The PIC/S GMP Guide and EU GMP are almost identical, revised Annex 1 included. In the EU the GMP principles are set in law and certificates sit in EudraGMDP; PIC/S authorities adopt the guide nationally and PIC/S certifies nothing. One design serves both; the EU inspection route is planned separately.
A failed acceptance criterion and a protocol step not executed as written are different OQ deviations. The first needs root cause, impact assessment, a fix and a repeat test; the second needs an assessment of whether the result still stands. Changing a criterion after a failure invites audit questions.
When deviations, change control and EM trends live in different systems, one event can carry three dates. Rehearse the first-hour inspection requests by pulling all seven records for one day six months back, then check that each system gives the same date and cross-reference for it.
A depyrogenation tunnel runs from the wash room into a Grade B filling room, across one or two grade steps, and releases heat next to the Grade A zone. Its own airflow balance and heat release per zone belong in the design basis, and the rooms are balanced with the tunnel at temperature.
Annex 1 expects the Contamination Control Strategy to be actively reviewed and updated. A yearly meeting catches changes late. Adding one question to every HVAC change control, EM adverse trend review and deviation closure form, asking whether the CCS still says the right thing, keeps it current with the site.
EU GMP Annex 15 asks for transport of finished products, investigational products and samples to be verified per route, seasons included. The plan is usually sound; the gap is the logger nobody checks before release. Name one role that downloads each logger on time, checks it against limits and can hold the stock.
GAMP 5 Second Edition (2022) scales validation effort to risk to patient safety, product quality and data integrity, and adds agile and cloud guidance. On a GMP BMS, test alarms, audit trail and data transfer deeply. A quick check: count low-risk functions that still carry full test scripts.
21 CFR 11.200(a)(1) requires two distinct identification components for non-biometric signatures. Later signings in one continuous session must still use a component only the signer can execute, usually the password. Define the session and its timeout in writing, then test both cases at OQ.
A test at minimum fan speed with a door held open is OQ evidence; PQ shows normal operation, with production or a representative simulation, over a stated period. A step that forces a condition usually belongs in OQ, and separate protocols allow OQ to close before PQ starts.
IQ verifies an installation against the approved drawings. When site changes are not red-lined and reviewed, IQ either raises late deviations or signs off drawings that no longer match the plant. Reviewed as-builts belong in the turnover package before IQ starts.
A timed airflow setback is simple to validate but blind to off-schedule activity. Particle-count-driven airflow follows real use and can save more, but counters and their failure modes join the validated system. Either way, the room must recover to its at-rest limits, pressures stable, before operations restart.
Measuring every room against one reference stops errors stacking but leaves room-pair differences at up to ±2 Pa with ±1 Pa sensors. Room-to-room sensors show airflow direction directly but add errors along the chain. Record one set, compute the other, and state the error budget.