as-is vs anhydrous basis: reporting and conversions

As-is vs anhydrous basis distinguishes reporting the same analyte quantity per mass of the sample in its stated condition from reporting it per that sample mass after subtracting the water quantity defined by the measurement method.

In analytical reporting practice, the measurand is the quantity intended to be measured. Define it through the analyte identity, the quantity being measured, the sample condition, the reporting unit and the denominator exclusions. Here, the principal model is mass fraction: analyte mass divided by the specified reference mass. A basis label alone does not fully define that measurement.

• As-is denominator: the sample mass in its stated condition, including water and all other constituents present. State whether that condition is as received or after a specified preparation; as-is does not identify the handling history by itself.
• Anhydrous denominator: that same sample mass minus the measured water mass attributable to it. Nonwater volatiles remain included unless the reporting convention expressly excludes them. Identify any such additional exclusion rather than leaving it implicit.

An anhydrous reporting basis does not establish that the physical material is water-free. The calculation changes the denominator, not the sample. Likewise, a dried basis need not equal a water-free basis: the drying procedure defines what is removed, which may not be water alone. Keep that distinction visible when considering water and loss-on-drying compensation.

The unit g/100 g means grams of analyte per hundred grams of the stated denominator; mg/g means milligrams of analyte per gram of that denominator. On an as-is basis, the denominator includes water; on an anhydrous basis, it excludes the method-defined water quantity. A percentage mass fraction expresses analyte mass per hundred mass units, not concentration per unit volume. Units such as mg/mL therefore describe a different quantity.

Keep analyte identity unchanged throughout. Chemical-form equivalence, salt-to-base conversion, dry-weight potency selection and manufacturing charge calculation are separate operations. This section addresses analytical reporting, not the subsequent use of a result in a potency calculation.

In common analytical reporting practice, these labels answer different questions. For mass-based results expressed as mg analyte/g material, the denominator is the material mass defined below.

As-received and as-weighed describe sample conditions; their relationship to an as-is result depends on when and how the sample was measured. Preserve that context at sample handoff.

Why a dried denominator need not be water-free is set out under Definitions; loss on drying versus Karl Fischer compares the methods.

Two results can differ solely because their denominators differ. Compare result and specification only after aligning analyte, unit, sample condition and reporting basis.

As a reporting practice, distinguish the basis declared for a source result from the target basis requested for comparison. Read these records together rather than treating a supplier CoA as the complete reporting instruction.

In a LIMS or worksheet, use separate fields for original value, original unit, original basis, target basis and derived value, with its unit. Preserve the supplier result. Record the analyte, sample and aliquot identifiers, sampling condition, measurement time, water or drying method, method version and source-document revision.

Give basis codes explicit definitions in controlled material master data: AS_IS uses original sample mass without exclusions; ANHYDROUS excludes water, with water fraction expressed as water mass divided by original sample mass. DRIED uses the method-defined residue mass; document the excluded components and drying-loss fraction as lost mass divided by pre-drying sample mass. A free-text dry label leaves these unresolved.

Link assay and water observations only where sample conditions are compatible. A shared lot number does not establish the same moisture state: storage, exposure or conditioning between aliquot measurements can change water content. Record the handling relationship, not just the lot reference.

Proposed record diagram: specification requirements + method definition → linked measured observations → derived calculation → human review → final report. Keep each record type distinguishable, with references connecting them; a calculated value is not a new observation.

For US finished pharmaceuticals, 21 CFR 211.194(a)(5) requires records of test calculations, including units, conversion factors and equivalency factors. The field structure above is a practical implementation, not a prescribed database layout.

The following mass-balance model is an analytical reporting convention, not a universal regulatory rule. It changes the reporting denominator while holding the analyte quantity fixed; it does not calculate a manufacturing charge.

Define the masses before converting

• m = original as-is sample mass, in grams.
• a = unchanged analyte mass within that sample, in grams.
• m_w = measured water mass attributable to that sample, in grams.
• w = m_w/m = water mass fraction referenced to the original as-is sample; its units are g/g, so it is dimensionless.

Assume that analyte quantity and identity remain unchanged, the water result applies to the assay sample condition, and both results describe the same underlying material. A water measurement from a separate aliquot is usable only insofar as it represents that condition.

Forward conversion and its inverse

For numerical results reported in g analyte/100 g of the stated basis, define P_as = 100a/m and P_anh = 100a/(m − m_w). The first denominator is as-is sample mass; the second is that mass minus measured water. Since m − m_w = m(1 − w), substitution gives:

• P_anh = P_as/(1 − w).
• P_as = P_anh × (1 − w).

