British Journal of Psychiatry (1989), 154 (suppl. 4), 91-95 Comparative Distributions of DSM - III Diagnoses In North and South American Clinical Samples CECILIA SOGI, DANTE WARTHON, JUAN E. MEZZICH, JESUS VALVERDE, ALFREDO SAAVEDRA-CASTILLO and CHUL W. AHN Growths in communication technologies, voluntary travel, and forced migration are fostering reciprocal awareness and a sense of interdependence among people from distant parts of the world. In line with these, there is increasing interest in health status and illness conditions across different geographical and cultural regions. Adequate identification of health problems in each region serves not only to enhance planning of treatment and preventive activities for both individuals and communities, but also may point out types of pathology that would require greater attention in diagnostic systems intended for wide international use, such as the prospective 10th Revision of the International Classification of Diseases (ICD) (Sartorius, 1988). Illustrating contrasting nosological distribution, Binitie (1988) has noted that anxiety disorders are the most common psychiatric disorders in Nigeria. Wig et al (1985) has suggested that among the prominent psychopathological conditions in the Third World are acute transient psychoses and neurotic conditions characterised by multiple physical complaints, and that these have not received adequate attention in diagnostic systems generated in technolo- gically developed countries. Comparative studies using sound methods should be helpful for further documenting and delineating these observations, but diagnostic instruments of adequate validity and reliability are required for this purpose. A major recent approach for the improve- ment of diagnostic reliability has been the use of explicit or specific criteria for the assignment of nosological entities - an approach pioneered by Berner and implemented most extensively through DSM- III (American Psychiatric Association, 1980) not only in its country of origin but also in many parts of the world (Mezzich, 1987). The other major approach towards enhancing diagnostic reliability has been the use of standardised evaluation instruments, which specify the areas and items to be covered and provide definitions and other guidelines for rating interviewee's responses (Richardson et ai, 1965). Additionally, the wording and order of the questions may characteristically be fixed in rigidly scheduled instruments. This arrange- 91 ment typically limits the possibility of probing intricate areas (e.g. psychotic symptoms), and leads to covering only a subset of the range of general psychiatric disorders. An alternative is represented by semi-structured procedures (standardised but only minimally scheduled), designed for well-trained clini- cians using experience and judgement in conducting the interview and probing the patient. Employing semi-structured evaluation procedures and pertinent diagnostic criteria, the present study was aimed at comparatively appraising DSM-III diagnostic distributions in clinical samples from North America (Pittsburgh) and South America (Lima). Method Clinical setting Two comprehensive psychiatric facilities served as sources of patient samples for this study - one in Lima, Peru, and the other in Pittsburgh, USA. The Peruvian centre, Instituto Nacional de Salud Mental 'Honorio Delgado-Hideyo Noguchi', was inaugurated in 1982 as a normative national institution with research, educational, and patient-care functions. Its area of clinical responsibility corresponds to the populous northern sector of the city of Lima. It offers care to patients of both sexes and all ages, experiencing all forms and levels of psycho- pathology. During the study period, it provided about 500 hospital admissions and over 11000 out-patient visits per year. The United States clinical sample was obtained from the Western Psychiatric Institute and Clinic of the University of Pittsburgh, Pennsylvania. This institution, like the Peruvian one, also has academic and patient-care responsi- bilities, the latter focused on Allegheny County and one of its catchment areas. It provides all modern psychiatric treatment modalities, through approximately 2000 in- patient admissions and over 100000 out-patient visits, to a population of all ages and demographic strata, presenting a wide variety of psychiatric problems. Evaluation procedure At both the Instituto Nacional in Lima and the Western Psychiatric Institute in Pittsburgh, the Initial Evaluation