Learnaboutgmp Community

The 10 Validation Commandments

[COLOR=#333333][COLOR=#333333][FONT=Arial]1. Thou shalt not include requirements that cannot be tested

  1. Thou shalt not hide requirements just because they are difficult to test

  2. Thou shalt not over complicate your validation protocols.

  3. Thou shalt not repeat the same requirement to bulk up your URS

  4. Thou shalt not execute a validation protocol without executing a dry run first

  5. Thou shalt include documented evidence of executed tests where required

  6. Thou shalt not make the tests over complicated

  7. Thou shalt execute prerequisites before protocol execution

  8. Thou shalt read the protocol before execution

  9. Thou shalt not pass a test even though it has failed[/color][/font][/color]

Any more …

Ouch… guilty of #9. :slight_smile:

I think #7 contradicts #1.

How about (in the medical world) “Thou shalt use Good Documentation Practices when recording test results.”

A few more:

  • Thou shalt only use confirmed calibrated equipment when measuring results

  • Thou shalt have known inputs (measured) if results are measured*

(* We had a situation where we fed in some material without first measuring it, operated on the material, then measured the results. Turns out the variability of the input caused the test to fail even though the equipment performed as expected.)

  • Thou shall not sign off on a deviation you dont agree with.

  • Thou shall not assume everything an engineer says in correct.

Thou shalt not change an expected result when executing a protocol.

  • Thou shalt not use pencils or tipex when writing on a protocol.

  • Thou shalt not rush an execution just because it is coming close to the end of the day.

  • Thou shalt document bugs found even though they don’t impact the execution of the protocol. (continuous improvement)