cross-cultural differences in the short-term prognosis of schizophrenic psychoses*

Transcription

1 102 SCHIZOPHRENIA BULLETIN cross-cultural differences in the short-term prognosis of schizophrenic psychoses* Norman Sartorius, Assen Jablensky, and Robert Shapiro Introduction More than 70 years ago, Kraepelin (1904) suggested that "comparative psychiatry," which he defined as "the observation of mental disorders in different groups of people," could substantially advance our knowledge about the causes and nature of mental illness. However, he issued the following warning: Reliable comparison is, of course, only possible if we are able to draw clear distinctions between identifiable illnesses, as well as between clinical states; moreover, our clinical concepts vary so widely that for the foreseeable future such comparison is possible only if the observations are made by one and the same observer. The International Pilot Study of Schizophrenia (IPSS) (World Health Organization 1973) attempted to overcome difficulties that had undermined many attempts to compare observations on psychiatric patients made by different researchers in different settings and that had obstructed the development of a common language for describing the major mental disorders. The study began in 1966 as a cross-cultural collaborative project carried out simultaneously by research centers in * This preliminary communication on the 2-year followup of patients included in the International Pilot Study of Schizophrenia of the World Health Organization was prepared by the authors on behalf of the collaborating investigators. Reprint requests should be addressed to the senior author at Division of Mental Health, World Health Organization, 1211 Geneva 27, Switzerland. nine countries that differed considerably in their sociocultural characteristics: China (Taipei), Colombia (Cali), Czechoslovakia (Prague), Denmark (Aarhus), India (Agra), Nigeria (Ibadan), the Union of Soviet Socialist Republics (Moscow), the United Kingdom (London), and the United States of America (Washington). It set out to lay methodological groundwork for future epidemiological and other research in schizophrenia and the other functional psychoses, and to expand knowledge about the nature of schizophrenia. With regard to methodology, the IPSS aimed: (1) to investigate the feasibility of large-scale international studies requiring the collaboration of psychiatrists and other mental health workers from different cultural and theoretical backgrounds; (2) to develop standardized instruments and procedures for psychiatric assessment that could be applied reliably in a variety of cultural settings; and (3) to train research workers in developing and developed countries to use such techniques to make comparable observations. With regard to substantive knowledge on the nature of schizophrenia, the study aimed: (1) to explore whether schizophrenic disorders exist in different parts of the world; (2) to identify similarities and dissimilarities between groups of patients diagnosed as schizophrenic in different cultures, as well as to compare such patients with patients suffering from other functional psychoses; (3) to determine to what extent dissimilarities between schizophrenic patients in different settings result from variations in diagnostic practices or from culture-related differences in the manifesta-

2 VOL. 4, NO. 1, 197S 103 tions of the disorder; and (4) to investigate whether the course and outcome of schizophrenia differ from country to country. The IPSS was not designed to be an epidemiological study, and no attempt was made to sample representative groups of patients. However, operationally defined selection criteria were used at each field research center to screen all patients contacting a psychiatric service during a 1-year period. In this way, it was possible to include in the project patients within the age range of 15 to 44 who had nonorganic psychotic illnesses of a recent onset and who would be available for a followup. A total of 1,202 patients were selected for study in the nine field research centers and were given a detailed standardized clinical examination at the initial evaluation. Of these 1,202 patients, 811 had received a clinical diagnosis of schizophrenia, 164 a diagnosis of affective psychosis, and 227 of other psychoses or nonpsychotic conditions. In the followup phase, the patients were traced and reexamined twice, 2 years and 5 years after the initial evaluation. Since the methodology and results of the initial evaluation phase of the IPSS have been described in detail elsewhere (World Health Organization 1973 and 1975), only the main points will be summarized here as a background to the findings of the 2-year followup study. Standardization of assessment was one of the main goals of the initial evaluation phase. A standardized semistructured interview guide, the Present State Examination (PSE) (Wing, Cooper, and Sartorius 1974) was adapted to the purposes of