Longitudinal data arise frequently in psychiatric investigations, and are most often analysed by multivariate analysis of variance (MANOVA) procedures. However, as routinely applied, the method is not satisfactory, particularly when the data are affected by subjects dropping-out of the study. More suitable methods are now available.

Problems with the MANOVA approach are discussed and the advantages of alternative procedures stressed.

Using MANOVA on complete cases to analyse unbalanced longitudinal data can be seriously misleading. More recently developed methods are far more suitable, but only if the missing values are non-informative.

Routine use of MANOVA for the analysis of longitudinal data, particularly when there is a substantial proportion of drop-outs, is ill advised. Statisticians have considerably enriched the available methodologies during the past decade, and psychiatric researchers dealing with such data should be aware of the advantages of the newer methods.

An instrument was required to quantify and thus potentially measure progress towards a Health of the Nation target, set by the Department of Health, “to improve significantly the health and social functioning of mentally ill people”

A first draft was created in consultation with experts and on the basis of literature review. This version was improved during four stages of testing: two preliminary stages, a large field trial involving 2706 patients (rated by 492 clinicians) and tests of the final Health of the Nation Outcome Scales (HoNOS), which included an independent study (n=197) of reliability and relationship to other instruments.

The resulting 12-item instrument is simple to use, covers clinical problems and social functioning with reasonable adequacy, has been generally acceptable to clinicians who have used it, is sensitive to change or the lack of it, showed good reliability in independent trials and compared reasonably well with equivalent items in the Brief Psychiatric Rating Scales and Role Functioning Scales.

The key test for HoNOS is that clinicians should want to use it for their own purposes. In general, it has passed that test. A further possibility, that HoNOS data collected routinely as part of a minimum data set, for example for the Care Programme Approach, could also be useful in anonymised and aggregated form for public health purposes, is therefore testable but has not yet been tested.

The aim was to compare the efficacy of intensive clinical case management (ICM) with standard community care in the management of ‘hard to treat’ patients with a severe mental illness.

A randomised controlled trial was carried out in East Lambeth, a deprived area of inner London. Seventy people with psychosis designated as ‘hard to treat’ by referring teams were included; 35 were randomised to ICM (case load eight patients per worker), and 35 to standard care, which offered follow-up by a community psychiatric nursing service (30 patients per worker). Outcome measures were admissions and hospital bed utilisation; contact with services; symptomatology; social behaviour; social functioning; quality of life; patients' satisfaction with care at 9 and 18 months.

There were no differences in patients' symptoms, social behaviour or social functioning. Quality of life was significantly improved in patients receiving ICM at 9 months. Satisfaction with care was significantly greater among case-managed patients. All ICM patients remained in contact with services throughout the study, while six control patients were refusing all contact with services at 18 months.

ICM failed to improve the clinical outcome of ‘hard to treat’ patients. The service was successful in maintaining contact with patients, was greatly appreciated and had a positive effect on their perceived quality of life.

In recent years, studies of the risk of recurrence in affective disorder in relation to the number of prior episodes have given contradictory results.

Survival analysis was used to calculate the rate of recurrence after successive episodes in a case register study including all hospital admissions with primary affective disorder in Denmark during 1971–1993. A total of 20 350 first-admission patients were discharged with a diagnosis of affective disorder, depressive or manic/cyclic type.

The rate of recurrence increased with the number of previous episodes in both unipolar and bipolar disorder. Initially, the two types of disorders followed markedly different courses, but later in the course of the illness the rate of recurrence was the same for the two disorders.

The course of severe unipolar and bipolar disorder seems to be progressive in nature despite the effect of treatment.

The risk of recurrence in affective disorder has been found to increase with each new episode. It is unclear whether it is universal without regard to gender, age and type of disorder.

Survival analysis was used to estimate the risk of recurrence in a case-register study including all hospital admissions with primary affective disorder in Denmark from 1971–1993. In this period 20 350 first-admission patients had been discharged with a diagnosis of affective disorder, depressive or manic/cyclic type.

The risk of recurrence increased with the number of previous episodes regardless of the combination of gender, age and type of disorder. Initially in the course of illness, unipolar and bipolar women experienced an equal greater risk of recurrence than men. The risk of recurrence after the first episode was increased for middle-aged and older unipolar women compared with the risk for younger women, while after all other episodes younger age at first episode increased the risk of recurrence.

The course of severe unipolar and bipolar disorder seems to be progressive in nature irrespective of gender, age and type of disorder.

The lifetime risks of suicide are generally quoted as 15% for affective disorder and alcoholism and 10% for schizophrenia, based on data from 1921–1975 and on calculations performed before computerised modelling techniques became available. This study recalculates the risk using contemporary data and modern techniques.

Twenty-seven mortality studies provided data for affective disorder, 27 for alcohol dependence and 29 for schizophrenia. The proportion of the cohort who had died was plotted against the proportion of deaths from suicide. Modelling techniques fitted curves through the data points extrapolating them to cohort extinction, thus estimating the lifetime risk of suicide for each disorder.

