Validity, reliability and objectivity of the family history method in psychiatry: A meta analysis

Abstract

BackgroundThe family history is a widely used method in psychiatry; but data on the method's objectivity, reliability and validity shows partly diverging results.MethodIn October 2005, a Medline search was conducted that yielded 7 studies regarding objectivity/reliability and 13 studies regarding validity. Results for six main groups of psychiatric diagnoses and any mental disorder were combined qualitatively for objectivity/reliability, and quantitatively for validity.ResultsObjectivity was generally high (κ in the 0.80 range). Reliability was high for any mental disorder, schizophrenia, substance abuse and depression (κ in the 0.70 range), and low or medium for anxiety (κ between 0.30 and 0.50). Results on validity displayed an OR = 148 for the family history for schizophrenia; OR = 64 for mania/bipolar disorder; and OR's between 8 and 194 for substance abuse, between 3 and 37 for depression, between 5 and 350 for personality disorders, between 2.5 and 49 for anxiety, and between 2.4 and 9 for any mental disorder.ConclusionThere is clear evidence that the family history provides results that are better than chance for all disorders examined. But variance among diagnostic groups and among studies is considerable.