The aim was to examine suicides by burning (SBB) occurring in England and Wales.

The method was a retrospective case note examination of death certificates, coroners' inquest notes and GP notes for all SBBs registered in 1991 by OPCS in England and Wales, and an examination of SBB statistics from 1979–1992.

In 24% of cases the incident involved others either as intended or actual victims or as witnesses, and three distinct groups (Asian-born women, schizophrenics, and homicide–suicides) were over-represented. The majority were not in contact with current psychiatric services, and 43% had never had contact with psychiatric services. There was no trend in SBB compared to other suicides in 1979–1992.

While SBB may be associated with Asian-born women, homicide–suicide and schizophrenia, the majority of cases did not have these characteristics, and the overall psychiatric contact rate was similar to suicide by unspecified means.

The findings from a controlled study of the effect of a public education campaign on community attitudes to mentally ill people are presented.

A census of neighbours' attitudes toward mental illness was conducted in two areas before the opening of supported houses for the mentally ill. In one area an educational campaign was conducted The attitude survey was then repeated in both areas and patients' social contact with neighbours was recorded.

Respondents exposed to the didactic component of the campaign showed only a small increase in knowledge about mental illness but there was a lessening of fearful and rejecting attitudes in the experimental area and not in the control area Neighbours in the experimental area were more likely to make social contact with both staff and patients. It was social contact which was directly associated with improved attitudes rather than education per se. Patients in the experimental area made contact and even friendships with neighbours whereas those in the control area did not.

The public education campaign did not lead to significant changes in neighbours' knowledge of mental illness. However, their attitudes improved and patients' social integration was enhanced.

The features of adolescents who had taken an overdose were assessed to determine the focus for a treatment trial.

Overdose cases were compared with psychiatric and community controls who had not taken an overdose in respect of mental disorders and family background.

Overdose cases had high rates of major depression, but most of them recovered from depression within six weeks of the overdose. There was a specific association between taking an overdose and family dysfunction.

Family dysfunction could be a useful focus in a clinical trial of the aftercare of adolescents who have taken an overdose.

The baseline findings from a controlled study of the effect of a public education campaign on community attitudes to mental illness are presented.

A census of attitudes to mental illness was conducted in two areas, prior to the opening of supported houses for the mentally ill Factor analysis of the Community Attitudes toward the Mentally III (CAMI) inventory revealed three components: Fear and Exclusion, Social Control and Goodwill.

The only determinant of Fear and Exclusion was having children. The main determinants of Social Control were social class, ethnic origin, age, having suffered mental illness and having children. The main determinant of Goodwill was educational level The attitude factors were predictive of respondents' behavioural intentions toward the mentally ill. Respondents with children and non-Caucasians were more likely to object to the mentally ill living in their neighbourhood.

Any intervention aimed at changing attitudes to mentally ill people in the community should be targeted at people with children and non-Caucasians, as these groups are more likely to object.

There is a paucity of research into the relationship between insight and variables including cognitive function, educational background and symptomatology in schizophrenia.

Sixty-four patients with DSM–III–R diagnosis schizophrenia were assessed with the Schedule for Assessment of Insight, Mini Mental State cognitive test, a new measure of knowledge about treatment (the Understanding of Medication Questionnaire), and educational background and compliance assessments.

Insight scores correlated significantly with a range of variables. In a multiple regression analysis only knowledge about treatment and number of years in education explained a significant proportion of insight.

Educational background and the intellectual ability required to learn complex concepts, such as models of mental illness, appear to be more important than previously considered. Research is limited by the lack of a generally accepted definition of insight.

Findings that the EE level of a relative may change over time support the idea that EE may represent the circumstances of the relationship between patient and caregiver. The present study examines to what extent EE levels in relatives are related to relatives' burden of care and their perceptions of patients' deficits in social role performance.

