Western media coverage of the violence associated with the 2003 US-led invasion of Iraq has contrasted in magnitude and nature with population-based survey reports.

The purpose of this study was to evaluate the extent to which first-hand reports of violent deaths were captured in the English language media by conducting in-depth interviews with Iraqi citizens.

The England-based Iraq Body Count (IBC) has methodically monitored media reports and recorded each violent death in Iraq that could be confirmed by two English language media sources. Using the capturerecapture method, 25 Masters' Degree students were assigned to interview residents in Iraq and asked them to describe 10 violent deaths that occurred closest to their home since the 2003 invasion. Students then matched these reports with those documented in IBC. These reports were matched both individually and crosschecked in groups to obtain a percentage of those deaths captured in the English language media.

Eighteen out of 25 students successfully interviewed someone in Iraq. Six contacted individuals by telephone, while the others conducted interviews via e-mail. One out of seven (14%) phone contacts refused to participate. Seventeen out of 18 primary interviewees resided in Baghdad, however, some interviewees reported deaths of neighbors that occurred while the neighbors were elsewhere. The Baghdad residents reported 161 deaths in total, 39 of which (24%) were believed to be reported in the press as summarized by IBC. An additional 13 deaths (8%) might have been in the database, and 61 (38%) were absolutely not in the database.

The vast majority of violent deaths (estimated from the results of this study as being between 68–76%) are not reported by the press. Efforts to monitor events by press coverage or reports of tallies similar to those reported in the press, should be evaluated with the suspicion applied to any passive surveillance network: that it may be incomplete. Even in the most heavily reported conflicts, the media may miss the majority of violent events.

Little is known about self-harm occurring during the period of untreated first-episode psychosis.

To establish the prevalence, nature, motivation and risk factors for self-harm occurring during the untreated phase of first-episode psychosis.

As part of the æSOP (Aetiology and Ethnicity in Schizophrenia and Other Psychoses) study, episodes of self-harm were identified among all incident cases of psychosis presenting to services in south-east London and Nottingham over a 2-year period.

Of the 496 participants, 56 (11.3%) had engaged in self-harm between the onset of psychotic symptoms and first presentation to services. The independent correlates of self-harm were: male gender, belonging to social class I/II, depression and a prolonged period of untreated psychosis. Increased insight was also associated with risk of self-harm.

Self-harm is common during the pre-treatment phase of first-episode psychosis. A unique set of fixed and malleable risk factors appear to operate in those with first-episode psychosis. Reducing treatment delay and modifying disease attitudes may be key targets for suicide prevention.

Research studies for the treatment of the putative prodromal phase of psychotic disorders have begun to appear

To obtain preliminary evidence of the short-term efficacy and safety of aripiprazole treatment in people with the psychosis prodrome

Fifteen participants meeting prodrome criteria (mean age 17.1 years, s.d.=5.5) enrolled in an open-label, single-site trial with fixed-flexible dosing of aripiprazole (5–30 mg/day) for 8 weeks

In the mixed-effects repeated-measures analysis, improvement from baseline on the Scale of Prodromal Symptoms total score was statistically significant by the first week. No participant converted to psychosis and 13 completed treatment. Neuropsychological measures showed no consistent improvement; mean weight gain was 1.2 kg. Akathisia emerged in 8 participants, but the mean Barnes Akathisia Scale score fell to baseline levels by the final visit. Adverse events were otherwise minimal

Aripiprazole shows a promising efficacy and safety profile for the psychosis prodrome. Placebo-controlled studies are indicated

Schizophrenia is known to be associated with a range of adverse outcomes, which have an impact at the societal level and are therefore of public concern

To examine the epidemiology and methods for measuring six adverse outcomes in schizophrenia: violence, victimisation, suicide/self-harm, substance use, homelessness and unemployment

Areview of the literature was carried out for each adverse outcome, with attention to critical appraisal of existing measurement tools

Schizophrenia is associated strongly with all six outcomes, although research has mainly focused on violence. Each outcome acts as a risk factor for at least some of the other outcomes. There are few standardised or validated measures for these'hard' outcomes. Each measure has inherent biases but a growing trend is for these to be minimised by using multiple measures

Asingle instrument which systematically measures multiple societal outcomes of schizophrenia would be extremely useful for both clinical and research purposes

Little is known about the determinants of violence in women with psychosis.

To identify predictors of violence in a community sample of women with chronic psychosis.

The 2-year prevalence of physical assault was estimated for a sample of 304 women with psychosis. Baseline socio-demographic and clinical factors were used to identify predictors of assault.

The 2-year prevalence of assault in the sample was 17%. Assaultive behaviour was associated with previous violence (OR=5.87, 95% CI 2.42–14.25), non-violent convictions (OR=2.63, 95% CI 1.17–5.93), victimisation (OR=2.46, 95% CI 1.02–5.93), African–Caribbean ethnicity (OR=2.24, 95% CI 1.02–4.77), cluster B personality disorder (OR=2.66, 95% CI 1.11–6.38) and high levels of unmet need (OR=1.17, 95%C11.01–1.35). An interaction between African–Caribbean ethnicity and cluster B personality disorder was identified in relation to violent outcome. Violent women were found to be more costly to services.

Nearly a fifth of community-dwelling women with chronic psychosis committed assault over a period of 2 years. Six independent risk factors were found to predict violence.