CPV

CPV is where validation stops being a project and becomes an operating discipline. Having proved at PPQ that the process can perform, Stage 3 asks whether it still does — batch after batch, through raw-material lot changes, seasonal shifts, operator turnover, and equipment ageing. It is the only stage that runs for the commercial life of the product.

The mechanism is statistical trending of the parameters and attributes that matter — the CPPs and CQAs identified in Stage 1 — usually with control charts and capability indices such as Cp and Cpk. The point is to detect a process moving within specification but away from its centre, because that drift is the early warning that precedes an out-of-specification result, not the result itself.

A CPV programme needs a written plan: what is monitored, at what frequency, by which statistical method, with what action limits, reviewed by whom, and what happens when a trend triggers. Without that, monitoring degenerates into collecting numbers no one acts on. Findings in this area are common precisely because the data is usually being gathered — it is the analysis and response that is missing.

• —Stage 3 of the FDA lifecycle — ongoing verification for the commercial life of the product.
• —Trend the CPPs and CQAs established in Stage 1, using control charts and capability indices (Cp/Cpk).
• —Detects drift inside specification — the early warning before an OOS, not the OOS itself.
• —Requires a written plan: parameters, frequency, statistical method, action limits, review, and response.
• —Feeds change control, CAPA, and the annual product review / PQR rather than sitting in isolation.
• —Collecting data without analysing and acting on it is the classic finding.

FDA, Process Validation: General Principles and Practices (2011), Stage 3; EU GMP Annex 15 (2015) — ongoing process verification; EU GMP Part I Chapter 1 §1.10 (Product Quality Review); ICH Q10 for the continual-improvement linkage.