Fifty patients recently admitted to hospital with DSM–III–R diagnoses of schizophrenia or schizophreniform disorder were assessed for positive and negative symptoms. Fifty relatives who were living or were in close contact with these patients were interviewed for the assessment of EE and burden of care, and to provide information about patients' social role performance and social and behaviour problems.

High-EE relatives had considerably higher mean scores for burden of care than low-EE relatives (12.5 v. 6.8, respectively, P=0.002), and perceived more deficits in patients' social functioning than low-EE relatives (means: 16.2 v. 6.9, respectively, P=0.004). The employment status of relatives was the only socio-demographic characteristic of relatives and patients associated with EE levels, those who were working being less likely to be high EE. Patients' psychopathology was not associated with EE levels and burden of care.

This study shows that EE and the burden of care are related. EE and burden both measure aspects of the relationship between relatives and patients. These findings suggest that EE and burden of care are more dependent on relatives' appraisal of the patient condition than on patients' actual deficits.

Paracetamol is now the most common drug used for self-poisoning in the UK and is associated with potentially fatal liver damage. Patients admitted to hospital because of paracetamol overdoses were studied in order to determine their characteristics and factors which might have deterred them from taking paracetamol or reduced the dangers of the overdose.

Eighty patients were studied in hospital using a structured interview schedule, measures of depression and suicidal intent, information collected through the Oxford Monitoring System for Attempted Suicide, and the results of liver function tests.

Acute liver dysfunction (25 patients) was associated with consumption of more than 25 tablets (odds ratio 4.46, 95% CI 1.31 to 17.41, P = 0.014). The proportionate use of tablets from blister packs (60%) and loose preparations (46%; 5 patients using both types) reflected their general availability. More of those who took tablets from a loose preparation consumed 25 or more tablets (69%) than those who used a blister-pack preparation (40%; odds ratio = 3.0, 95% CI 1.12 to 9.95, P = 0.028). Only 20 patients thought that any type of warning label would have deterred them from taking a paracetamol overdose.

Establishing a maximum number of tablets (e.g. 25) that can be available in individual preparations is likely to reduce the dangers of paracetamol self-poisoning. The potential effects of other measures are uncertain.

The reduction of beds in long-stay hospitals has led to concerns over the quality of care offered to the remaining residents as well as that provided in the community. This study seeks to compare the quality of care and quality of life (reported satisfaction) from residents in both types of setting.

A cross-sectional comparison was made of community residential homes and hospital wards drawn randomly from lists provided by local authorities in the outer London area. Samples were drawn from all the main provider types (local authority, housing association, private and joint NHS/voluntary sector). Measures were taken of the quality of the physical environment, staff and resident characteristics, external management arrangements and internal management regimes, resident satisfaction and staff stress. Direct observations were also made of the amount and quality of staff-resident interactions.

In general, the most disabled residents were found to be still living in hospital in the worst conditions and receiving the poorest quality of care. Although there were some problems with missing data, hospital residents also seemed most dissatisfied with their living situation. There were few differences between community providers regarding either the quality of care provided or the levels of reported satisfaction. Quality of care in the community homes seemed to be much more determined by the personality and orientation of project leaders.

Purchasers and providers still need to give attention to the problems of selectively discharging the most able residents to the community, leaving the most disabled being looked after in progressively deteriorating conditions. All residential providers need to review their internal management practices and try to ensure that residents are offered, as far as possible, the opportunity to make basic choices about where and how they will live. Staff training and quality assurance practices need to be reviewed in order to improve the direct quality of care offered to the most disabled individuals.

The diagnostic category of generalised anxiety disorder (GAD) was originally intended to describe residual anxiety states. Over the years clinical criteria have been refined in an attempt to describe a unique diagnostic entity. Given these changes, little is known about the clinical course of this newly defined disorder. This study investigates the longitudinal course, including remission and relapse rates, for patients with DSM–III–R defined GAD.

