The government has highlighted the service needs of mentally disordered offenders (MDOs) and has introduced policies which have driven MDO needs assessment.

Based on a literature review and a national survey of agencies likely to be conducting needs assessment, the paper reviews government policy regarding MDO needs assessment and examines the particular problems of conducting needs assessment for MDOs.

A methodological framework for MDO needs assessment is presented, and different needs assessment approaches are discussed. The methodological and practical limitations of conducting MDO needs assessment are emphasised, as is the role of subjective judgement in both research design and interpretation.

Needs assessment is value-laden. Also, no single needs assessment approach can address all the issues of importance to purchasers and providers. Substituting the term ‘perspectives on need’ for the term needs assessment would help to reflect the lack of a single ‘truth’ about need and emphasise the role of subjective political judgement in designing and interpreting research on ‘need’.

Neuroimaging has proven useful in confirming diagnoses of certain neuropsychiatric conditions, but neuroimaging studies in learning disability are at an early stage.

A review of recent structural neuroimaging research in relation to learning disability was carried out.

Brain abnormalities can be detected in cases of idiopathic and non-idiopathic learning disability, but their significance is not clear due to discrepancies in study findings and the small cohorts involved.

Although the role of structural neuroimaging in the assessment of people with learning disability is not clear, new magnetic resonance imaging technology holds great promise for future research.

Despite growing evidence of the effectiveness of cognitive–behavioural therapy (CBT) for psychosis, typically only about 50% of patients show a positive response to treatment. This paper reports the first comprehensive investigation of factors which predict treatment outcome.

In a randomised controlled trial of CBT for medication-resistant psychosis (see Part I) measures were taken at baseline of demographic, clinical and cognitive variables. Changes over time were assessed on the Brief Psychiatric Rating Scale and the relationship between potential predictor variables and outcome was investigated using analysis of variance and covariance.

A number of baseline variables were identified as predictors of good outcome in the CBT group. Key predictors were a response indicating cognitive flexibility concerning delusions (P=0.005) and the number of recent admissions (P=0.002). Outcome was less predictable in the control group and was not predicted by any cognitive variable.

Good outcome is strongly predicted in patients with persistent delusions by a cognitive measure, while this was not the case in controls. Thus we argue that positive outcome in CBT is due in part to specific effects on delusional thinking.

Factors related to the outcome of depression and anxiety in primary care are not fully understood.

Adult patients in general practice with depressive, anxiety or panic disorder (n=148; DSM–III–R criteria) were studied prospectively for six months to determine the factors most closely associated with outcome. The Psychiatric Assessment Schedule, Hamilton Depression Rating Scale, Clinical Anxiety Scale and Life Events and Difficulties Schedule interviews were performed at index consultations and repeated six months later. Variables associated with outcome were assessed by multiple regression analysis.

Good outcome was predicted by mild depression at initial assessment, high educational level, and being in employment. At follow-up the most important predictor of improvement was reduction in marked difficulties over the six months. Recognition and management by the GP was most frequent in patients with severe disorder; such patients were least likely to improve because of the severity of their depression and marked social difficulties.

This naturalistic study helps to provide a framework for further studies with more precisely defined groups of people with depression. An effective treatment strategy for people with marked depression and ongoing social difficulties is especially needed.

The social origins of depression in people of Pakistani origin in the UK are not fully understood; they may not be the same as for the indigenous White population.

Patients attending a GP's surgery were screened with the Personal Health Questionnaire to detect depressive illness, which was confirmed using the Psychiatric Assessment Schedule; the estimated prevalence of major depression was 42%. Forty-four people with depression and 33 non-depressed comparison subjects were successfully interviewed using the Life Event and Difficulty Schedule, either in Urdu or English.

Seventy-five per cent of the sample were Pakistani born; 57% of cases and 33% of non-cases spoke English with difficulty. Seventy-five per cent of the cases and 36% of the non-cases had experienced an independent severe event and/or a major difficulty (P=0.0015). Difficulties associated with depression were most commonly in the marital, health and housing categories; overt racial harassment and discrimination were rare.

This is a preliminary study because of the small sample of GP attenders. The somatic presentation of depression in this population is clear. Acceptable forms of social support and help with marital, housing and employment problems will be needed to help relieve the depression in this population.

