Review each batch production record before disposition and decide every exception
- Only the entries that break a rule
- Each sum shown with the pages behind it
- Logbooks and solvent lots read beside the record
A batch record review for drug substance plants that reads each API batch as its stages, redoes every yield and hold clock, reads the equipment logbooks and recovered-solvent lots beside the record, and hands QA only the entries that break a rule.
See one API batch reviewed screen by screen ↓An API batch record is several stage records written across weeks by several shifts and handed to QA as one stack. The checks that matter cross pages: a stage yield needs a wet mass from one page and a loss on drying from another, and a hold clock starts on an isolation page and stops on the next stage's charge page. A page-by-page read finds the missing initials and leaves the cross-page arithmetic to a calculator and a long memory.
Logbooks, maintenance requests and recovered-solvent releases are separate records, opened only when the batch record points to them. A valve adjusted by hand during a cooling ramp can be written in the logbook and nowhere else, so it reaches no deviation, and the reactor fails the same way on a later batch.
Built forPlant QA headsQA batch record reviewersMSAT process engineersProduction shift supervisors
QA reads each stage record page by page, while the event that matters is often written only in an equipment logbook.
Checked by hand on each batch
Built-in controls
QA decides exceptions instead of reading pages, and each client receives a signed review it can check.
The QA reviewer runs the agents, steers them to the maintenance records, decides each exception and hands the Majors to the plant QA head, up to a disposition that waits for each open item.
demoSample data: each batch, vessel, reviewer and figure, and both companies named on screen.
Production answers by the box, MSAT reads the campaign, and only QA signs.
Review each batch production record before disposition and decide every exception
Add the second signature on Majors, approve learned rules and disposition the batch
Answer each correction with a dated late entry or a deviation
Watch yields and hold clocks across the campaign
The agents read, sum, time and challenge; the reviewer and the QA head decide.
Read each stage record against the master instructions, one reader per stage.
Redoes each stage's mass and molar yield from the wet mass and measured loss on drying.
Times each hold from isolation to the next charge against its hold-time study.
Checks each recovered-solvent lot's release, reuse cycle and status at the charge.
Reads the logbooks, maintenance requests and historian trends of each vessel the batch used.
Pins each exception to its cell with its rule, sources and a proposed action.
Argues as an inspector would against the exceptions and, as hard, against what the agents cleared; the others answer or hand the point to a person.
Drafts the client's summary from the signed review.
Four findings on one sample batch that a page-by-page read tends to pass.
Stage 4 was written as 87.2% on an assumed 5.0% loss on drying; the measured 6.8% gives 85.6%, and both sit inside 82–90%.
The Stage 2 intermediate waited 73 h 45 min against a study that supports NMT 72 h, timed from its isolation page to the Stage 3 charge.
The brine valve was slow to open at 21:40 and the cooling was adjusted by hand; it stuck on a later batch, and the entry is now evidence in that investigation.
The agents first cleared the recovered THF on its release record; asked about its status at the charge, they re-marked it a Major.
demoSample data, from the working solution.
Past the one batch: a second product, the campaign, corrections in flight, the queue, learned rules and the trail.
demoSample data, from the working solution.
Batch record review is the quality unit's check of each batch production record before an API batch is dispositioned, which ICH Q7 asks for on critical process steps. Here agents read the stage records and the logbooks, maintenance requests and solvent releases around them, redo the sums and leave QA only the entries that break a rule.
Charges against their ranges and second-person checks, in-process results, yields at each isolated stage, hold times against their studies, allowed loops such as drying on, recovered-solvent lots, label reconciliation, GDP corrections and the equipment logbooks. Here the checklist is read from each product's own master instructions.
Each entry is still read and cited; what changes is who reads it first. A person decides the entries that break a rule, and items cleared by a rule, such as drying continued and re-tested, are listed with the rule instead of hidden.
Yes. Stage records may be scanned paper, hybrid or electronic, and the logbooks, maintenance requests and solvent releases sit in other systems. A scanned page that cannot be read is marked as not read, and a rescan is asked for.
No. The review is signed only once each exception is decided and each correction accepted or raised as a deviation, and the plant QA head's disposition waits for the deviations, the linked investigation and release testing. The client's qualified person decides for the medicine made from it.
A summary of the signed review for its own batches only, with each exception, its decision and its state, which the QA head opens before it is sent. A summary withdrawn after a later change stays listed as withdrawn.
The holder finds its own gap and tells the client first, so each client's filing rests on a file that matches the plant.
1-min demoImpurity Risk Assessment (ICH M7, Q3C, Q3D)The filing's impurity section is ready before validation starts, and every purge claim shows its sum.
1-min demoProcess Development Report AuthoringOne sourced set of claims carries the report, the impurity assessment and the PPQ protocol, and the client's reviewer can check each one.
1-min demoSend one closed API batch record with its logbooks and solvent releases, and we will lay it out stage by stage and mark its exceptions.