Analysis of the 164 patients with GAD participating in the Harvard Anxiety Research Program. Patients were assessed with a structured clinical interview at intake and re-examined at six month intervals for two years and then annually for one to two years. Psychiatric Status Ratings were assigned at each interview point. Kaplan–Meier curves were constructed to assess likelihood of remission.

Comorbidity was high, with panic disorder and social phobia as the most frequently found comorbid disorders. The likelihood of remission was 0.15 after one year and 0.25 after two years. The probability of becoming asymptomatic from all psychiatric symptoms was only 0.08.

This prospective study confirms the chronicity associated with GAD and extends this finding to define the one and two year remission rates for the disorder. Likelihood of remission for GAD and any other comorbid condition after one year was half the annual remission rate for GAD alone.

Although a cognitive–behavioural formulation of morbid jealousy has been described there is little empirical research into the practical usefulness of this model. This study evaluated the effectiveness of treating non-psychotic morbid jealousy using a cognitive approach.

Cognitive–behavioural and emotional measures of jealousy were calibrated by comparison with 40 non-jealous normal controls. These instruments were used as measures of change to assess the effectiveness of cognitive therapy in altering cognitive errors in 30 morbidly jealous out-patient referrals, divided into delayed and immediate treatment subgroups to assess the stability of the condition. Both groups completed all measures immediately before and after treatment, and at follow-up.

The instruments demonstrated significant differences between jealous and non-jealous subjects on cognitive–behavioural and emotional aspects of jealousy. The delayed treatment group showed no significant alteration in scores on any of the instruments after 12 weeks on the waiting list, confirming the stability of the condition. In the majority of cases cognitive therapy aimed at the modification of dysfunctional cognitive processes resulted in a significant improvement on all jealousy measures, immediately after treatment and at follow-up. The improvement reported by patients was supported by the partner's ratings.

The results support the postulation of the cognitive model that modification of cognitive schema by a schema-focused treatment package results in a significant reduction in disturbance in all aspects of the jealousy syndrome.

Severe personality disorder (SPD) is an imprecise but useful term referring to some notoriously difficult to treat psychiatric patients. Their long-term psychiatric treatment is often unsuccessful, in spite of hospitalisation. The specialist expertise of in-patient psychotherapy units (IPUs) can successfully meet some of SPD patients' needs.

Relevant literature on the subject is summarised and integrated with the authors' specialist clinical experience.

Many clinical problems with SPD patients are interpersonal and prevent any effective therapeutic alliance, which is necessary for successful treatment. With in-patients, inconsistencies in treatment delivery and issues surrounding compulsory treatment reinforce patients' mistrust of professionals, compromising accurate diagnosis and an assessment of the need for specialist IPU referral.

General psychiatric teams are well-placed to plan long-term treatment for SPD patients which may include IPU treatment. Timely referral of selected SPD patients to an IPU maximises a successful outcome, especially if there is appropriate post-discharge collaboration with general psychiatric teams to consolidate gains made.

The possibility that treatment with tricyclic antidepressants, in the form of dothiepin, might attenuate benzodiazepine withdrawal symptoms was investigated in a double-blind trial.

Eighty-seven non-depressed psychiatric out-patients with putative normal dose benzodiazepine dependence had their benzodiazepines reduced in stepwise amounts of 20% of the original dose for eight weeks. The patients were randomised to receive dothiepin (with dosage increasing to 150 mg/day) or placebo as an aid to withdrawal before benzodiazepine reduction and these drugs were taken for four further weeks before being stopped.

Fewer patients entered and completed the study than expected and a Type II error was possible in the results. Although there was some evidence of withdrawal symptoms being less marked in those patients allocated to dothiepin this was independent of any antidepressant effect as depression scores were lower in the placebo group in the early phase of withdrawal (P<0.01). Of those completing the study, greater satisfaction (P=0.03) was recorded by those who had received dothiepin; no other differences reached statistical significance.