the study through iterative translation/backtranslation from English into the languages spoken in the different study settings. Investigators in the centers underwent intensive training in its use until within-center and across-center reliability of assessment was achieved. Instruments for recording past history information and social data about the patients were also developed. Although similar training procedures were applied, the reliability assessment of these two tools had to be less rigorous than for the PSE because of the cultural specificity and pilot nature of some of the variables included. Statistical comparisons between patients and centers were based on several procedures: computing of average scores on individual symptoms, plotting of scores on groups of symptoms to obtain psychopathological profiles, and rank-ordering of symptoms most frequently rated as present within centers and in particular groups of patients. Patients were diagnosed using three independent methods. First, clinical diagnosis of the condition was made at each field research center; second, the PSE ratings on all patients were used as input for a computer diagnostic program (CATEGO), which produced a reference classification; and third, a mathematical clustering technique was applied to group patients who had the greatest number of characteristics in common. While variations emerged between the centers with regard to proportions of patients classified differently by the different techniques, the average agreement between clinical diagnosis and the computer reference classification was high (87 percent). Of particular interest was a group of 306 patients from all centers who received both a clinical and a computer diagnosis of schizophrenia, and who were classified into the three statistical clusters that contained more schizophrenic patients than could be expected on a random basis. The patients in this group were designated as "concordant," and their symptom profiles turned out to be even more similar to each other than the profiles of all schizophrenic patients. The main conclusion of the initial phase of the IPSS was that schizophrenic patients with similar characteristics could be identified in all nine of the cultures in which the study took place and that the degree of similarity among them could be specified due to the application of standardized assessment methods ensuring cross-cultural comparability. Furthermore, in all cultures patients diagnosed as suffering from schizophrenia could be distinguished clearly in terms of clinical symptomatology from patients with a diagnosis of affective psychosis or other functional mental disorder. Two-year Followup of IPSS Patients An attempt was made to reevaluate as many of the original 1,202 patients as possible 2 years after

3 104 SCHIZOPHRENIA BULLETIN their initial evaluation (World Health Organization, in press). As a result, 97.1 percent of all patients were traced, and in all centers except one, an average of 82.1 percent were seen and reexamined using the PSE and a followup history and social description schedule. The reliability of psychiatric assessments during the followup phase was maintained by continuing training and reliability exercises both within and between centers. Meetings at regular intervals and exchanges of visits of collaborating investigators, which provided an opportunity for discussion of evolving study plans, proved to be an essential mechanism for coordination of work. Data obtained with the PSE at followup were used for comparisons of symptomatology within and between diagnostic groups and centers, and for comparisons with the initial evaluation profiles. The information collected with the followup psychiatric history and social description schedules (supplemented by narrative summaries) was used to rate each patient on the following dimensions of course and outcome: Length of the initial psychotic episode Proportion of the followup period during which the patient was psychotic Pattern of course (e.g., continuous, remitting with or without relapses) Clinical type of subsequent episodes Degree of social impairment Proportion of the followup period during which the patient was not hospitalized Overall outcome (a combined measure taking into account the proportion of the followup period during which the patient was psychotic, degree of social impairment, and type of remission) These measures were used to compare the course and outcome of patients across diagnostic groups and centers and to assess the predictive power of characteristics of the patients and their illness at initial evaluation. They were also used in analyses of the validity of the different systems of classification that were applied in the study. The results of the 2-year followup (World Health Organization, in press) are summarized below. Variability of the Course and Outcome of Schizophrenia Schizophrenic patients who had similar symptoms on initial evaluation and whose disorders met strict diagnostic