Method Detection Limits: Build a Traceable Sample-to-Report Workflow

A method detection limit is easy to reduce to one number in a worksheet. The real operational question is harder: which analytical process does that number represent, what evidence supports it, and where should it appear when a result is reviewed and reported?

For work governed by 40 CFR Part 136, EPA defines the method detection limit (MDL) as the minimum measured concentration that can be reported with 99% confidence that the measured concentration is distinguishable from method blank results.[1] That definition makes the blank population part of the decision. It also makes an MDL different from a general promise about accuracy, quantitation, or fitness for a particular use.

A dependable information workflow should preserve that distinction from the study record through the final report.

Start with the analytical scope

Before entering an MDL, identify the scope the value is meant to represent. At minimum, the record should make the analyte, method and revision, matrix, preparation path, instrument or instrument group, units, and laboratory location explicit.

That scope prevents a convenient number from being reused more broadly than the supporting work allows. It also gives reviewers a stable key when a method changes, an instrument is added, or a different matrix enters the laboratory.

Decide whether MDLs will be represented per instrument or across a defined instrument group, then document the governing procedure, technical basis, and approval for that choice.

Keep the two estimates visible

Revision 2 evaluates low-level spiked samples and method blanks. EPA calls the spike-based estimate MDL_S and the blank-based estimate MDL_b; the initial MDL is the greater of the two.[1][2]

Those values should not disappear behind a final editable field. A reviewable record should retain:

• the low-level spike results and the calculated MDL_S;
• the method-blank population and the calculated MDL_b;
• the calculation procedure or controlled calculation artifact;
• any excluded data and the documented basis for exclusion;
• the selected MDL and which estimate controlled it; and
• the reviewer, approval date, effective date, and superseded value.

This structure makes a blank-driven increase understandable. It also lets a reviewer distinguish a calculation change from a change in analytical scope.

Design for ongoing data, not an annual scramble

Revision 2 is intended to use performance data distributed through the year rather than a single best-case study date.[1] EPA's procedure calls for at least seven spiked samples and seven method blanks for annual verification, while allowing the spike data to be drawn from the previous two years when one instrument is in use.[2]

The practical systems problem is collection. If low-level spikes, routine blanks, batches, methods, and instruments use consistent identifiers, the annual dataset can be assembled from connected records. If those identities live in separate spreadsheets or free-text notes, the laboratory has to reconstruct the population later.

A useful workflow links each qualifying spike and blank to its batch, analytical method revision, instrument, matrix or reference matrix, run date, result status, and any rejection or reanalysis decision. The annual verification record should then point back to the included source records rather than store only a pasted summary.

Separate MDL, reporting limit, and result interpretation

An MDL answers a procedure-defined detection question. A reporting limit answers a routine reporting question set by the laboratory or applicable program. A quantitation term may have a method- or program-specific definition. Treating all three as one threshold creates avoidable ambiguity.

Store each threshold as a distinct, named record with its basis and scope. Then test how the reporting workflow handles:

• a result below the MDL;
• a result at or above the MDL but below the reporting limit;
• a result affected by dilution or a different sample volume;
• a value converted into another reporting unit;
• a qualified or estimated result; and
• an amended report issued after a threshold changes.

The test should compare the approved source result with the reviewer view, electronic deliverable, and human-readable report. The same scenario should produce consistent units, qualifiers, significant figures, threshold labels, and narratives at every output boundary.

Version the decision, not just the number

Overwriting an MDL removes context from historical work. An effective-dated model is safer: a newly approved value becomes active for defined work without retroactively changing the threshold associated with an already authorized result.

Each version should answer four questions:

1. What scope does this value represent?
2. What study and source records support it?
3. Who approved it, and when did it become effective?
4. Which results and reports used it?

If a correction is necessary, preserve the previous version and record the reason for change. That history is more useful than a generic audit entry that says only that a field was edited.

Run a boundary-focused demonstration

When evaluating a laboratory system, use a sanitized scenario rather than a feature checklist. Include one established MDL, a higher blank-based recalculation, a future effective date, one result between the MDL and reporting limit, and one previously issued report.

Ask the team to show the supporting spike and blank records, approval path, effective-date behavior, reviewer display, final report, amended-report path, and exportable history. The demonstration should reveal whether the system preserves the scientific decision or merely stores the latest number.

A LIMS can preserve these records; the laboratory still determines and approves the applicable scientific and program rules.

Choose the next step around your actual gap

If your decision is whether a proposed workflow can connect MDL evidence, approved thresholds, result review, and report output, evaluate the workflow with Clearline LIMS using a sanitized example of that path.

Carry result qualification into the COA release review

The handoff from threshold approval to a certificate of analysis (COA) should preserve the decision context, not just copy a limit into a report field. Use the laboratory's approved method and program rules to define each expected output.

1. Select a sanitized result between the applicable MDL and reporting limit. Record its method, matrix, units, threshold versions, and any dilution or conversion context.
2. Have the authorized reviewer determine the qualifier and narrative required by the applicable procedure. Do not infer a universal qualifier from the numerical range alone.
3. Compare that decision with the COA and any scoped electronic deliverable: result, units, limit labels, qualifier meaning, and approval state should agree.
4. Exercise a corrected-result or amended-report case and verify that the earlier issued report remains distinguishable from the replacement.

Download the COA release review checklist (CSV) to capture the source record, expected output, observed output, discrepancy owner, and release decision. Continue with the COA reporting workflow guide or the water-testing LIMS workflow for the surrounding sample-to-report context.

This checklist does not calculate or approve an MDL and does not replace the EPA procedure cited below, the applicable program requirements, or the laboratory's scientific review. Report behavior and any integration must be confirmed in the proposed scope.

Sources

1. U.S. EPA — Method Detection Limit: Frequent Questions
2. U.S. EPA — Definition and Procedure for the Determination of the Method Detection Limit, Revision 2