To overcome this, there have been suggestions in the past that the DSM should be divided into two: a scientific version, for use by researchers and psychiatrists, and a pragmatic version, for everyone else.
Finally, you use the non-supporting evidence to develop new conclusions that are more tempered and pragmatic than your initial Chicken Little panicking.