Assay-Adjusted Charge

Assay-Adjusted Charge is the recalculation of an active ingredient's weigh-out target at the moment of dispense so the batch delivers the label-claim quantity of active despite the actual lot assaying above or below the formula's reference potency. The mechanic is one line of arithmetic: adjusted target = nominal target × (reference potency ÷ measured potency), evaluated on a basis that matches the way the formula was written and the way the certificate of analysis reports the result.

The problem the adjustment solves is structural. Master Manufacturing Records are written against a reference potency — typically 100.0% on the active's pharmacopeial monograph (USP, EP, JP) basis. Real lots vary; supplier CoAs commonly report 95–105% assay against that reference. If the operator weighs the nominal mass without correction, the batch silently under-delivers (if the lot is under-potent) or over-delivers (if it is over-potent) the active ingredient. For a narrow-therapeutic-index drug or a finely-titrated supplement, that error lands in the finished product, escapes through release testing if release relies on the same assay method, and reaches the patient. The recalculation is the controlling step.

Before the recalculation runs, the inputs must be on the same basis as the formula. Mismatching bases is the most common quiet failure mode — the math is correct, the basis is wrong, the batch under- or over-delivers by 1–10%, and the gap is invisible until finished-product assay or a regulator's mass-balance challenge surfaces it.

1. Anhydrous vs as-is — most APIs and many supplements report assay on the anhydrous basis. If the formula was written 'as-is' but the CoA reports 'on anhydrous basis', the recalculation must also adjust for measured moisture (Loss-on-Drying or Karl Fischer). See Water/LOD Compensation. Mixing the two bases produces a 0.5–8% silent error depending on hygroscopicity.
2. Free-base vs salt form — when the active is dosed as a salt (HCl, sulfate, mesylate, sodium, magnesium…) but label claim is the free base, a Salt-to-Base Conversion Factor must be applied first. Confusing the two is the second-most-common assay-related 483 observation in solid-dose pharma.
3. Marker compound vs total botanical — for herbal / botanical supplements, the marker compound assay (e.g. ginsenosides for Panax ginseng, withanolides for Ashwagandha) is the only quantitative handle the supplier reports. The MMR must declare the marker, the % range expected, and the basis; deviating produces silent sub-potency under 21 CFR 111.70.
4. Activity vs mass — for enzymes, probiotics (CFU), and biologics, the assay is in activity units (U/g, IU/mg, CFU/g) not mass-fraction. The recalculation flips: adjusted mass = (target activity per batch) ÷ (measured activity per gram). Treating an activity-based active like a mass-based one is the single most common error in enzyme-blended supplements.

The general form is adjusted target = nominal target × PF × (1 / (1 − LOD)) × SBF, where PF is the assay potency factor (reference ÷ measured), LOD is the loss-on-drying fraction (0.025 = 2.5% moisture), and SBF is the salt-to-base conversion factor (>1 when the API is the salt form and the label claim is the free base). When a factor is not applicable it is set to 1.0.

Each adjusted target is the figure the operator weighs to; the nominal target is shown for context only. The audit trail captures both, the factors applied, the lot's CoA reference, and the operator + verifier signatures.

When more than one correction applies — common for APIs sold as a salt on an anhydrous basis with batch-variable assay — the order matters. Applying the factors in the wrong order can produce a 0.5–3% error compounded from rounding at each step. The canonical sequence:

1. Salt-to-base normalise — if the label claim is the free base but the API is supplied as a salt, multiply by MW(salt) / MW(base) to convert the target mass from base-mass to salt-mass.
2. Anhydrous-basis convert — divide by (1 − LOD) to convert from anhydrous mass to as-is mass. This step is skipped if the formula and assay are both already on the as-is basis.
3. Apply assay PF — multiply by reference potency ÷ measured potency to correct for batch-to-batch potency variability.
4. Recompute counter-balance excipient — subtract the total adjusted active mass from the planned batch size, redistribute the delta into the designated counter-balance excipient (typically lactose, MCC, maltodextrin, dicalcium phosphate). Total batch mass holds; active delivery is exact.

Adjusting the active up or down changes total batch mass unless something else is adjusted in the opposite direction. The counter-balance excipient is the designated diluent that absorbs the delta: if the active is recalculated from 50.000 kg up to 51.440 kg, the counter-balance is recalculated 1.440 kg down. Total batch mass stays at the planned value, total batch volume stays close to planned (within excipient/active density differences), the mixing vessel capacity check (see Mixing Vessel) is preserved, and downstream fill weights remain correct.

