Background. Mental health service research continues to use only outcome measures that are available rather than develop measures that are important. This paper argues that it is necessary to select and then define a set of ethical principles that can be operationalized and validated as outcome measures to provide a wider balance of information for health policy and clinical service decisions.

Methods. The method used is to adopt a five stage procedure: (i) to select ethical principles most directly relevant for mental health services and their evaluation at the local level; (ii) to propose definitions of these principles; (iii) to validate these definitions; (iv) to translate the defined principles into operationalized outcome measures; and (v) to use these outcome measures in mental health services research, within the context of evidence-based medicine.

Results. We address steps (i) and (ii) of this five-stage procedure. Nine principles are selected and defined: autonomy, continuity, effectiveness, accessibility, comprehensiveness, equity, accountability, coordination and efficiency. These principles can together be referred to as the three ACEs.

Conclusions. Of these nine principles, only two (effectiveness and efficiency) have so far been fully translated into quantitative outcome measures, upon which the evidence-based medicine approach depends. We propose that further concepts also be developed into a more complete multi- dimensional range of fully operationalized outcome measures.

Background. There is some uncertainty about the rate and correlates of autism.

Method. Twenty-three epidemiological surveys of autism published in the English language between 1966 and 1998 were reviewed.

Results. Over 4 million subjects were surveyed; 1533 subjects with autism were identified. The methodological characteristics of each study are summarized, including case definition, case-finding procedures, participation rates and precision achieved. Across surveys, the median prevalence estimate was 5·2/10000. Half the surveys had 95% confidence intervals consistent with population estimates of 5·4–5·5/10000. Prevalence rates significantly increased with publication year, reflecting changes in case definition and improved recognition; the median rate was 7·2/10000 for 11 surveys conducted since 1989. The average male/female ratio was 3·8[ratio ]1, varying according to the absence or presence of mental retardation. Intellectual functioning within the normal range was reported in about 20% of subjects. On average, medical conditions of potential causal significance were found in 6% of subjects with autism, with tuberous sclerosis having a consistently strong association with autism. Social class and immigrant status did not appear to be associated with autism. There was no evidence for a secular increase in the incidence of autism. In eight surveys, rates for other forms of pervasive developmental disorders were two to three times higher than the rate for autism.

Conclusion. Based on recent surveys, a minimum estimate of 18·7/10000 for all forms of pervasive developmental disorders was derived, which outlines the needs in special services for a large group of children.

Background. Lifetime rates of depression reported in epidemiological surveys are generally only twice the 12 month rates. Either people forget the symptoms of depression or many people who have a depressive episode remain depressed for many years. Both may be true. There is a need to examine the long-term clinical validity of interviews that are used to make lifetime diagnoses.

Methods. Forty-five patients who were part of a long-term follow-up study of depression were interviewed 25 years after the index episode. The diagnoses from the original, fully structured interviews were compared with the responses people made for that period when interviewed using the CIDI 25 years later.

Results. Twenty-seven patients met CIDI DSM-III-R criteria for depression at index episode. At the 25 year follow-up, 19 of the 27 reported the essential symptoms of ‘depression or loss of interest’ being present at the index time, and in 14 of the 27 the depressive symptoms recalled met criteria for DSM-III-R major depressive episode at that time.

Conclusions. Seventy per cent of people who were hospitalized for a major depressive episode can recall being depressed but only half can recall sufficient detail to satisfy the diagnostic criteria when interviewed 25 years later. As depressive episodes, especially those severe enough to warrant admission, are recalled better than many other diagnoses, one must be cautious about the lifetime rates for mental disorders reported in retrospective epidemiological surveys.

Background. It has been suggested that giving people the opportunity talk about a traumatic experience may prevent the development of later disorder. We tested the efficacy of two brief interventions, education and psychological debriefing, designed to prevent adverse psychological reactions to criminal victimization.