If W is the numerical water percentage in g water/100 g as-is sample, enter w = W/100, not W. Keep the unit labels attached: P_as/(1 − w) retains the g/100 g reporting scale but now refers to water-excluded mass. An assay percentage and a water percentage describe different quantities despite sharing a percent sign. The distinction also matters in water and LOD compensation.

Moving between excluded-mass bases

Let d be the method-defined drying-loss mass divided by original as-is mass. Then P_dry = P_as/(1 − d), but only when the unchanged-analyte mass-balance model is appropriate. Apply the dried-basis distinction covered above rather than substituting drying loss for water; drying mass balance provides related context.

More generally, let f_source and f_target be the fractions excluded from the source and target denominators, respectively, both referenced to the same original mass m. For results on the same g/100 g scale, P_source = 100a/[m(1 − f_source)], so P_target = P_source × (1 − f_source)/(1 − f_target). Consequently, P_anh = P_dry × (1 − d)/(1 − w).

Do not use this simple model with incompatible reference masses, analyte loss, chemical change, unknown exclusion definitions or a nonpositive denominator. Resolve those conditions through the applicable analytical method before converting.

These cases are hypothetical reporting practice, not product specifications. All stated masses are exact example assumptions for demonstrating arithmetic—not experimental precision, acceptance limits or typical product compositions. The stated analyte remains the same throughout each case. In the tables, excluded fractions are grams excluded per gram of original as-is material; denominators are the corresponding reference masses.

Case A — Supplement ingredient: hypothetical practice

Assume a 100 g as-is sample contains 72 g of the stated analyte, 4 g water and 24 g other nonvolatile constituents. Excluding only water leaves a 96 g water-free denominator. The analyte mass remains 72 g: the calculation changes its reporting denominator, not the amount present.

Equivalently, 72 g/100 g as-is is 720 mg/g as-is, and 75 g/100 g anhydrous is 750 mg/g anhydrous. Reversing the conversion gives 75 × 0.96 = 72 g/100 g as-is. Reporting decision: select the requested basis and retain its label beside the value; neither value represents additional analyte.

Case B — Chemical intermediate: hypothetical practice

Assume a 100 g sample contains 76 g analyte, 5 g water, 15 g nonwater solvent and 4 g other solids. Assume the specified drying procedure removes all water and solvent without changing or losing analyte. The anhydrous denominator is 95 g, retaining the nonwater solvent; the dried denominator is 80 g, excluding both water and solvent.

The dried-to-anhydrous route confirms the result: 95 × 0.80 ÷ 0.95 = 80 g/100 g anhydrous. Multiplication first restores the as-is value; division then excludes water alone. This is the practical distinction behind loss on drying versus Karl Fischer. Reporting decision: use 80 g/100 g for anhydrous reporting, or 95 g/100 g for the assumed drying procedure—not interchangeable labels.

Case C — Pharmaceutical ingredient CoA: hypothetical practice

Assume the certificate of analysis reports 98 g/100 g anhydrous and a compatible water result of 2 g/100 g as-is. For a 100 g as-is reference mass, the water-free denominator is 98 g. The source assay already uses that water-free basis.

Do not divide the source assay by 0.98 again: that would double-correct an already-anhydrous result. Water / LOD compensation therefore starts with identifying the source basis. Reporting decision: retain 98 g/100 g for an anhydrous target; report 96.04 g/100 g only when the target is as-is.

NIST TN 1297 Appendix A.1 recommends that a measurement model include all quantities contributing significant uncertainty to the result. This is general metrology guidance, not a product-specific manufacturing rule; it does not establish an uncertainty allowance for an as-is-to-anhydrous conversion.

As reporting practice, apply that principle to the assay input and water input, then assess whether their sampling conditions represent a compatible material state. Consider relevant dependence between measurements rather than automatically treating them as independent. This extends the assay uncertainty assessment to the converted result, not merely the original assay.

For the conversion model above, a smaller retained mass fraction makes the result more sensitive to denominator changes. For the same absolute uncertainty in water mass fraction, its relative influence on the converted result increases as the retained fraction shrinks. Evaluate that influence for the actual measurements rather than applying a universal allowance.

Preserve available input precision through calculation, then apply the established reporting convention at the output. Additional displayed digits do not establish additional measurement certainty. The hypothetical cases demonstrate exact arithmetic under stated assumptions; real assay and water measurements remain uncertain. If supplier inputs are available only as rounded values, evaluate that limitation before reporting the conversion. Calculation cannot reconstruct discarded digits or reveal the underlying measurement uncertainty.

A reporting basis connects a result to its calculation and comparison standard. Keep the record obligations for US finished pharmaceuticals under 21 CFR 211 separate from measurement guidance and the reporting practices discussed here.