criteria showed a marked variability of 2-year course and outcome, both within and across centers. When the range of possible outcomes was divided into five overall outcome categories (figure 1), 26 percent of the schizophrenics fell into the best outcome group (i.e., were psychotic during less than 15 percent of the followup period, were not socially impaired, and had full remission); 18 percent fell into the worst outcome group (i.e., were both continuously psychotic and severely impaired); and the remaining 56 percent were distributed over the three intermediate categories. With regard to pattern of course, 27 percent of all schizophrenics had a single, relatively short psychotic episode followed by full recovery without relapses and social impairment. On the other hand, 26 percent never had a full remission during the 2-year period. At the time of the 2-year followup evaluation, one third of the schizophrenic patients were entirely free of symptoms; more than a quarter, however, were still in the psychotic episode that had occasioned their inclusion in the study. Variability of Course and Outcome of Schizophrenia According to Center When course and outcome of the schizophrenic patients were analyzed by centers and by groups of centers, striking and consistent differences emerged between patients in centers in the developing countries and those in the developed

4 VOL. 4, NO. 1, Figure 1. Distribution of schizophrenics in developing and developed countries over five categories of 2-year outcome % 40r Patients In developing countries u Good outcome i. I 1 Patients in developed countries Poor outcome Note. The categories of outcome are as follows (1) Very favorable psychotic for less than 15 percent of the followup period, then full remission; no severe social impairment. (2) Favorable (a) psychotic for less than 15 percent of the followup period, then remission with tome residual symptoms but' no severe social impairment; (b) psychotic for 16 to 45 percent of the followup period, then full remission and no severe social Impairment. (3) Intermediate (a) psychotic for less than 15 percent of the followup period, then remission with or without residual symptoms, but with social impairment; (b) psychotic for 16 to 45 percent of the followup period, then either remission with residual symptoms but no severe social impairment or full remission with severe social impairment; (c) psychotic for 46 to 75 percent of the followup period, then remission (full or with residual symptoms) but no severe social impairment. (4) Unfavorable (a) psychotic for 46 to 75 percent of the followup period, then full remission but severe social impairment; (b) psychotic for over 76 percent Of the followup period, then remission (full or with residual symptoms) but no severe social Impairment. (5) Very unfavorable (a) psychotic for 46 to 75 percent of the followup period, then remission with residual symptoms and severe social Impairment; (b) psychotic for over 76 percent of the followup period, then remission (full or with residual symptoms) and severe social Impairment.

5 106 SCHIZOPHRENIA BULLETIN countries. 1 On virtually all course and outcome measures, a greater proportion of schizophrenic patients in Agra, Cali, and Ibadan had favorable nondisabling course and outcome than was the case in Aarhus, London, Moscow, Prague, and Washington. When individual centers were compared, schizophrenic patients in Ibadan and Agra had the best course and outcome among schizophrenic patients in all centers. Fifty-eight percent of the followed-up schizophrenic patients in Ibadan and 47 percent in Agra were found to be free of symptoms at the time of 2-year followup. Furthermore, 36 percent of Ibadan schizophrenics and 27 percent of Agra schizophrenics had an initial psychotic episode that lasted for less than 1 month and was followed by a full remission without any relapses. Only 5 percent of Ibadan schizophrenics fell into the worst overall outcome group. For every outcome variable, the schizophrenic patients in Ibadan and Agra had a significantly better outcome than patients in Aarhus, London, Washington, and Prague with patients in Cali and Moscow being intermediate. The markedly different distributions of patients in centers in developing and developed countries over the five categories of overall outcome are shown in table 1. These differences in the course and outcome of schizophrenic disorders in developing and developed countries could not be related to a systematic bias in the selection of patients for study in one of the two groups of centers. For example, it was considered possible that patients with a longer previous history of schizophrenia might be overrepresented in the centers in developed countries, while among the patients selected for study in the developing countries there might be an excess of acute cases of a very recent onset. In fact, the proportion of patients for whom the onset of disorder occurred 6 months or more before the initial