The MMR must designate the counter-balance partner explicitly. Picking the wrong partner — e.g. a binder rather than a true diluent — changes dissolution profile or compactability. For supplements the counter-balance is usually a low-functionality bulk excipient (microcrystalline cellulose, maltodextrin, dicalcium phosphate). For solid-dose pharma, the diluent specified in the formulation development report (FDR) is the only acceptable choice; substituting requires change control.

Every assay-adjusted weigh-out is a Part 11 e-signature event under 21 CFR 211.101(d). The operator who performs the dispense signs the adjusted-charge record; a second individual (or automated equipment under §211.68) verifies. The verifier signature must occur before the next operation in the batch sequence; back-filling at end-of-shift breaks the contemporaneity rule (MHRA GMP Data Integrity Guidance 2018) and is one of the most common 483 observations in dispensing audits.

Annex 16 (the QP release contract) requires the QP to take personal responsibility for the batch — which in practice means the QP cannot release a batch where the dispense audit trail is incomplete or where assay adjustments were applied without verifier signature. The signature chain has to land in the BMR before release; missing signatures are the second-most-common QP-rejection cause after open deviations.

1. Basis mismatch — formula written 'as-is', CoA reported 'anhydrous', no LOD compensation applied. Silent 2–8% under-delivery.
2. Salt-form confusion — label claim is the free base, formula and dispense weigh the salt without applying SBF. Silent 5–20% under-delivery.
3. Marker-compound drift — botanical adjusted to nominal marker without re-baselining counter-balance, batch mass drifts up over campaign.
4. Counter-balance is a critical excipient — adjusting active up forces a critical disintegrant or binder down, dissolution moves out of spec.
5. Verifier signature back-filled at shift end — breaks contemporaneity, deviation per MHRA 2018; QP can refuse release.
6. Recalculation done on a paper worksheet — operator transcribes back into kiosk, transcription error escapes verifier. Common with hybrid paper/electronic dispense.
7. Assay-as-is used after retest — original CoA assay used long after stability or re-assay shifted PF; lot 'aged' against the original certificate.
8. Rounding cumulated across factors — each factor rounded to 3 sig figs before the next applies; final adjusted target off by 0.5–1.5% from true.

• Adjusted-dispense coverage % — fraction of active weigh-outs where the recalculation engine ran (target 100% for assayed actives).
• Verifier-signature contemporaneity % — fraction of dispense events where the verifier signature timestamp falls within X minutes of the operator signature (target ≥99%).
• Out-of-tolerance adjustment rate — fraction of recalculations that landed outside the formula's allowed adjustment window (target ≤2%, investigate trend).
• PF override rate — fraction of dispense events where an operator or supervisor manually overrode the computed PF (target ≤1%; every override is a deviation).
• Counter-balance reconciliation variance — average absolute delta between planned and actual counter-balance mass after adjustment (target ≤0.1%).
• Yield-vs-adjusted-charge correlation — does actual yield track the adjusted-charge prediction? Drift suggests dispense weighing error or upstream loss.
• CoA-to-dispense lag — average days between CoA receipt and first dispense using that CoA (long lag risks using stale assay against re-tested lot).

1. Item master holds reference potency, basis (anhydrous / as-is), salt form, free-base MW, salt MW, marker compound, and designated counter-balance partner.
2. Lot master holds assay %, LOD %, salt-form override, CoA reference, retest dates, PF (computed) — populated from supplier CoA import or manual entry under change control.
3. MMR carries the formula on reference basis with tolerance windows for the adjusted charge (e.g. ±5% from nominal as hard limit).
4. WO release snapshots the formula into work_orders.mmr_snapshot — the rendering basis is locked at release.
5. Kiosk dispense screen shows nominal target, adjusted target, and the full factor formula ('50.000 kg × 100/97.2 × 1/(1−0.042) = 52.430 kg'); operator weighs to adjusted target.
6. Counter-balance excipient is automatically re-targeted in the same batch; the kiosk shows the delta and the new target.
7. Operator + verifier two-person e-signature gates the next dispense step (Part 11 §11.50 + §11.70).
8. Out-of-tolerance adjusted charges (e.g. PF outside 0.90–1.10) auto-open a deviation and block release until QA disposition.
9. BMR/BPR/DHR renders the adjusted-charge math, factors, signatures, and CoA reference contemporaneously — no post-batch transcription.
10. Yield reconciliation at batch close runs against adjusted theoretical (not nominal), so 21 CFR 211.103 variance is computed correctly.

See the FAQs section below for short answers on common operational questions — when adjustment is mandatory, how it interacts with overage, how repacked lots inherit PF, and what to do when no CoA assay is available.