Discussion

Abstract

Despite the enormous progress made in investigative methods in basic sciences, social- and psychodynamics and in epidemiological and social-psychiatric research, increasing stagnation has occurred in acquiring of knowledge of the psychoses of the “schizophrenic spectrum” during the last several years. It has still not yet been settled whether the spectrum of schizophrenic and schizophrenia-like psychoses is a disease continuum with continuous boundaries or whether it is composed of different diseases with very different causes. The great hope placed in modern, atheoretical and operational diagnostic systems has not yet been realized. Further, there are very few findings which have not been contradicted. It therefore seems justified to question whether the mere compilation of a number of specific symptoms based on expert consensus, in order to form diagnostic categories and/or “disorders” without a nosological background, is actually sufficient to advance schizophrenia research. This expert consensus is a compromise of various individual viewpoints and should serve first of all to improve mutual understanding. In attaining a high degree of diagnostic correspondence between different investigators (reliability), however, it is often necessary to pay the price of over-simplification. It is also incorrect to automatically assume that the consensus criteria are also valid. Reliability does not necessarily mean validity (Gottesman and Shields 1982). A systematic twin study which compares diagnostic systems based on “expert consensus” (DSM-III-R, ICD 10) with Leonhard’s nosology based on clinical-empirical research appears to be appropriate for checking the validity of diagnostic subgroups within the schizophrenic spectrum.