evaluation was higher in developed countries, but the proportions of patients with a very recent onset (less than 8 weeks before initial evaluation) were roughly similar in the two groups of centers. However, within each of these two 1 Taipei was not included in these analyses because it did not fit clearly into either of the two groups. subgroups of patients (i.e., those having onset more than 6 months, and those having onset less than 8 weeks before initial evaluation), course and outcome were better for patients in developing countries. The hypothesis that an overrepresentation of chronic patients could explain the poorer prognosis in developed countries was therefore rejected. Type of onset and family history of mental disorder were among the variables whose relationship to the observed intercenter differences in course and outcome was explored. In the total sample of schizophrenic patients (all centers combined), cases with acute sudden onset had a significantly better 2-year prognosis than cases with a slow insidious onset. The acute-onset cases had a generally favorable prognosis, and no significant differences with regard to 2-year overall outcome emerged between such patients in developing and in developed countries. In contrast to this, there were significant differences in overall outcome between patients in developing and in developed countries within the group of cases with insidious onset. Insidious-onset patients in developing countries tended to have a better overall outcome than patients with the same type of onset in developed countries, but this difference reached statistical significance only for patients who had been ill for more than 6 months before the initial evaluation. These findings can be put in another somewhat speculative way. Assuming, on the basis of observed differences in the 2-year prognosis, the existence in the developing countries of a hypothetical "factor" protecting schizophrenic patients from chronic deterioration (or, by the same logic, the presence of a "factor" with a negative influence on course and outcome in the developed countries), then it would seem that the effects of this factor are manifest only in patients with an insidious onset of the psychosis and only after the patient has been ill for some months. The operation of such a factor could not be demonstrated in acute-onset cases, which appear to have a propensity to recovery regardless of the cultural setting. Two groups of variables the presumed biological basis of the disorder and the omnibus

6 VOL. 4, NO. 1, Table 1. Overall outcome: Percentage distribution of schizophrenic patients by center Outcome category Centers in developing countries Centers in developed countries All centers Agra Cali Ibadan Aarhus London Moscow Prague Washington Very favorable 48 Favorable Intermediate Unfavorable Very unfavorable 'See figure 1 for definitions , assortment of variables subsumed under "culture" could be thought of as candidates for the hypothetical factor (although the factor could, of course, result from the interaction of both groups of variables). Unfortunately, available data did not permit an investigation of sufficient depth, but an attempt was made to relate the outcome findings to the presence or absence of a family history of mental illness in the corresponding groups of patients. In the sample as a whole, schizophrenic patients without a family history of mental illness had a significantly better overall outcome than patients with a histery of mental illness in relatives. In the developing countries, however, this difference did not appear, and patients with good and poor prognosis could not be differentiated by family history of mental illness. In the developed countries, a further cross-tabulation indicated that only within the subgroup of acute-onset did cases with a family history of mental illness have a poorer prognosis than cases without such a history. In insidious-onset cases, family history of any type of mental disorder appeared to be unrelated to overall outcome. As tentative clues, these preliminary findings are not incompatible with a hypothesis of cultural influence on the medium-term course and outcome of schizophrenic psychoses. At the same time, they also suggest that considerable differences may exist between schizophrenic illnesses with acute " , " " " " and insidious onsets. These two aspects of the findings were further amplified by multivariate statistical analyses aimed at identifying predictors of course and outcome. Predictors of Course and Outcome Potential predictors, selected from the variables assessed at initial evaluation, were divided into three classes: sociodemographic predictors, past history predictors, and predictors related to characteristics of the initial psychotic episode. Course arid outcome variables were defined in terms of the measures described above, and a stepwise multiple regression analysis was applied to predictors. The major findings are summarized below: The proportion of variance for any of the course and outcome variables that could be explained by the five best predictors was relatively low, ranging from 8 to 22 percent. For the 15 best predictors taken together, this proportion did not exceed 27 percent. Therefore, no single factor and no combination of a small number of "key" factors appeared to be strongly associated with the course and outcome of schizophrenia. It is possible, however, that part of the variance

