How to structure a qualification protocol

• Title, unique identifier, version, approvals (author, reviewer, quality).
• Purpose, scope and system description.
• References: URS, DQ, drawings, procedures, standards.
• Responsibilities.
• Prerequisites: calibration status of test equipment, training, completed prior stages.
• Test procedures, each with: requirement reference, method, equipment, acceptance criterion, result field, executed by/date.
• Deviation handling and how results are recorded.
• Attachments: calibration certificates, raw data, diagrams.

The report mirrors the protocol test by test, adds the deviations and their resolution, and states whether the system is qualified — signed by the same roles that approved the protocol.

The section that decides whether the protocol works

Everything above is administration except one part: the test procedures. A protocol succeeds or fails on whether each test can be executed by a competent person who was not in the room when it was written, and whether the result can only be read one way. That means every test carries five things — what requirement it traces to, exactly what to do, what equipment to do it with, what result counts as a pass, and space to record what actually happened.

Approve before you execute, and mean it

The protocol is approved before execution because that is the only moment at which the acceptance criteria are not yet influenced by the result. A criterion adjusted after seeing the data is not a criterion, and an inspector reading two signature dates in the wrong order will draw exactly that conclusion. Where a criterion genuinely turns out to be wrong, the route is a recorded deviation and a protocol amendment with its own approvals — not a quiet edit.

Executing it

• Record raw data contemporaneously, in the protocol or in an attachment referenced from it. A result written up the following week from memory is not a record.
• Attach the evidence: logger exports, screenshots of alarm receipt, calibration certificates for every instrument named, photographs of sensor placement.
• Correct errors the GxP way — single line through, initial, date, reason. Never overwrite and never use correction fluid.
• Raise a deviation for anything that departs from the written method, including departures that turned out fine. The deviation log is evidence of control, not a confession.
• Have the same roles that approved the protocol approve the report, with the qualification statement drawn from the results rather than asserted alongside them.

A protocol nobody can execute without asking the author a question is a draft. The cheapest review is to hand it to a colleague from another team and watch where they stop.