The point about background rates not accounting for health status is valid, but the fact that the ratio is still below one after adjustments suggests the effect is robust.
The point about background rates not accounting for health status is valid, but the fact that the ratio is still below one after adjustments suggests the effect is robust.