21 CFR 820.250 - Statistical Techniques (QMSR)

What it says

Where appropriate, valid statistical techniques must be used to establish, control and verify the acceptability of process capability and product characteristics, with sampling plans, when used for inspection, based on a valid statistical rationale.

Paraphrased for plain-language clarity. Always verify against the current published regulation text.

What it means in practice

• This provision requires genuine statistical validity behind sampling and acceptance decisions, conceptually paralleling the statistical rigor described in USP <1210> for analytical procedure validation, but applied here to broader process and product acceptance decisions.
• A sampling plan used for inspection must have a documented statistical basis, meaning arbitrary sample sizes chosen for convenience rather than statistical adequacy do not satisfy this requirement.
• Statistical process control techniques, where used to monitor and control manufacturing processes, fall within this provision's scope, connecting statistical rigor to ongoing process monitoring, not just point-in-time acceptance decisions.

What FDA inspectors look for

• Whether sampling plans used for inspection and acceptance decisions have a documented, defensible statistical rationale.
• Whether statistical process control data, where used, is genuinely analysed using valid statistical methods, not simply plotted without meaningful statistical interpretation.
• Whether statistical techniques applied are appropriate for the specific data type and decision being made, rather than a generic approach applied regardless of actual applicability.

Related regulations