7 108 SCHIZOPHRENIA BULLETIN in course and outcome might be related to factors not included among the variables assessed. Indirect support for this assumption was derived from the fact that the inclusion of the centers themselves in the regression equation as "blanket" variables comprising an unspecified variety of cultural and other factors increased the proportion of variance explained. The three classes of predictors sociodemographic factors, past history factors, and factors related to the initial pychotic episode were found to be about equal in their predictive power. Three sociodemographic predictors appeared consistently among the best predictors, of this class: social isolation associated with a poor outcome, marital status (widowed, divorced, or separated) associated with a poor outcome, and marital status (married) associated with a good outcome. Three past history predictors consistently emerged among the best predictors of this class: history of past psychiatric treatment, poor psychosexual adjustment, and unfavorable environment, all associated with poor outcome. Among the characteristics of the initial psychotic episode, two factors appeared consistently among the best predictors: length of the psychotic episode before initial evaluation, and insidious onset, associated with poor outcome. Other predictors included presence of precipitating factors, presence on initial evaluation of derealization and affective symptoms (all associated with good outcome) and presence of flatness of affect, associated with poor outcome. The proportion of course and outcome variance that could be explained by the best predictors was different for the groups of schizophrenics in the developing and developed countries, The power of the best predictors appeared to be considerably higher in the developed countries. Furthermore, there were differences in developed and developing countries with regard to which specific factors best predicted outcome (table 2). Thus, social isolation assessed on initial evaluation and length of the psychotic episode before the patient's entrance in the study were among the best predictors in both developing and developed countries. However, marital status, educational level, and history of contacts with nonmedical agencies because of mental disorder were among the best predictors in developing countries only, while sex (female sex predicting better outcome), history of physical illness or disability (associated with a tendency to better outcome), and occupation (professional and managerial occupations predicting better outcome) were among the best predictors only in developed countries. There were major differences between the best predictors for schizophrenia and the best predictors for affective psychoses. Some of the main findings about the predictors of 2-year course and outcome in schizophrenic patients are summarized in table 3. The IPSS results provide some support for hypotheses previously held (Vaillant 1962) about the relationship between particular predictors (e.g., previous length of illness, type of onset, presence/absence of precipitating factors, presence of affective symptoms) and outcome. At the same time, they suggest that predictive factors identified in European and North American cultures may be neither relevant nor sufficient for the study of prognosis in other cultures. A large part of the variance in the course and outcome of schizophrenia may be due to factors that have not yet been identified. This conclusion is, by and large, in agreement with findings reported earlier by Raman and Murphy (1972) based on observations of the course and outcome of schizophrenic patients in Mauritius. Discussion and Conclusions As well as demonstrating that international comparisons of psychopathology and prospective observations on psychiatric patients are feasible, the IPSS generated a wealth of cross-cultural data that give rise to hypotheses about the nature of

