit stays attached to the correct row. The row without a limit remains incomplete. A qualified reviewer can then decide the appropriate analytical treatment using the original evidence.

For non-detect reporting, bring the result labels, limits and required states to a document-workflow scope discussion. Inspect the public example first, while treating laboratory interpretation as a separate qualified review. The invoice trial does not supply a non-detect analysis workflow.