Previous studies show that criticism by relatives during acute depressive illness predicts relapse over nine months, but the course of criticism over time and its relation to the course of illness, have not been examined in depressed patients.

Thirty-nine depressed patients and their partners, were interviewed separately at three-monthly intervals over about one year. Illness was assessed by Research Diagnostic Criteria and Hamilton Depression Rating Scale. Criticism was measured by the Camberwell Family Interview.

Criticism when patients were most depressed predicted neither remission nor subsequent relapse. Patients who fully recovered, with or without later relapse, had partners who were consistently uncritical, or critical only at presentation. Patients with residual symptoms during remission had more persistently critical partners.

The nature of the association between criticism and depression could not be unequivocally established, although it appeared likely that continuing criticism was a result of continuing depression. Relatives' understanding of the illness may mediate between levels of criticism and the course of depression.

Serotonin receptors may play an important role in the pathophysiology of affective disorders. We studied type-2 serotonin (5-HT2) receptors in the brain of patients with major depression.

Using positron emission tomography (PET) and the selective radioligand [18F]altanserin, we investigated 5-HT2 receptor distribution in eight drug-free unipolar depressed patients and 22 healthy subjects. Data were analysed using Statistical Parametric Mapping 95.

In depressed patients, [18F]altanserin uptake was significantly reduced in a region of the right hemisphere including the posterolateral orbitofrontal cortex and the anterior insular cortex. A trend to similar changes was found in the left hemisphere. No correlation was found between the uptake and the Hamilton rating scale score.

Pathophysiology of depression may involve changes in 5-HT2 receptor in brain regions selectively implicated in mood regulation.

Currently the Mini Mental State Examination (MMSE) is widely used as a screening instrument for dementia syndrome. Diagnostic validity may be lowered in old age by normal age-related cognitive decline. Furthermore, visual impairment, occurring frequently in old age, leads to missing values which prevent an interpretation of the test result.

In the Berlin Ageing Study (n=516, age range 70–103 years) MMSE and clinical dementia diagnosis, made by a psychiatrist investigating all subjects by the Geriatric Mental State–A and History and Aetiology Schedule interviews, were investigated independently. The MMblind was analysed, an MMSE version for vision impairment in which all items requiring image processing are omitted. The study sample is population-based; dementia cases (DSM–III–R) were excluded on the basis of the clinical diagnosis.

Norms are reported for very old age regarding MMSE as well as MMblind. There is a considerable age effect on MMSE scores. In contrast to MMSE, sensitivity and specificity of the shorter MMblind version are not reduced.

The considerable age effect requires the adaptation of cut-off values for old age. The blind version of the MMSE seems to be a valid instrument improving the applicability of the MMSE in old age.

We assessed the publication trends of papers on schizophrenia through an analysis of the articles published by three general psychiatric journals (Archives of General Psychiatry (AGP), the British Journal of Psychiatry (BJP), and the Australian and New Zealand Journal of Psychiatry (ANZJP) from three continents.

For each of the journals, we covered the period between 1980 and 1994. We carried out both a quantitative analysis, assessing the trends over time in the publication of papers on schizophrenia, and a qualitative analysis, classifying the articles into eight scientific fields.

During the study period a total of 943 articles dealing with schizophrenia were published in the three journals evaluated; the proportion of papers focusing on schizophrenia was higher in the AGP and in the BJP (18 and 15%) as compared with the ANZJP (5.6%). A substantially higher proportion of basic science articles was published in the AGP as compared with the BJP and the ANZJP, whereby a somewhat larger representation of epidemiological and psychosocial research was found in the latter journals.

Given the importance of schizophrenia in psychiatric research and practice, it will be useful to regularly monitor the research trends in this specific field.

Recent findings indicate that the established association between social indicators of deprivation and psychiatric admission rates may not hold across all diagnoses.

Admission rates in individuals aged 16–64 years for 71 electoral wards in North Staffordshire were calculated for six diagnostic groups using data from the Korner Episode System for 1987–1993. These were correlated with selected individual census variables, Townsend and Jarman indices. The ability of regression models to predict admission rates was tested.