8 VOL. 4, NO. 1, Table 2. Best predictors of overall outcome in patients with initial evaluation diagnosis of schizophrenia All patients Patients in developing countries Patients in developed countries 1. Social isolation 2. Length of the episode before initial evaluation 3. Past history of psychiatric treatment 4. Marital status 5. History of behavior disorder in adolescence Percentage variance explained by: (a) 5 best predictors= 19% (b) 15 best predictors = 23% 1. Marital status 2. Social isolation 3. Length of the episode before initial evaluation 4. History of contacts with nonmedical agencies because of psychiatric problem 5. Educational level Percentage variance explained by: (a) 5 best predictors = 17% (b) 15 best predictors = 18% 1. Social isolation 2. Length of the episode before initial evaluation 3. Occupation: managerial or clerical 4. Psychophysical symptoms 5. Occupation: professional Percentage variance explained by (a) 5 best predictors 30% (b) 15 best predictors 34% Table 3. Predictors of several measures of course and outcome in schizophrenic patients (all centers) Course and outcome variable Best predictors Variance explained Length of the initial psychotic episode Proportion of the followup period in which patient was in psychotic episodes Full remission without relapses after the initial episode Length of episode before initial evaluation 15% Social isolation Score on derealization Psychiatric treatment in the past History of behavior disturbance Social isolation 12% Length of episode before initial evaluation Psychiatric treatment in the past Sex (male sex correlating positively with proportion of followup period in psychotic episodes) Type of onset Sudden onset 14% No psychiatric treatment in the past No personality change Married Short duration of episode before initial evaluation

9 110 SCHIZOPHRENIA BULLETIN Table 3. Predictors of several measures of course and outcome in schizophrenic patients (all centers) Continued Course and outcome variable Best predictors Variance explained Remitting course with relapses Continuous illness, no remissions Social impairment on followup schizophrenia and the effects dn its course and outcome of social and cultural environment. In considering the findings of the IPSS one must be wary of making unwarranted generalizations. The study was not designed to select epidemiologically representative samples of patients in the different settings where the research centers were located. The findings of the 2-year followup that seem most compelling are: (1) the very marked variability of course and outcome among schizophrenic patients whose diagnoses were based on strict symptomatologic criteria; and (2) the consistent differences between groups of schizophrenic patients in different cultural settings, with patients in developing countries having more favorable course and outcome than their counterparts in developed countries. The observed variability of course and outcome can support both the view that many schizophrenic patients do surprisingly well in later life, and the view that many patients have a very poor Female sex 7% Depression or elation Higher occupational level Neurotic complaints Absence of precipitating stress Long duration of episode before initial 14% evaluation Social isolation Divorced, separated, or widowed Absence of derealization History of behavior disturbance Social isolation 20% Long duration of episode before initial evaluation Past psychiatric treatment Marital status other than currently married No physical illness or disability in the past prognosis and are subject to severe impairments and disability. The emphasis on the one or the other aspect will depend on the issue in question, but in any case the findings so far indicate that a clinical diagnosis of schizophrenia alone (even if based on strictly defined criteria) does not say very much about the patient's likely pattern of course, probability of relapses and remissions, or degree of social impairment in the future. A finding of a greater difference between the course and outcome and subsequent symptomatology of schizophrenic and depressive patients when prediction was based on CATEGO computer classification, rather than on clinical diagnosis, does, however, suggest that the more strictly diagnostic categories are defined, the greater their predictive validity. At the same time, the overlap found between the short-term course and outcome of schizophrenia and of affective psychoses appears to invalidate the belief that schizophrenia invariably has a poorer outcome. The degree of overlap calls for caution in predictions of differences in

10 VOL. 4, NO. 1, length of episodes, proportion of tinu 1 psychotic, level of social functioning, and pattern of course, based on clinical diagnosis alone. Many coui'se and outcome variables may be influenced by factoi-s other than the characteristics usually considered in making clinical diagnoses. A number of such factoi-s were identified in the analysis of predictor of course and outcome. Some of the predictors (e.g., social isolation, marital status) point to the importance of psychosocial influences on the course and outcome of schizophrenic disorders, while othei's (e.g., type of onset, presence or absence of precipitating factors, psychosexual adjustment) suggest that a more reliable assessment of such characteristics in a conventional clinical evaluation can yield a more accurate prognosis. Because the usual predictors do not account for a veiy high proportion of the variance of the course and outcome of schizophrenia, especially