The lifetime risk was estimated at 6% for affective disorder, 7% for alcohol dependence and 4% for schizophrenia.

The lifetime suicide risk figures often quoted in the literature appear to be too high.

Factors associated with aggression among psychiatric in-patients are still poorly understood. Inconsistent published findings are partly attributable to methodological problems. Much previous research has been undertaken in specialist settings, where conditions are likely to differ from those in acute admission wards.

Levels of aggression were ascertained weekly by staff report, using a reliable measure, for every patient on five acute admission wards and one locked intensive care ward at two hospitals prospectively over a five-month period.

Levels of aggression varied considerably among the admission wards. Aggression was more common on the locked ward and among younger patients. Factors associated with aggression changed with time since admission (seldom examined in previous studies). Complex associations were found with gender, ethnic group and diagnosis.

Some of the findings might be explicable by differences in ward environment, and staff–patient interactions. Further investigation into these is warranted.

We compared the anti-depressant and cognitive effects of up to eight sessions of bilateral, brief pulse electroconvulsive therapy (ECT) administered twice (ECT x 2) or three times weekly (ECT x 3), to confirm that ECT x 3 acts more rapidly although the two schedules are equivalent in antidepressant outcome, and to establish whether ECT x 3 is indeed associated with more severe memory impairment.

Patients with major depression, endogenous subtype were randomly assigned to ECT x 3 or ECT x 2 plus one simulated ECT per week, both up to a maximum of eight real ECT. Depression was evaluated by the Hamilton Depression Scale the day after each treatment and cognitive function by a test battery administered before and after the ECT series and at one month follow-up.

Assessed categorically or parametrically, there was no significant difference in antidepressant outcome between the two schedules. Rate of response was significantly more rapid with ECT x 3 but was associated with more severe memory impairment.

Twice weekly administration is an optimum schedule for bilateral ECT unless clinical indications require the more rapid antidepressant effect of three times weekly treatment.

Animal experimental studies suggest that repeated administration of selective serotonin reuptake inhibitors (SSRIs) produces complex adaptive changes in brain serotonin (5-HT) pathways. The effect of these adaptive changes on different aspects of brain 5-HT neurotransmission and their clinical consequences are not well understood.

We studied the effect of repeated administration of the SSRI, paroxetine (20 mg daily), on the Cortisol responses to the 5-HT precursor, 5-hydroxytryptophan (5-HTP), in healthy subjects and depressed patients.

In healthy subjects, following one week of paroxetine treatment there was a large increase in the Cortisol response to 5-HTP. This increase had all but disappeared following 3 weeks treatment. In contrast, in depressed patients treated with paroxetine for 8 weeks, the Cortisol response to 5-HTP was significantly increased.

SSRI treatment in depressed patients produces a persistent increase in the Cortisol response to 5-HTP, a probable measure of neurotransmission at central 5-HT2 receptors. Potentiation of 5-HT2 neurotransmission is unlikely to account for the efficacy of SSRIs in major depression but might underlie their actions in obsessive–compulsive disorder and also perhaps certain of their adverse effects, notably sexual dysfunction.

Little is known about the outcome of common mental disorders (CMD) in primary care attenders in low income countries.

Two and 12 month (T1 and T2) follow-up of a cohort of cases of CMD (n=199) recruited from primary health, traditional medical practitioner, and general practitioner clinics in Harare, Zimbabwe. The Shona Symptom Questionnaire (SSQ) was the measure of caseness.

The persistence of case level morbidity was recorded in 41% of subjects at 12 months. Of the 134 subjects interviewed at both follow-up points, 49% had recovered by T1 and remained well at T2 while 28% were persistent cases at both T1 and T2. Higher SSQ scores, a psychological illness model, bereavement and disability predicted a poor outcome at both times. Poorer outcome at T1 only was associated with a causal model of witch-craft and an unhappy childhood. Caseness at follow-up was associated with disability and economic deprivation.

A quarter of cases of CMD were likely to be ill throughout the 12 month follow-up period. Targeting risk groups for poor outcome for interventions and policy interventions to reduce the impact of economic deprivation may provide a way of tackling CMD in primary care in low income countries.

Little is known of the behavioural adjustment of children of postnatally depressed mothers. Previous studies have relied on maternal reports, and have produced inconsistent findings.

In a prospective, longitudinal study of the five-year-old children of a community sample of postnatally depressed and well women, evidence was collected concerning the children's adjustment in the context of school, teachers being asked to complete questionnaires after the children had finished their first term.

Family social class and the child's gender had the most pervasive influences on adjustment. However, both postnatal and recent maternal depression were associated with significantly raised levels of child disturbance, particularly among boys and those from lower social class families.

