Run the quality risk management program and check each moved score
- Every new event read against every assessment
- Each decision to check, rejections included
- An evidence export ready for an inspection
Quality risk management that reads each new investigation, change and CAPA for the control it breaks, moves the score with its source, and waits for the risk owner to accept.
See one granulation FMEA move after an investigation ↓An FMEA is written by a cross-functional team, approved, then reviewed on a calendar, often by re-scoring the table from memory. Deviations, investigations, changes and CAPAs arrive in between, and each form's tick box asks whether a risk assessment is impacted; the investigator answers from what they remember of the file.
The assumptions sit in the current controls column, where nobody can search them. An investigation finds that a software update widened an alarm limit, yet the failure modes scored low because of that alarm keep their scores, and so does the sister machine. The inspector compares the risk file with the deviation log and finds the gap first.
Built forQA risk management leadsRisk ownersQA headsSite quality heads
Risk files are re-scored on a calendar from memory, so a control that stopped working still holds its risks low.
Compared from memory today
Built-in controls
A closed investigation now reaches each risk file that leaned on the failed control, at each plant.
A root cause is approved: a software update widened the high-torque alarm on a wet granulator. The risk owner, the QA risk lead and the QA head carry the plant's granulation FMEA from that finding to a reopened file.
demoThe CDMO, its plants, clients, people, records and scores are sample data.
The risk owner rules on each move, QA checks the ruling, the QA head accepts what stays high, and other plants decide on their own files.
Run the quality risk management program and check each moved score
Own process FMEAs and decide each proposed move on their files
Accept high risks for a set time and approve learned rules
Decide on their plant's assessments when another plant's finding reaches them
A software update widened the high-torque limit, the change control never compared it with the validated value, and the file still scored detection on that alarm.
Granule density used it as prevention and the over-wet granule as in-step detection; the challenger's question brought both into the read.
The loss-on-drying limit at the end of drying still catches an over-wet granule, so the draft score was revised down a step.
Another product's FMEA on the sister granulator rests on the same widened alarm, and joins the same cross-functional review.
Each one reads plant and QA records alone, puts a source behind each proposal, and leaves each decision to a named role.
Reads each deviation, investigation, change and CAPA for the assumptions it breaks or confirms.
Finds each assessment, at any plant, whose controls, frequencies or severity basis rest on that assumption.
Proposes the moved severity, occurrence or detection with the scale definition and the passage.
Plays the ICH Q9 reviewer and the inspector, testing the agents' own scores as hard as anyone's.
Shows how this risk owner and QA scored similar evidence before, and proposes rules from their rulings.
Plays forward interim controls and CAPA actions to show where each pin would land.
Drafts the reopening record, the review agenda and the invitations to reviewers.
Behind the map sit the records an inspector asks for next, and the place where subjectivity is measured.
demoSample data, from the working solution.
ICH Q9 defines quality risk management as a systematic process for the assessment, control, communication and review of risks to product quality across the lifecycle. Here each assessment stays current: agents read new deviations, investigations, changes and CAPAs against it, and the risk owner decides each proposed move.
After new knowledge and experience, from planned or unplanned events such as root causes from failure investigations, at a frequency set by the level of risk. The review may reconsider earlier risk acceptance decisions, which is why the challenger flags them for the QA head.
Formality as a spectrum, risk-based decision-making, managing and minimizing subjectivity, and product availability risks. Blind pre-scores, with the spread between reviewers shown and discussed, answer the subjectivity section with data.
Only on the plant's own scales. Occurrence needs the events and the batch count behind them, and detection follows the step where the failure was actually found. The score mover cites both, and the challenger checks every occurrence move against that rule.
Deviations, investigations, change controls, CAPAs, batch records, historian trends, maintenance records, configuration exports, procedures and validation reports, each read against the assessments that rest on it.
The client's quality agreement sets the notice; it states the move and its interim control, waits for the QA head's high-risk acceptance, and goes out once the QA risk lead has read it.
The record gap is found by your own back room, disclosed and logged, before an investigator reads the page.
1-min demoContamination Control Strategy (Annex 1)A weak barrier shows the day its records change, not at the yearly review.
1-min demoClient Batch Release PackagePackages reach the client already checked the way the client's own reviewer checks them, so its release is not held up by send-backs.
1-min demoSend one FMEA with the deviations, changes and CAPAs since its last review, and we will show which scores they move.