@strypey The point is twofold:
- These communities are small and yet provide evidence for vaccine efficacy.
- They are still larger than the group of people who will not vaccinate due to medical reasons, making that group a vanishingly small confounding factor.
I do not have a study to reference for this and I will not be digging for it soon. I will be satisfied with saying that peer review and other professionals have not pointed this out as something that undermines epidemiological studies on efficacy.