The findings indicate a persistent effect of postnatal depression on child adjustment. They highlight the need for resources devoted to supporting mothers of young children and particularly routine screening and treatment for postnatal mood disorder.

Attachment theory argues that psychological development and functioning are affected by our earliest attachments to care-givers. Failed or pathological attachment in childhood may give rise to repetition of maladaptive attachment patterns in adulthood.

Analysis of therapeutic relationships in the light of attachment theory.

Relationships between patients and both psychiatric care-givers and institutions may resemble attachment relationships.

An attachment perspective may be useful for understanding common behavioural disturbances in general psychiatric settings, and support the use of clinical strategies which focus on containment of arousal and the management of anxiety states.

Moclobemide, a reversible inhibitor of monoamine oxidase A, previously has been reported to have efficacy in the treatment of social phobia.

Seventy-seven non-responders to one week of single-blind placebo were randomly assigned to moclobemide or placebo for eight weeks of double-blind treatment. Outcome was assessed by independent evaluator, treating psychiatrist and self-ratings. After eight weeks, patients who were at least minimally improved continued treatment for a further eight weeks.

Intention-to-treat sample response rates at week 8 were 7/40 (17.5%) for the moclobemide group and 5/37 (13.5%) for placebo (NS). Moclobemide was significantly superior to placebo on 2 of 10 primary outcome measures. Moclobemide was well tolerated.

Moclobemide may have efficacy in the treatment of social phobia, but absence of significant differences on most primary outcome measures and small effect sizes for all outcome measures suggest that the magnitude of its clinical effect is small.

This paper evaluates the effects on knowledge and clinical stability of an educational intervention about tardive dyskinesia.

Fifty-six patients receiving antipsychotic maintenance completed a questionnaire assessing their knowledge about tardive dyskinesia. After random allocation to either an educational intervention or a control group, their knowledge was reassessed at six months.

Ninety-five per cent of patients completed the study. The study patients gained significantly more knowledge than the controls, who made modest gains. There were no significant differences in clinical outcome between the groups.

Patients can learn about serious toxic effects of antipsychotic treatment with a low risk of non-compliance. Discussion about tardive dyskinesia is necessary in the process of obtaining informed consent to treatment.

A 17-year-old boy suffering from a severe schizophrenic disorder of the paranoid type and mental retardation did not respond to treatment with typical antipsychotics, whereas under clozapine treatment he showed a favourable response. Discontinuation of clozapine led to an acute psychotic relapse. During clozapine treatment the patient developed severe neutropenia.

Due to the history of unsatisfactory response to traditional antipsychotics, clozapine treatment was continued despite white blood cell (WBC) decline. Concomitant treatment with G-CSF was followed by a rapid normalisation of WBC.

This case report is not intended to challenge the clinical practice of discontinuing clozapine upon the development of neutropenia/agranulocytosis, but rather to stimulate further research in the pathophysiology and clinical consequences of a clozapine rechallenge after a WBC decline, especially in patients with a rather complex symptomatology where no sufficient therapeutic results can be achieved with any other pharmacological intervention than clozapine.

For more than a century, the idea of particular personality/behavioural characteristics being associated with people with Down's syndrome has been explored, but with inconclusive results.

The Disability Assessment Schedule was used to ascertain the behavioural profiles of 360 adults with Down's syndrome and 1829 adults with learning disabilities of other aetiologies, who were the whole identified population within a defined geographical area. Comparison was made between the two total groups and additionally for the subgroups aged < 35 years and aged ⩾35 years. Comparison was also made with regards to cluster analysis findings.

Despite an equal age and developmental quotient, the Down's syndrome group were less likely to demonstrate maladaptive behaviours. The behaviour characteristics of the adults with Down's syndrome remained constant in the younger and older age groups. Cluster analysis demonstrated adults with Down's syndrome to have an increased prevalence in cluster groupings with lower rates of maladaptive behaviours.

This study confirms there to be a behaviour phenotype among adults with Down's syndrome. The reasons for this (e.g. genetic/psychological/social) require further research. Such research may establish a better understanding of the aetiologies of maladaptive behaviours among people with learning disabilities in general.

Refugees who have suffered traumatic events present complex therapeutic challenges to health professionals. There is little research into post-exile factors that may be amenable to change, and therefore reduce morbidity. We examined the importance of social factors in exile and of trauma factors in producing the different elements of psychological sequelae of severe trauma.

Eighty-four male Iraqi refugees were interviewed. Adverse events and level of social support were measured. Various measures of psychological morbidity were applied, all of which have been used in previous trauma research.

Social factors in exile, particularly the level of ‘affective’ social support, proved important in determining the severity of both post-traumatic stress disorder and depressive reactions, particularly when combined with a severe level of trauma/torture. Poor social support is a stronger predictor of depressive morbidity than trauma factors.

Some of the most important factors in producing psychological morbidity in refugees may be alleviated by planned, integrated rehabilitation programmes and attention to social support and family reunion.