Methods. Individuals who had been the victims of a violent crime within the past month were written to and invited to take part in a study of their attitudes to crime and punishment: 2161 were contacted and 243 replied, of whom 157 were eligible and were randomly assigned either to an education condition, to a psychological debriefing plus education condition, or to an assessment only condition. Education involved providing information about normal post-traumatic reactions. Debriefing involved in-depth probing about events, thoughts and feelings experienced during the crime. Subjects were recruited from police and hospital sources and interviewed in their own homes: 138 were followed up at 6 months, and 92 at 11 months.

Results. Outcome was assessed using a DSM-III-R diagnosis of PTSD, the Post-traumatic Symptom Scale, the Impact of Event Scale and the Beck Depression Inventory. All groups improved over time but there were no between-group differences.

Conclusions. No evidence was found to support the efficacy of brief one-session interventions for preventing post-traumatic symptoms in individual victims of violent crime.

Background. Expressed emotion (EE) is a measure that has been used to assess the quality of the relationship between patient and their key relative. It has been shown to be strongly predictive of clinical outcome in a range of psychiatric and medical disorders. This study investigated the effect of EE on treatment outcome in chronic post-traumatic stress disorder (PTSD).

Methods. A prospective design was adopted. The key relatives of 31 PTSD patients participating in a treatment trial comparing imaginal exposure with cognitive therapy were interviewed and rated on EE prior to treatment allocation. The effect of EE on post-treatment clinical outcomes was assessed.

Results. Sixteen patients (52%) had high EE and 15 (48%) low EE relatives. Patients with high EE relatives showed lesser change scores on the main outcome variable of the trial, the total CAPS score, and on all the secondary outcome variables than those with low EE relatives. Using different multiple regression models the EE scales of criticism and hostility predicted just under 20% of the outcome variance. These two scales were highly correlated and criticism marginally predicted the greatest variance (19·7%).

Conclusions. The results highlight the importance of the quality of the patient's social environment in influencing their response to cognitive and behavioural treatments.

Background. We examine potential sources of the sex differences in post-traumatic stress disorder (PTSD) in the community.

Methods. Data were obtained from a representative sample of 2181 persons aged 18–45 years in the Detroit primary metropolitan statistical area, which is a six-county area containing more than four million residents. A random digit dialling method was used to select the sample and a computer-assisted telephone interview was used to obtain the data. DSM-IV PTSD was assessed with respect to a randomly selected trauma from the list of qualifying traumas reported by each respondent.

Results. The lifetime prevalence of exposure and the mean number of traumas were lower in females than males. The overall conditional risk of PTSD (i.e. the probability of PTSD among those exposed to a trauma) was approximately twofold higher in females than males, adjusting for the sex difference in the distribution of trauma types. The sex difference was due primarily to females' greater risk following assaultive violence. The sex difference in the avoidance and numbing symptom group following assaultive violence exceeded the sex differences in other symptom groups.

Conclusions. Future research should focus on sex differences in the response to assaultive violence, including potential explanations for females' greater probability to experience avoidance and numbing.

Background. This study aimed to assess psychiatric morbidity and to collect information on disability, life events and family support in a representative sample of patients admitted to a general hospital.

Methods. On the basis of information collected in a pilot study a systematic sample of patients consecutively admitted to seven general medical and seven surgical wards of the Academic General Hospital of Verona was selected and interviewed using a two-phase screening procedure and standardized instruments (GHQ-12, HADS, BDQ and CIDI-PHC). All data were analysed using appropriately weighted logistic regression procedures.

Results. A total of 1039 patients completed the GHQ-12 and 298 (28·7%) were high-scorers: 363 patients were interviewed with CIDI-PHC. The prevalence of ICD-10 cases was 26·1%. The most common psychiatric diagnoses were current depression (12·8%) and generalized anxiety disorder (10·8%), followed by alcohol related disorders (5%). A higher prevalence of ICD-10 cases was found in medical wards, among females, patients older than 24 years, unemployed and separated/divorced people. Life events were associated with psychopathology, and so was the number of disability days. Although 49·8% of ICD-10 cases were identified by the hospital doctors as having a psychological disorder, 23·1% of ICD-10 cases were referred to the liaison psychiatric service.

Conclusion. The results of the present study stress the need to collect epidemiologically-based data on psychological disorders and their recognition not only in general practice, but also in general hospital settings, in order to have a more complete picture of the pathways to specialist care.