The strongest correlations were found for total admissions (r 0.44–0.79). Strong correlations were found for neurotic disorders/depression (r 0.29–0.62), schizophrenia (r 0.24–0.59), all non-psychotic disorders combined (r 0.41–0.71) and all psychotic disorders combined (r 0.33–0.67). Predicted admission rates for total admissions, psychotic and non-psychotic admissions using regression models showed strong positive correlations with observed admission rates.

The strong correlations between social indicators of deprivation and total psychotic admission rates are consistent with the results of previous studies. The strong associations between social indicators and admissions for non-psychotic disorders is contrary to previous findings and may be partly explained by the relatively high admission rates for neurotic disorders.

Information on suicide in ethnic and immigrant groups in England and Wales is limited.

A three-year (1991–1993) survey was conducted of all unnatural deaths of residents of an urban area. ‘True likely’ and official’ age-adjusted suicide rates were compared by ethnicity and, for Whites, birthplace.

Irrespective of verdict, 329 likely suicides were identified. Relatively few ethnic minority and White immigrant suicides had received a suicide verdict. Afro-Caribbeans had relatively low, and young Indian women relatively high suicide rates. Rates of Scottish- and Irish-born residents were 2.1 to 2.9 times higher than the local base rate. Young White males' rates were higher than those of the elderly.

Classification of suicide is biased with respect to ethnicity and national origin. Rate patterns for ethnic minority groups reflect patterns seen in attempted suicide. In this deprived area, young White male suicide rates have surpassed those among the old.

To study driver suicides, all motor vehicle driver fatalities in Finland from 1987 to 1991 were analysed.

Cases were collected from all driver fatalities (n=1419), as those that accident investigation teams considered possible suicides (n=99). Drivers committing suicide classified according to ICD–9 by two forensic pathologists were selected as cases (n=84). Drivers of unintentional motor vehicle fatalities served as the control group.

While 5.9% of all driver fatalities were classified as suicides, the figure given in the official statistics was 2.6%. Driver suicides accounted for 1.2% of all suicides. Fifty per cent of driver suicides were committed by men aged between 15 and 34 years. The victims had often suffered from life-event stress, mental disorders and had alcohol misuse problems. The cases were usually head-on collisions between two vehicles with a large weight disparity.

Misclassification of driver suicides does not significantly influence the total suicide rate. However, people who commit suicide by this method could often be recognised, and prevention of these events would improve traffic safety.

We aimed to study the effect of weather conditions on the rate of all deaths due to fatal self-harm (FSH) reported in a population of 350 000. In this study the term FSH refers to all coroners' verdicts which were recorded in a specified period and included suicide, misadventure and open verdict.

Five-year data of FSH in North Cheshire was analysed in relation to meteorological data, which were measured at the nearest meteorological office to the study population.

Small but significant positive correlations were demonstrated between occurrence of FSH and hours of sunshine (P < 0.01) and diurnal variations in daily temperature (P < 0.05). These effects were independent of gender and psychiatric history.

Weather may influence the occurrence of FSH, probably interacting with biological and social variables. The rate of FSH death appears to be positively related to fine weather conditions, during early summer, and not to extreme weather conditions as previously reported.

Post-traumatic stress disorder is a well recognised condition which can have long-term consequences. A rail accident to commuter trains in south London provided an opportunity, within the context of a Psychological Therapies Service, to study the prevalence and associations of the condition in a follow-up study.

Postal questionnaires were completed by 187 survivors (52% response rate) 10 to 22 months after the accident, and their symptoms were compared with a commuter control group.

The accident group scored significantly higher on the Impact of Events scale (IES) than the control group. Severity of physical injury was significantly associated with high IES scores. Thirty per cent of those moderately or severely injured showed high levels of intrusion or avoidance symptoms (sub-scores of 20 or more). Thirty per cent of those nearest the accident focus had high sub-scores, associated variables being the accident experience of risk of death, feeling trapped and seeing death. For those less directly involved the proportions with symptoms were similar to those of controls.

It is recommended that disaster services should provide for long-term psychological effects, particularly for those survivors who are injured, who experience risk of death, feeling trapped or who witness death.