Accuracies and inaccuracies of the family history method: a multivariate approach

Roy M‐A, Walsh D, Kendler KS. Accuracies and inaccuracies of the family history method: a multivariate approach. Scand 1996: 93: 224–234. © Munksgaard 1996. This paper reports on 1459 first‐degree relatives of probands with schizophrenia or affective illness and matched community controls. We sought (i) to validate psychiatric diagnoses obtained by family history (FH) against those obtained by a best estimate (BE) procedure based on personal interview and (ii) to explore the factors influencing the accuracy of the FH report. We found relatively poor agreement between the FH and BE diagnoses, and the disagreements were influenced by numerous factors, including gender, psychiatric status of the informant or proband's diagnosis. When validated against a BE diagnosis, the overall accuracy of the FH method is relatively poor, and is furthermore subject to several biases. Therefore, substituting the FH method for BE diagnosis may be an important source of error in the investigation of familial/genetic factors in psychiatric disorders. Earlier versions of this paper were presented at the Canadian Psychiatric Association Meeting (1993), at the International Congress in Psychiatric Genetics, (1993), at the Behavior Genetics Association Meeting (1994) and the Congress of the International Genetic Epidemiology Society (1994). An earlier version of this paper received the 1993 R.O. Jones award for the best paper by a regular member of the Canadian Psychiatric Association.