Background. We developed a comprehensive, 20-hour training programme for primary-care physicians, that sought to improve their ability to detect, diagnose and manage depression. We evaluated the effects of physician training on patient outcomes, using a pre-post design.

Methods. In the pre-training phase of the study, we sampled 1834 consecutive patients of 17 primary-care physicians and evaluated 518 of these patients for the presence of depression. We measured outcomes of all patients with depression at 3 months and 1 year. The outcome measures were: severity of psychopathology; duration of depressive episode; and level of daily functioning. After the 17 physicians completed the training, we drew a new sample from their practices (498 of 1785 consecutive patients were evaluated for depression) and measured outcomes for the depressed patients.

Results. We found an effect of the training on short-term outcome, particularly for patients with a recent-onset depression. At 3-month follow-up depressed patients whose physicians had received training had less severe psychopathology and patients with recent-onset depression also showed higher levels of daily functioning than patients of the same physicians prior to the training. The patients with a recent-onset depression that was recognized by trained physicians had shorter depressive episodes, but this was not statistically significant. At 1-year follow-up, all training effects had faded away.

Conclusions. Training primary-care physicians to recognize, diagnose and manage depression can improve short-term patient outcomes, especially for patients with a recent onset of depression. Patients suffering from a recurrent or chronic depression may need more specific interventions, both for acute treatment and long-term management.

Background. While cross-sectional and longitudinal studies have consistently found depressive illness and disability to be related, understanding whether depression leads to subsequent onset of disability is limited.

Methods. In the context of the multi-centre international WHO Collaborative Study on Psychological Problems in General Health Care, we followed prospectively consulting non-elderly primary care patients who were essentially disability free at baseline but who differed in baseline depression status, comprising 1051 patients free of physical disability at baseline including 14% depression; 914 free of social disability including 9% depression. Depression status was assessed with the CIDI; patient-reported physical disability with the MOS physical functioning scale and social disability with the BDQ role functioning and number of disability days measures; investigator-rated social disability with the Occupational section of Groningen Social Disability schedule; and the treating physicians rated the severity of physical illness.

Results. In patients essentially disability free at baseline, depressive illness resulted in a 1·5-fold (at 3 months) and a 1·8-fold (at 12 months) increase in risk of onset of physical disability, after controlling for physical disease severity. Depressive illness also resulted in a 2·2-fold (at 3 months) and a 2·3-fold (at 12 months) increase in risk of onset of social disability, after controlling for physical disease severity, physical disability and onset of physical disability.

Conclusions. Among non-elderly primary care patients, depressive illness is associated with onset of physical disability and shows an even stronger association with onset of social disability.

Background. Depression, anxiety and fatigue are among the most common symptoms presented in primary care. Whether such symptoms indicate discrete psychological syndromes or whether they result from a common vulnerability is not clear. This study examined longitudinally the patterns of co-morbidity between prolonged fatigue and other forms of psychological distress in patients attending general practitioners.

Methods. Adults attending primary care completed questionnaires designed to detect cases of prolonged fatigue and psychological distress at presentation and 12 months later.

Results. Of 652 patients, the prevalence rates of ‘prolonged fatigue’ alone, ‘psychological distress’ alone, ‘prolonged fatigue+psychological distress’ and ‘no disorder’ were 7%, 19%, 15% and 59% respectively at initial assessment. Of those patients with any prolonged fatigue syndrome initially, 58% still reported fatigue 12 months later (representing 13% of the total sample). Most importantly, the risk of developing prolonged fatigue was not increased in patients who initially had psychological distress (OR = 0·95; 95% CI 0·2–3·6), neither was the risk of developing psychological distress increased in patients who initially had prolonged fatigue (OR = 1·4; 95% CI 0·6–3·4).

Conclusions. This study indicates that prolonged fatigue is a persistent diagnosis over time. The longitudinal patterns of co-morbidity with psychological distress do not support an aetiological model that proposes a common vulnerability factor for these disorders. Psychiatric classification systems may be better served by treating prolonged fatigue and psychological distress as independent disorders.