when applied to patients in developing countries, it would seem that it is important to look for other potential predictors of outcome. Fresh approaches may be needed in the study of schizophrenia in cultures other than European or North American, and the simple transfer of knowledge derived from schizophrenia research in Europe and North America to other settings probably ceases to be meaningful beyond certain limits. The finding that schizophrenic patients in the centers in developing countries (particularly in Ibadan and Agra) had a better course and outcome on all variables than the schizophrenics in the centers in the developed countries is perhaps the most important outcome of the 2-year followup study. If confirmed by 5-year followup results, it would have significant theoretical and practical implications. As indicated earlier, a number of hypotheses can be formulated tentatively to explain these differences, and future work will focus on investigations to test them. First, it is conceivable that biologically different disorders with a different natural history may manifest a number of very similar symptoms, although subtle phenomenological distinctions may still be possible. Such distinctions have been proposed for example, between schizophreniform psychoses and true schizophrenia (Langfeldt 1939 and Welner and Stromgren 1958) or between reactive and process schizophrenia. The applicability of such distinctions to the series of IPSS patients is supported by the identification in all centers of remitting schizophrenic patients characterized by an acute onset, usually associated with a precipitating stress, and by such symptoms as derealization, depression, and elation. The presence of these patients could not explain the markedly better 2-year course and outcome of schizophrenics in the developing countries, however, since the prognostic differences remained when acute cases, which could be labeled "schizophreniform" psychoses, were excluded from the analysis. Second, one could consider the social or cultural environment as the possible key to understanding the observed prognostic differences between patients in developing and developed countries. Such a hypothesis is compatible with the findings that social isolation, marital status, educational level, and occupation were predictors of 2-year course and outcome. Unfortunately, a hypothesis of this kind is difficult to test on the basis of IPSS data because the study's design focused far less on social and cultural variables than on clinical ones. The results obtained so far, however, fully warrant a more intensive study of the relationship between the prognosis of schizophrenic psychoses and the sociocultural context. A 5-year followup of IPSS patients has been carried out and data analyses for this phase of the study are under way. These analyses will clarify whether prognostic differences in different cultures and diagnostic groups become greater, less, or remain the same over a long period of time. In addition, a new study aimed at replicating the findings on cross-cultural differences in the prognosis of schizophrenia and at testing some of the hypotheses about possible influences of culturerelated factors is being undertaken. Summary Results of the 2-year followup of the patients included in the International Pilot Study of Schizophrenia (World Health Organization 1973)

11 112 SCHIZOPHRENIA BULLETIN indicate that patients diagnosed schizophrenic on the basis of standardized assessments and clearly specified criteria demonstrated very marked variations of course and outcome over a 2-year period. Schizophrenic patients in the centers in developing countries had, on the average, considerably better course and outcome than schizophrenic patients in the centers in developed countries. Part of the variation of course and outcome was related to sociodemographic (e.g., social isolation and marital status) and clinical (e.g., type of onset and precipitating factors) predictors, but another larger part remained statistically unexplained. This suggests that variables usually used to describe psychopathology, the environment, and history of psychiatric patients in European and North American cultures may not account for cross-cultural differences. Clinical diagnosis on initial evaluation appeared to be a good predictor of subsequent symptomatology, but not of the length of the episodes, the total time during which the patient would be psychotic, pattern of course, or the degree of social impairment. A 5-year followup of the IPSS patients has also been completed, and the collected data are being analyzed. References Kraepelin, E. Vergleichende Psychiatrie. (1904) Translated by H. Marshall. In: Hirsch, S.R., and Shepherd, M., eda. Themes and Variations in European Psychiatry. Bristol: John Wright & Sons, pp Langfeldt, G. The Schizophreniform States. Copenhagen: Munksgaard, Raman, A.C., and Murphy, H.B.M. Failure of traditional prognostic indicators in Afro-Asian psychotics: Results of a long-term follow-up study. Journal of Nervous and Menial Disease, 164: , Vaillant, G.E. Prediction of recovery in schizophrenia. Journal of Nervous and Mental Disease, 135: , Welner, J., and Stromgren, E. Clinical and genetic studies on benign schizophreniform psychoses based on a follow-up. Ada Psychologica et Neurologica Scandinavica, 33: , Wing, J.K.; Cooper, J.E.; and Sartorius, N. The Measurement and Classification of Psychiatric Symptoms. London: Cambridge University Press, World Health Organization. Report of the International Pilot Study of Schizophrenia. Vol. 1. Geneva: World Health Organization Press, World Health Organization. Schizophrenia: A Multinational Study. (Public Health Papers No. 63) Geneva: World Health Organization Press, World Health Organization. Schizophrenia. An Internaiional Follow-up Study. New York: John Wiley & Sons, in press. Acknowledgment The work reported was sponsored by the World Health Organization and funded jointly by the World Health Organization, the National Institute of Mental Health, and the participating field research centers. The collaborating investigators in this study have been: Dr. N. Sartorius, principal investigator, Dr. T.-Y. Lin, former principal investigator, and Ms. E. M. Brooke, Dr. F. Engelsmann, Dr. G. Ginsburg, Mr. W. Gulbinat, Dr. A. Jablensky, Mr. M. Kimura, Dr. A. Richman, and Dr. R. Shapiro at the Headquarters of the World Health Organization in Geneva, Switzerland; Dr. E. Stromgren, chief collaborating investigator, and Drs. A. Bertelsen, M. Fisher, C. Flack, and N. Juel-Nielsen at the field research center in Aarhus, Denmark; Dr. K. C. Dube, chief collaborating investigator, and Dr. B. S. Yadav at the field research center in Agra, India; Dr. C. Leon, chief collaborating investigator, and Drs. G. Calderon and E. Zambrano at the field research center in Cali, Colombia; Dr. T. A. Lambo, chief collaborating investigator, Dr. T. Asuni, and Dr. M. 0. Olatawura at the field research center in Ibadan, Nigeria; Dr. J. K. Wing, chief collaborating investigator, and Drs. J. Birley and J. P. Leff at the field research center in London, United Kingdom; Dr. R. A. Nadzharov, chief collaborating in-

12 VOL. 4, NO. 1, vestigator, and Dr. N. M. Zharikov at the field research center in Moscow, USSR; Dr. L. Hanzlicek, chief collaborating investigator, and Dr. C. Skoda at the field research center in Prague, Czechoslovakia; Dr. C. C. Chen and Dr. M. T. Tsuang at the field research center in Taipei, China; and Dr. L. Wynne, Dr.. J. Strauss, and Dr. W. Carpenter, chief collaborating investigators, and Dr. J. Bartko at the field research center in Washington, D.C., USA. For a list of other staff contributing to the IPSS, see World Health Organization (1973). i aew nida publications The Authors Norman Sartorius, M.D., is Director, Division of Mental Health, World Health Organization, Geneva, Switzerland, Assen Jablensky, M.D., is Senior Medical Officer, Division of Mental Health, World Health Organization, Geneva, Switzerland. Robert Shapiro, M.D., is associated with the Massachusetts Mental Health Center, Boston, Mass. Single copies of the following publications may be obtained by writing the National Clearinghouse for Drug Abuse Information (NCDAI), NIDA, 5600 Fishers Lane, Room 10A-56, Rockville, Md : NIDA National Training System Course Calendar and Source Book, September 1977-June Provides those who work in substance abuse programs with a convenient source of current information on training : available through NIDA's National Training System. = Psychodynamics of Drug Dependence. A pioneering effort toward understanding the role of a-person's own psychodynamics in drug dependence. NIDA Research Monograph Series 12. Drug Treatment in New York City and Washington, D.C.* FoUowup Studies. NIDA Services Research Monograph Series. Can Drug Abuse Be Prevented in the Black Community? A look at the why's and wherefore's of drug abuse in the black community. First National Asian-American Conference on Drug Abuse Prevention, February 20-21, One in a series of publications which explore drug usage with the focus on this nation's ethnic minorities. This booklet examines the drug problem in the Asian-American community. Addicted Families and Their Children. A summary of a conference dealing with maternal and neonatal narcotic addiction, and treatment programs for pregnant drug dependent women and their children. An Investigation of Selected Rural Drug Abuse Programs. A report which explores and examines needs and concerns peculiar to the delivery of drug abuse treatment services within rural communities. Self-Concept and Drug Addiction: A Controlled Study of White Middle Socioeconomic Status Addicts. Technical paper. A Method for Estimating Heroin Use Prevalence. Technical paper. Guide to the Investigation and Reporting of Drug-Abuse Deaths. Problems and methods. Treatment program monograph. The Epidemiology of Drug Abuse. Research monograph. Drugs and Driving. Research monograph.