Dothiepin (and by implication other tricyclic antidepressants) might have some value in reducing benzodiazepine withdrawal symptoms but does not aid drug withdrawal.

We set out to determine whether and to what degree life events independent of illness increase the risk of relapse in schizophrenia following withdrawal from medication in the previous 6 months, either by triggering a relapse in the following 4 weeks or by acting cumulatively over time.

Seventy-one patients fulfilling DSM–III–R criteria for schizophrenia with chronic illness were followed for 48 weeks and assessed on the LEDS scale. Half were treated with regular neuroleptic medication and half had been recently withdrawn from medication. A subgroup was randomised double-blind to treatment or placebo.

A proportional hazards regression model showed that life events made a significant cumulative contribution over time (P < 0.05) to the risks of relapse and that ceasing medication made an independent contribution. The risk of relapse increased in proportion to the number of life events but no interaction between medication status and events could be detected, i.e. life events were not more closely associated with relapse on medication than off medication. For those of the sample exposed to the mean rate of life events during the study period, it was estimated that 23% of the relapse risk could be attributed to life events, and for those with twice the mean rate of events, 41%. In contrast, patients who continued on regular medication had 80% less risk of relapse than those who had been withdrawn from medication either by choice or under double-blind controlled conditions.

A contribution of life events to the risk of relapse in schizophrenia was confirmed by this study but the hypothesis that life events trigger relapse was not supported, nor was the hypothesis that life events are more relevant to relapse in patients on maintenance medication than in patients off medication.

Rapid-cycling bipolar disorder is defined as four or more affective episodes yearly. The conventionally recognised limit in episode duration is usually considered 24 hours (i.e. a cycle duration of 48 hours). We report a small series of intensively observed bipolar patients who showed much faster patterns of mood oscillation.

Detailed, systematic, longitudinal assessment of five bipolar patients during extended in-patient psychiatric evaluation were conducted, including retrospective life charting and prospective evaluation of daily mood by self and blinded observer ratings, and motor activity recording.

Our data demonstrate a spectrum of cycling frequencies in rapid-cyclers, including distinct, clinically robust mood shifts that occur at frequencies faster than once per 24 hours. Affective oscillations spanned a range of cycling frequencies from four episodes per year (rapid cycling) to those occurring within the course of weeks to several days (ultra-rapid cycling), to distinct, abrupt mood shifts of less than 24 hours' duration (ultra-ultra rapid or ultradian cycling). The time of onset and duration of these ultradian affective fluctuations are highly variable and they are observed in bipolar patients without evidence of personality disorder.

The potential clinical and theoretical implications of these first systematic observations of ultra-rapid and ultradian cycling in the context of the evolution of otherwise classical bipolar affective illness are discussed.

We test the hypothesis that negative attitudes to mentally ill people may be fuelled by a lack of knowledge.

A census of knowledge of mental illness was conducted in two areas prior to the opening of long-stay supported houses for the mentally ill in each area. Three attitudinal factors (Fear and Exclusion, Social Control and Goodwill) which had been extracted by factor analysis of the Community Attitudes toward the Mentally III (CAMI) inventory (see previous paper) were analysed in respect of their associations with knowledge of mental illness.

Most respondents (80%) knew of somebody who had a mental illness but a substantial proportion of respondents had little knowledge about mental illness. Social Control showed an association with knowledge of mental illness. Groups who showed more socially controlling attitudes (especially those over 50 years old, those of lower social class, and those of non-Caucasian ethnic origin) had less knowledge about mental illness. Regression analysis revealed that when knowledge was taken into account age had no effect on Social Control, and the effect of social class and ethnic origin was diminished Respondents with children, who showed more Fear and Exclusion, were not less knowledgeable about mental illness.

The results support the hypothesis that negative attitudes, especially in older people, are fuelled by a lack of knowledge. Negative attitudes among people with children are not related to a lack of knowledge.