Background. Fatigue and psychiatric symptoms are common in the community, but their association and outcome are sparsely studied.

Method. A total of 1177 patients were recruited from UK primary care on attending their general practitioner. Fatigue and psychiatric disorder was measured at three time points with the 12-item General Health Questionnaire and the 11-item Fatigue Questionnaire.

Results. Total scores for fatigue and psychiatric disorder did not differ between the three time points and were closely correlated (r around 0·6). The association between non-co-morbid (‘pure’) fatigue and developing psychiatric disorder 6 months later was the same as that for being well and subsequent psychiatric disorder. Similarly, having non-co-morbid psychiatric disorder did not predict having fatigue any more than being well 6 months previously. Between 13 and 15% suffered from non-co-morbid fatigue at each time point and 2·5% suffered from fatigue at two time points 6 months apart. Less than 1% of patients suffered from non-co-morbid fatigue at all three time points.

Conclusions. The data are consistent with the existence of ‘pure’ independent fatigue state. However, this state is unstable and the majority (about three-quarters) of patients become well or a case of psychiatric disorder over 6 months. A persistent, independent fatigue state lasting for 6 months can be identified in the primary-care setting, but it is uncommon – of the order of 2·5%. Non-co-morbid (pure) fatigue did not predict subsequent psychiatric disorder.

Background. The literature documents that functional impairment is associated with affective disorders. Nevertheless, the choice among thorough, yet brief, well-validated assessments of functional impairment is limited. The objective of this study was to evaluate the psychometric properties of a brief scale of functional impairment, the Range of Impaired Functioning Tool (LIFE–RIFT).

Method. The study sample included subjects who presented with major depressive disorder at intake into the NIMH Collaborative Depression Study (CDS). The LIFE–RIFT is composed of items that are included in the Longitudinal Interval Follow-up Evaluation (LIFE). The reliability and validity were examined using data from LIFE–RIFT assessments conducted at four points in time: 6, 12, 18 and 24 months after intake into the CDS.

Results. Cross-sectional one factor models accounted for the covariance structure among the four scale items. A longitudinal factor model, with an invariant factor structure over time, also fitted the data well and indicated that the scale items are measures of one construct, namely functional impairment. The internal consistency reliability of the scale was supported with alpha coefficients ranging from 0·81 to 0·83. The inter-rater reliability intraclass correlation coefficient (ICC) was 0·94. Mixed-effect linear regression models showed that those in episode were significantly more impaired than those in recovery. Furthermore, in analyses of predictive validity, impairment was positively associated with subsequent recurrence and negatively associated with subsequent recovery.

Conclusions. This psychometric evaluation provides empirical support for the reliability and validity of the LIFE–RIFT, a brief measure of functional impairment.

Background. Scales to measure the severity of different dimensions of auditory hallucinations and delusions are few. Biochemical and psychological treatments target dimensions of symptoms and valid and reliable measures are necessary to measure these.

Method. The inter-rater reliability and validity of the Psychotic Symptom Rating Scales (PSYRATS: auditory hallucination subscale and delusions subscale), which measure several dimensions of auditory hallucinations and delusions were examined in this study.

Results. The two scales were found to have excellent inter-rater reliability. Their validity as compared with the KGV scale (Krawiecka et al. 1977) was explored.

Conclusions. It is concluded that the PSYRATS are useful assessment instruments and can complement existing measures.

Background. For adolescents, there is no specific needs assessment instrument that assesses significant problems that can benefit from specified interventions. A new instrument (S.NASA) was developed by incorporating and adapting three well established adult needs assessment instruments. The S.NASA covers 21 areas of functioning including social, psychiatric, educational and life skills.

Method. Client and carer interviews were conducted by different researchers. A week later the interviews were repeated using a crossover design. Significant (cardinal) problems were generated from the clinical interviews using a pre-defined algorithm. Final need status (three categories) was made by clinicians assessing the cardinal problems against defined interventions. The interventions were generated from discussions with clinicians and a survey of appropriate professionals working with adolescents.

