This is most definitely an issue in psychology. Granted, it's improved significantly since the 60s and 70s (see the Rosenhan experiment - both enlightening and amusing ). But both professionals and laymen alike still have problems with stigmatizing certain labels and making assumptions about a person based on a label.

Any propositions on how we can remedy this?

We could do away with labels and diagnoses entirely, but that would make treatment, research, and communication difficult for psychologists.