Two words can look like a complete answer. Their meaning depends on the substance, sample preparation and measurement limits attached to them.
A hypothetical report arrives with an apparently reassuring entry: “Impurity A: not detected.” The reader looks for a zero and finds none. Another laboratory reports a small signal for the same impurity. Before deciding that the reports contradict each other, there is a more useful question: what could each procedure reliably detect in the sample it actually measured?
The answer starts with a measurement limit. ICH Q2(R2), the pharmaceutical analytical-validation guidance published by FDA in March 2024, treats detection and quantitation as separate capabilities.[1] A report becomes easier to interpret when those capabilities travel with its result.
A detector has a background
Imagine listening for a quiet sound in a room with a fan running. The fan does not prevent every sound from being heard, but it changes the evidence needed to recognize one. The comparison is an analogy, rather than a model for calculating a laboratory limit.
Analytical procedures also require ways to distinguish a relevant signal from background. Q2(R2) describes several approaches to estimating lower limits, including evaluation of known low-concentration samples, signal-to-noise comparisons and calculations involving response variability and calibration slope.[1] The chosen approach needs to suit the procedure; one numerical shortcut does not describe every measurement.
The idea has a substantial history. In NIST's account of Lloyd Currie's work, Richard Lindstrom explains why recognizing an observed response, estimating a procedure's detection capability and obtaining a quantitative value need separate definitions.[2] Those distinctions help explain why a blank-looking trace cannot support an unlimited claim of absence.
Detection and quantitation answer different questions
The detection limit describes a procedure's ability to detect an analyte. The quantitation limit concerns how little can be measured with suitable accuracy and precision.[1] Recognizing a signal and assigning it a dependable amount therefore require different supporting evidence.
A laboratory may have evidence of an analyte at a level where its quantitative procedure is not sufficiently reliable to issue an ordinary numerical result. A phrase such as “detected, below the quantitation limit” can preserve that distinction, provided the report defines how the finding was established. Reading it as zero would erase information; reading it as a fully supported quantitative value would add information the method has not established.
Detection is also a statistical capability. Lindstrom's NIST explanation distinguishes the rule for recognizing an observed response from the true signal level expected to lead to detection under stated error assumptions.[2] A quoted detection limit should consequently not be treated as an absolute physical boundary below which detection never occurs or above which every observation is guaranteed.
| Term on a report | Question to ask |
|---|---|
| Not detected | Which analyte was sought, and under what detection procedure and stated limit? |
| Detected below quantitation | What supports identification, and why is a numerical amount not being reported? |
| Below a reporting limit | How has this particular reporting threshold been defined? |
| Numerical result | Is it within the procedure's supported range and on the stated sample basis? |
The table is an interpretive guide. It is not a replacement for a laboratory's definitions or reporting rules.
The vial and the measured solution are different reference points
A limit written in micrograms per millilitre needs a location in the preparation record. Does it apply to the original liquid, a diluted aliquot or some other measurement solution? Q2(R2) distinguishes the range presented to the instrument from the reportable range associated with the result.[1] The preparation steps connect them.
Consider this explicitly hypothetical example. A research sample is prepared to a final volume of 20.0 millilitres. An aliquot then undergoes a tenfold dilution for analysis. Assume that all steps are documented, that the analyte transfers quantitatively, and that the stated limits apply to this complete procedure and matrix.
Suppose the detection limit in the final measurement solution is 0.010 micrograms per millilitre and the quantitation limit is 0.030 micrograms per millilitre. The corresponding original-solution concentrations are ten times larger: 0.10 and 0.30 micrograms per millilitre. Multiplying those concentrations by the original 20.0 millilitres gives equivalent amounts of 2.0 and 6.0 micrograms for the prepared sample.
Those figures are arithmetic translations of invented limits. They are not results for a real vial, guarantees of recovery or absolute upper bounds on an undetected amount. If preparation losses or matrix interference were unresolved, the translation would need more evidence before supporting a sample-level conclusion.
The example shows why “0.010” copied from an instrument method can mislead when the report concerns the original vial. The units may be correct while the reference point is wrong. A useful report makes both identifiable.
A reporting threshold needs its own definition
A laboratory may use a defined threshold for how results are presented. The reader needs to know whether that threshold is the detection limit, quantitation limit, a specification-related value or another documented reporting rule. Q2(R2) connects the required reportable range to the intended use of the procedure and relevant specification or reporting limits.[1]
This is a practical reading habit: look for the limit's name, units, sample basis and method. If those details are absent, ask for them. A bare abbreviation such as “ND” supplies too little context to reconstruct the measurement.
It also helps to keep the analytical target visible. A procedure directed at Impurity A does not automatically investigate every other possible contaminant. The statement should remain attached to the analyte actually sought and the method used to seek it.
Return to the two reports
The hypothetical reports can now be compared more carefully. One procedure might support a lower detection limit. The samples might have undergone different dilutions. One report might use “not detected” while another distinguishes detection from quantitation. These are possibilities to investigate, rather than explanations established by the example.
The useful answer is the result together with the measurement conditions that give it meaning. “Not detected” can be an informative finding. Its value grows when the report makes clear what was investigated, how the original sample reached the measurement, and which limits apply.
Research and attribution note
This article explains analytical concepts from ICH Q2(R2) and NIST's historical account of detection terminology. It reports no Chudik experiment or product finding. The dilution example and its limits are invented for explanation. Pharmaceutical-validation guidance is used for its measurement principles; it does not establish a Chudik method, a product specification or suitability for human use.
For related reading, see Reading a batch-linked analytical report and Reading the Peaks.