PPQProcess Performance Qualification

Process Performance Qualification (PPQ) is Stage 2 of FDA's three-stage Process Validation (PV) lifecycle defined in the 2011 PV guidance: Stage 1 — Process Design (build the process based on development knowledge); Stage 2 — Process Qualification (qualify the facility and equipment, then run PPQ); Stage 3 — Continued Process Verification (monitor for life). PPQ is the prospective demonstration on commercial-scale batches that the process, as designed and qualified, consistently produces product meeting all CQAs.

EU GMP Annex 15 uses different terminology — Performance Qualification (PQ) at the process level, or process validation as a discrete activity — but the substance is the same: a prospective protocol, predetermined acceptance criteria, commercial-scale evidence, and a report concluding the process is fit for routine use.

The 'three consecutive successful batches' convention predates the 2011 guidance — it was the de facto rule under the 1987 PV guidance. The 2011 revision explicitly removed the fixed three-batch rule: the number of PPQ runs must be justified by the statistical confidence required, the variability observed in Stage 1, and the risk of the product. Three remains the default for low-risk processes with extensive Stage 1 data; high-variability or novel processes routinely need five, ten or more batches before acceptance.

FDA's Compliance Program 7356.002 inspector guidance now expects a statistical rationale for the batch count. Sites that submit three batches with no rationale draw a 483 even if all three pass.

• Process description — every unit operation, every parameter, every set-point, with references to the master batch record version under qualification.
• Equipment list with completed IQ/OQ/PQ references — PPQ cannot start on equipment that is not qualified.
• Sampling plan — typically 3–5× the routine sampling frequency, with sample plan justified statistically.
• Critical quality attributes (CQAs) with acceptance criteria tighter than release specs — typically the inner 95 % of the release spec, with Cpk acceptance (commonly ≥1.33) on quantitative CQAs.
• Critical process parameters (CPPs) with acceptance ranges — process parameters that drift outside the validated range trigger a deviation even if the CQAs pass.
• Hold-point list — where the process pauses for QA decision before continuing.
• Pre-defined deviation handling — minor / major / critical classification with the action for each, and the explicit list of events that constitute PPQ failure (vs a minor anomaly).
• Acceptance criteria for the protocol as a whole — number of batches that must succeed, what success means, what happens on a failed batch.

PPQ samples the process more aggressively than routine production. For a tablet press, routine might be content-uniformity n=10 every hour; PPQ might be n=30 every 15 minutes with samples taken from the beginning, middle and end of the compression run and stratified across the punch positions. The elevated sampling has two purposes: (1) to detect any within-batch heterogeneity the design data didn't surface, and (2) to provide a denser statistical baseline against which Stage 3 monitoring can detect drift.

The elevated sampling regime usually persists for the first 3–10 commercial batches after PPQ closes (Stage 3a), then steps down to routine sampling once statistical evidence supports the reduction. The step-down criteria must be in the protocol — not invented after the fact.

Every CQA in the PPQ protocol must have a pre-specified, scientifically justified acceptance criterion. 'Meets release spec' is not enough — PPQ criteria are tighter than release specs to provide margin against process drift. Typical pattern: release spec 95.0–105.0 %, PPQ acceptance 97.0–103.0 % on the mean and Cpk ≥ 1.33 on the spread. The criteria must be tighter than the release spec by an amount the protocol justifies.

For Stage 3 CPV planning, the PPQ data also feed the establishment of process control limits — typically ±3σ from the PPQ mean, narrower than the spec, used as out-of-trend signals during routine operation. A PPQ that runs near the edge of the spec leaves no room for OOT alerts.

1. Three-batch rationale absent — number of batches not justified statistically against Stage 1 variability.
2. Acceptance criteria identical to release spec — no margin, no protection against drift.
3. PPQ batches run on equipment whose IQ/OQ/PQ was not closed at the time of PPQ start.
4. Sampling plan less aggressive than routine — defeats the purpose of PPQ.
5. Deviations during PPQ classified as minor without investigation, then PPQ declared successful.
6. Stage 3 monitoring not designed — CPV protocol does not exist when PPQ closes.
7. Process changes after PPQ start without re-baselining — change-control through a PPQ batch invalidates the data.
8. Hold-point QA decisions documented after the batch released — paper PPQ.

• PPQ protocols are first-class objects with elevated sampling rules that flow into the work-order BPR automatically.
• Equipment-qualification status is enforced — work orders cannot start a PPQ batch on equipment whose IQ/OQ/PQ is not current.
• CQA acceptance criteria are stored separately from release spec — PPQ runs evaluate against the tighter criteria and surface the gap in the report.
• Cpk and other capability metrics compute live as PPQ batches complete; the report PDF renders from the same data as the dashboard.
• Deviations during PPQ route to the validation team automatically with the protocol's classification rules pre-applied.
• Stage 3 CPV control limits are derived from the PPQ data on protocol closure, with the sampling step-down schedule applied automatically.