• For US finished pharmaceuticals under 21 CFR 211, 211.194(a)(5) identifies all test calculations, including units of measure, conversion factors and equivalency factors, as laboratory-record content.
• For US finished pharmaceuticals under 21 CFR 211, 211.194(a)(6) identifies test results and their comparison with established standards of identity, strength, quality and purity as laboratory-record content.
• For US finished pharmaceuticals under 21 CFR 211, 211.160(a) requires the subpart’s requirements to be documented at the time of performance.
• For US finished pharmaceuticals under 21 CFR 211, that same provision requires deviations from written specifications, standards, sampling plans, test procedures or other laboratory-control mechanisms to be recorded and justified.

For food, supplements, devices, chemicals and other jurisdictions, consult the applicable product requirements, current method and specification. Do not use the US finished-pharmaceutical provisions as substitute requirements for those categories.

Keep compendial applicability material-specific: consult the current applicable method and specification for the basis declaration. Treat the conversion model discussed earlier as reporting practice, not a universal USP convention or a legal conversion formula. The review question is whether the declared basis matches the method and comparison standard—not whether one basis sounds more authoritative.

Use this recommended reporting workflow as an operational control, not as a substitute for the applicable laboratory procedure.

1. Identify the sample. At sample login and handoff, connect the sample identifier, lot, source and sampling condition to the analytical record.
2. Capture the source result. The analyst records the value, unit, declared basis and method version before selecting any conversion inputs. Preserve the original CoA or laboratory result.
3. Select the comparison target. Identify the applicable specification version and its reporting basis. Review whether the assay and supporting water or volatile measurement represent compatible sample conditions and the intended denominator.
4. Choose the route. Retain an already compatible result unchanged; calculate a justified conversion when the inputs support it; or pause for clarification when the basis or supporting measurement remains unresolved. Record why the selected route applies.
5. Reconstruct and review. The analyst documents input selection, units, equation and rounding. The reviewer confirms the comparison basis, checks the calculation against its inputs and assesses the documented reasoning. A recalculation review should remain traceable to the original result.
6. Authorize the report. Retain the source result beside any derived result, with applicable method and specification versions. Use human review to authorize reporting; keep that authorization separate from material disposition. Neither a completed calculation nor a software action constitutes material release.

Lifecycle summary: Identified sample → captured result and selected specification → compatibility decision → retain / convert / pause → review → report authorization.

As an integration practice, treat LIMS, ERP, supplier portals and spreadsheets as generic system categories, not guarantees of reporting meaning. Exchange the value, unit, analyte, basis code, method version and source-record identifier together. Keep original observations separate from derived values, with links between them; a converted result should not overwrite its input.

An explicit as-is declaration is not equivalent to a blank basis field. Route missing, unknown or unmapped codes for clarification rather than supplying a default. Maintain controlled mappings alongside material master data, preserving the supplier’s original declaration so a mapping can be checked later.

Assign each conversion a deterministic identity using the source measurements and their immutable revisions, target basis and formula version. Include the assay and water measurement references, not just the lot identifier. On an interface retry, use that identity to retrieve the existing calculation rather than transform the result again.

For an amended CoA or revised water measurement, retain the earlier source and calculation, identify affected reports, and document what changed and why. Obtain review of the replacement result through a controlled recalculation review workflow, linking it to the superseded result. Recovery should restore traceability and comparison on the intended reporting basis—not merely make displayed numbers match.

These practice controls apply the definitions and Switching bases model above. For transferred or corrected records, also use the integration controls described earlier.

For a result outside the physically meaningful range of its stated measurand, investigate identity, method, units, denominator and input compatibility. Do not clamp the value or treat it as an automatic disposition decision.

Use this practical checklist at report review; refer to the definitions, conversion model and reporting lifecycle above for explanations.

• Identify the analyte, source and target mass denominators, and units for every input and output.
• Confirm compatible sample conditions and applicable water or drying measurements; revisit the distinction in water/LOD compensation.
• Record specification and method versions, source-document revision, and the unchanged source result with its basis.
• Preserve the traceable formula, input references, rounding convention, reviewed output and specification comparison.
• Link original and corrected reports, the correction reason and reviewer authorization through the audit trail.

For implementation testing, replay the hypothetical worked cases above. Test a forward-and-reverse calculation using unrounded intermediate values; confirm that an already compatible source result remains unchanged. Test an unknown basis code and an incompatible measurement pairing: both should stop conversion pending resolution, rather than invoke a default.

Final review question: can another qualified person reconstruct the denominator, inputs, calculation and comparison from the record?