Results. Pre-piloting led to the final version being administered to 40 adolescents from secure units, forensic psychiatric and adolescent psychiatric services. There were 25 males and 15 females, mean age 15·5 years. Overall there were moderate to good inter-rater and test–retest reliability coefficients, the test–retest reliability coefficients for the total scores on the needs assessment interviews ranged from 0·73 to 0·85. Consensual and face validity was good, the adolescents and staff finding the instrument useful and helpful.

Conclusions. This new needs assessment instrument shows acceptable psychometric properties. It should be of use in research projects assessing the needs and the provision of services for adolescents with complex and chronic problems.

Background. Schizophrenia patients, particularly those with symptoms such as thought insertion, passivity experiences and hallucinations, may share an underlying cognitive deficit in monitoring the generation of their own thoughts. This deficit, which has been referred to as ‘autonoetic agnosia’, may result in the conclusion that self-generated thoughts come from an external source. Previous work supports this notion, yet the statistical approaches that have been used have not enabled a distinction between specific deficits suggesting autonoetic agnosia and more general cognitive dysfunction.

Methods. Autonoetic agnosia was assessed using source-monitoring paradigms in 28 patients with schizophrenia and 19 control subjects. Multinomial model analyses, which allow the distinction between deficits in recognizing information, remembering its source, and response biases, were applied to the data.

Results. Schizophrenia patients were impaired in discriminating between words that came from two external sources, from two internal sources, and one internal and one external source. In a condition requiring subjects to distinguish between words they had heard from those they had imagined hearing, when schizophrenic patients did not remember the source of the information, they showed a stronger bias than controls to report that it had come from an external source.

Conclusions. The application of multinomial models to source monitoring data suggests that schizophrenia patients have source monitoring deficits that are not limited to the distinction between internally-generated and externally-perceived information. However, when schizophrenia patients do not remember the source of information, they may be more likely than controls to report that it came from an external source.

Background. Previous research has mainly focused on the cross-sectional structure of symptoms in schizophrenia. This meta-analysis examined the association of duration of illness with the structure of symptoms.

Methods. Using explicit criteria, 22 studies reporting on the correlations of symptoms in 2665 schizophrenic patients were selected. From each study, symptom-pair correlations for negative symptoms as rated by Scale for the Assessment of Negative Symptoms (SANS) and positive symptoms as rated by the Scale for the Assessment of Positive Symptoms (SAPS) were extracted. Variability among symptom-pair correlations across studies was assessed using tests of homogeneity. For symptom-pair correlations which were not found to be homogeneous, the association of average duration of illness with the symptom-pair correlations were examined.

Results. There was considerable variability in symptom-pair correlations across studies. Part of this variability was explainable by variations in average duration of illness. Longer duration of illness was associated with lower negative–negative symptom-pair correlations and higher negative–positive symptom-pair correlations.

Conclusions. The findings suggest that the structure of symptoms in schizophrenia evolves over time, following a consistent pattern. In the early stages of illness, negative and positive symptoms form cohesive dimensions. With time, these dimensions become less cohesive and the boundaries between them, less clear.

Background. The study explored the genetic and environmental risk factors for both the behaviours and attitudes characteristic of disordered eating.

Methods. In three waves of data collection, information was collected from female twins regarding their eating and attitudes towards eating, weight and shape. The first assessment consisted of a self-report questionnaire (1988–9) with 1682 women. The second assessment consisted of a semi-structured psychiatric interview schedule (1992–3), completed by 1852 women, many of whom had completed Wave 1 assessment. The third assessment, with 325 women chosen from Waves 1 and 2 (1995–6), consisted of a semi-structured interview (the Eating Disorder Examination).

Results. As only one twin pair was concordant for lifetime bulimia nervosa at Wave 3 assessment, ordinal measures of all assessments were used in a multivariate genetic analysis. Results indicated that additive genetic and non-shared environmental influences best explained variance in liability to disordered eating, with about 60% (95% CI 50–68) of the variance explained by genetic factors. Comparison with a model allowing for the effects of shared environment indicated genetic factors accounted for a similar degree of variance (59%, 95% CI 36–68).

