Cultural diversity in subject selection is strongly encouraged. Clinical studies should be inclusionary, designed in such a way that the subject pool is representative in gender, race, ethnicity, and age of the local population affected by the condition being studied. Exclusion of particular subgroups may be justified in the following cases:

1. A specific scientific question is more appropriately addressed with a more targeted group,
   
2. Specific known risks of intervention (or other participation) warrant exclusion of specific groups,
   
3. Gender- or racial-neutrality has been previously demonstrated,
   
4. Redress of gaps in medical knowledge about specific groups,
   
5. Multicenter studies may achieve overall balance by choosing centers with complementary minority representation.

Subject enrollment need not be sufficient to provide high statistical power for calculating effects of gender and minority status on intervention effect in all cases. Investigators planning phase 3 clinical trials in which prior studies suggest a clinically “significant difference” in effect among subgroups or those in which differential effects are uncertain are encouraged to design the subject pool in sufficient numbers so as to provide a “valid analysis” in each subgroup.