QSIT includes two tables that provide sampling plans for the FDA Investigator. When I talk to companies about sample sizes, I find they tend to use small convenience sample sizes (3 to 5 records). Following the QSIT tables, an Investigator would request a larger sample size and, consequently, identify nonconformances not disclosed in the smaller audit sample.
The implication is that an internal audit program with small samples might “be happy” with records while an FDA Investigator would find “observations” that result in a 483.
Without getting into the statistics of sampling plans, I’m curious about
- a) the frequency at which FDA Investigators actually use the QSIT tables
- b) the frequency at which internal audits don’t find problems while the Inspection does because of larger sample sizes
------------------------------
Dan O'Leary
Swanzey NH
United States
------------------------------