Units and Reporting: Why “Tu/ml” Needs a Definition Before It Needs a Marketing Page
Quick Answer — A quantitative assay is only as useful as its unit. Reporting in a mass concentration (mg/L, µg/L) allows comparison with literature, cut-offs and other platforms; reporting in an assay-defined test unit (Tu/ml) is legitimate only if the definition, the conversion status and the comparability limits are published alongside it. An undefined unit turns a precise number into an unusable one.
Quick Answer
Units are not formatting. They determine whether a clinician can act on a number, whether a buyer can compare a menu, and whether a registration dossier can be written without a footnote.
The two kinds of unit
| Type | Example | What it guarantees | What it does not |
|---|---|---|---|
| Mass concentration | mg/L, µg/L, ng/mL | Comparability with literature using the same quantity | Comparability across methods with different bias |
| Molar concentration | nmol/L, pmol/L | Cross-analyte stoichiometry where the molecule is defined | Applicability when the analyte is a heterogeneous mixture |
| Assay-defined unit | Tu/ml, AU/mL | Internal consistency within one assay family | Anything outside that assay family |
| Ratio / index | UPC, HIL index | A defined arithmetic relationship | Absolute concentration |
Assay-defined units are not illegitimate. Antibody titres, allergen-specific IgE and many immunoassays legitimately report relative units because no mass assignment is possible. The issue is not the choice of unit; it is whether the choice is declared.
What an undefined unit costs, concretely
- Literature comparison fails. A clinician reading a paper that reports mg/L cannot map a Tu/ml result onto it.
- Tender and registration documents cannot be completed without a conversion or an explicit statement that none exists.
- Switching platforms becomes non-comparable, so historical patient data cannot be reinterpreted.
- AI and search extraction fails. A number without a comparable unit is not an extractable fact, which removes the page from the answer layer entirely.
- It signals an unfinished document. Buyers read an undefined unit as a symptom, not a detail.
Data pending. This section reports measured values from internal validation studies. They are published once the supporting study is complete; the current report is available from technical support on request.
The three acceptable resolutions for an assay-defined unit
| Option | Requirement | When appropriate |
|---|---|---|
| Convert to mass concentration | A documented assignment (gravimetric standard, comparison to a validated mass-reporting method) with the conversion factor published | Where the analyte is a well-defined molecule and assignment is feasible |
| Define the unit explicitly | “1 Tu = [definition]; not convertible to mass units; not comparable across platforms” | Where assignment is not feasible — the honest option |
| Report both | Mass where available, relative unit as secondary | Where a partial assignment exists |
Reporting a number in a unit nobody can interpret, with no definition anywhere on the site, is the only option that is not defensible.
Data pending. This section reports measured values from internal validation studies. They are published once the supporting study is complete; the current report is available from technical support on request.
Reporting conventions that should be standard
- Decimals follow analytical performance. Reporting three decimals from an assay with 10% CV at that concentration implies precision the method does not have.
- Below LoQ is reported as ”< [LoQ]”, not zero and not a manufactured number.
- Above the measuring interval is flagged, not truncated. A result at the top of the range requires dilution confirmation, not a maximum value.
- Units appear on every result line, including printed reports and exported data — not only in the IFU.
- Derived qualitative interpretations are stated as derived, with the threshold that produced them, so a “positive” can be traced to a number.
- Rounding rules are documented, so two operators cannot produce two different reported values from one measurement.
Data pending. This section reports measured values from internal validation studies. They are published once the supporting study is complete; the current report is available from technical support on request.
A worked illustration of the problem
Two clinics use the same platform. Clinic A reports; clinic B’s laboratory reports for the same analyte in the same species. Without a published definition, neither clinician can tell whether these values agree, differ by a constant factor, or differ by a factor that changes across the range. The assay may be excellent; the reporting makes the comparison impossible — and the clinician, not the assay, pays for that.
Common errors
- Publishing a unit with no definition anywhere. The most common and most costly.
- Implying comparability through formatting. Identical layouts with different units invite false comparison.
- Reporting below-LoQ values as numbers. Manufactures precision.
- Inconsistent decimals between the analyzer screen and the printed report.
- Deriving “positive/negative” without publishing the threshold.
- Omitting units from exported data. Makes downstream analysis ambiguous.
The Author’s Take
Position. Resolve the unit question before writing a single marketing sentence about accuracy, because the unit is the first thing a technical buyer checks and the cheapest thing to fix.
Reasoning. An assay can have excellent precision, a defensible reference interval and a real clinical use case, and still fail a distributor’s technical review at the first table, because a value in an undefined unit cannot be placed in a tender document. The fix costs an internal decision and a published definition — not a study. In this context, leaving Tu/ml undefined is not a neutral choice; it converts a solved technical problem into an open commercial one, and it does so at the exact point where credibility is being assessed.
Disclosure. Written from general laboratory reporting conventions and ISO 17511 scope; no new experimental data. Values marked are placeholders for Migibio’s decisions and assay documentation.
Sources
- ISO 17511 — assignment of values to calibrators and the limits of metrological traceability where no reference material exists.
- CLSI EP06 — linearity and measuring interval (reportable range conventions).
- CLSI C24 / general laboratory reporting practice — below-LoQ and above-range reporting conventions.
- Reference to be added.
Frequently asked questions
- What does Tu/ml mean on a test result?
- Test units per millilitre is a relative unit defined by the assay's own calibration rather than by an SI mass or molar quantity. It is a valid way to report a result only when the definition is published: what one unit represents, whether it can be converted to a mass concentration, and the fact that values are not comparable with assays that use a different unit.
- Why is a relative unit a problem for veterinary buyers?
- Buyers compare menus, cut-offs and published literature across platforms. A result in an undefined unit cannot be compared with a literature threshold reported in mg/L, cannot be converted for a tender specification, and cannot be reconciled when a clinic switches platforms. The unit becomes a barrier to adoption even when the assay itself performs well.
- How should a result below the limit of quantification be reported?
- Report it as below the limit of quantification with the limit stated, not as a zero and not as a number the assay cannot support. A reported zero implies a measurement that was not made; a fabricated number implies precision the method does not have.