It's done by demonstrating to CASA, within the framework of their regulations, that the risk is acceptable to them.
And there, writ large, is the fundamental error made by the zealots on a medical crusade.
The risk that is acceptable is already built into the regulations and does not require anything to be acceptable or unnaceptable to anyone in CASA. If they actually read the regulations and complied with them, the zealots on a medical crusade would realise that there is actually
no role to play for any medical expert in the process of demonstrating compliance with the colour perception standard.
Let's remind ourselves of what the actual rule about demonstrating compliance with the colour perception standard actually says:
(6) A person must demonstrate that he or she meets the criterion in item 1.39 of table 67.150 by:
(a) in daylight, or artificial light of similar luminosity, readily identifying a series of pseudo‑isochromatic plates of the Ishihara 24‑plate type, making no more than 2 errors; or
(b) for somebody who makes more than 2 errors in a test mentioned in paragraph (a), readily identifying aviation coloured lights displayed by means of a Farnsworth colour‑perception lantern, making:
(i) no errors on 1 run of 9 pairs of lights; or
(ii) no more than 2 errors on a sequence of 2 runs of 9 pairs of lights; or
(c) for somebody who does not satisfy paragraph (a) or (b), correctly identifying all relevant coloured lights in a test, determined by CASA, that simulates an operational situation.
A person who passes the Ishihara test in (a) has demonstrated compliance with the standard. It doesn't matter whether that outcome is acceptable to CASA or not. That test can be administered by a trained monkey.
A person who passes the Falent test in (b) has demonstrated compliance with the standard. It doesn't matter whether that outcome is acceptable to CASA or not. That test can be administered by a trained monkey.
A person who passes the simulated operational situation test in (c) has demonstrated compliance with the standard. It doesn't matter whether that outcome is acceptable to CASA or not. That test cannot be administered by a trained monkey. It must be administered by a person with operational expertise.
The test for the purposes of (c) can't be determined by a medical expert. It must be determind by an operational expert. The point of the test is not to find out whether the candidate has a defect in colour perception: We already know that s/he does. The point of the test is to find out whether the candidate is able to perform simulated safety-critical operational tasks that depend on the identification of the meaning of lights that are coloured, as effectively and efficiently as candidates who don't have colour vision deficiency. Medical experts wouldn't have a clue about that. It's been proved: The CAD test doesn't simulate sh*t.
But because zealots on a medical crusade are superior and need to save the world, mere trivialities like the law are but a minor irritation. It's all about
them and what's "acceptable" to
them. (That's why they need to pry into and know everything about your personal life, Kharon.)