Conclusion. Liability to the development of the behaviours and attitudes characteristic of eating disorders is best explained by both environmental and genetic factors, with covariation between the three measures best explained by a single latent phenotype of disordered eating which has a heritability of 60%.

Background. Empirical data regarding prevalence of sexual and physical abuse histories in Japanese patients with eating disorders is lacking, in contrast to Western countries. This study investigated the prevalence of traumatic events in Japanese patients with eating disorders, and examined the relationship between such traumatic events and clinical features.

Methods. Subjects consisted of 33 patients with anorexia nervosa restricting type (RAN), 40 patients with anorexia nervosa binge eating/purging type (AN–BP), 63 patients with bulimia nervosa purging type (BN) and 99 healthy controls. All were female and diagnoses were based on DSM-IV. The Physical and Sexual Abuse Questionnaire (PSA), Eating Disorder Inventory (EDI) and Dissociation Experience Scale (DES) were administered to all of the subjects.

Results. Paradoxically, victims of minor sexual abuse committed by Chikan (a Japanese word indicating a person who commits minor sexual crimes) were more prevalent among controls than among patients with RAN, AN–BP or BN. However, physical punishment histories tended to be more prevalent among patients with AN–BP or BN than among RAN or controls. Only AN–BP and BN patients with physical punishment histories had twofold higher scores for DES and significantly more frequent histories of self-mutilation (67% v. 33%) compared with patients without such histories.

Conclusion. An abuse history is not essential or a prerequisite to developing an eating disorder in Japan.

Background. Current systems of describing personality pathology have significant shortcomings. A polydiagnostic approach is used to study the relationship between psychological, psychoanalytical and psychopathological models of personality.

Methods. The subjects were 256 patients enrolled in treatment studies of major depression and bulimia nervosa. Subjects were assessed using the Temperament and Character Inventory (TCI), the Defense Style Questionnaire (DSQ) and the Structured Clinical Interview for DMS-III-R personality disorders (SCID-II).

Results. Subjects had high rates of DSM-III-R personality disorders with 52% having at least one personality disorder. Cluster A personality disorders were correlated with low reward dependence, high harm avoidance and low self-directedness and cooperativeness. Cluster B personality disorders were related to high novelty seeking and low self-directedness and cooperativeness. Cluster C personality disorders were correlated with high harm avoidance and low novelty seeking and low self-directedness. Immature defences were related to DSM-III-R personality symptoms, but individual defences were not related to personality clusters in a predictable way. Immature defences were strongly related to low self-directedness and cooperativeness. Both TCI self-directedness scores and immature defence scores were moderately predictive of the presence and number of personality disorders.

Conclusion. A widely accepted clinical nosology (DSM-III-R personality disorders) rated using a clinical interview correlates reasonably predictably with two theoretical models derived from different paradigms and rated using self-reports. This might be seen as providing concurrent validity for all three models. However, serious methodological shortcomings confront studies of this type, including sample selection and measurement of personality dysfunction. One way to begin to resolve these problems is to study which personality measures are best related to treatment response and prognosis.

Background. Previous studies on male patients with sex chromosome abnormalities (SCA), namely XYY and XXY, suggest that such patients commit criminal acts more frequently than expected. Most of these studies are affected by ascertainment bias.

Methods. Using a population-based sample of men with SCA, identified by screening 34380 infants at birth between 1967 and 1979, comparison between 16 XYY men, 13 XXY men and 45 controls were made in terms of frequency of antisocial personality disorder (APD) using the Schedule for Affective Disorders and Schizophrenia lifetime version. Rates of criminal convictions were examined in 17 XYY men, 17 XXY men and 60 controls.

Results. XYY males showed a significantly higher frequency of antisocial behaviour in adolescence and adulthood and of criminal convictions than the controls, but multiple regression analysis showed this to be mediated mainly through lowered intelligence. Property offences constituted the majority of offences in all groups. The XXY men did not show an increased rate of criminal convictions. It is possible that this apparently negative result relates to the relatively small numbers of cases and hence low power of this study.

Conclusions. The findings of this study carry the advantage of not being affected by ascertainment bias and the disadvantage of having low power. It provides evidence for a slightly increased liability to antisocial behaviour in XYY men.