High Scores on the Montgomery-Åsberg Depression Rating Scale and Psychotic Symptoms Predict Suicide: A Prospective Cohort Study of Psychiatric Acute Ward Patients.

Objective: To investigate the role of depression severity in suicide risk by studying the predictive value of psychotic symptoms and depression scale scores, controlled for suicidal behavior and gender.Methods: We conducted a prospective cohort study of consecutive psychiatric acute ward admissions between 2005 and 2014 from a Norwegian catchment area. Inclusion criteria were an ICD-10 diagnosis of unipolar or bipolar depression with a current depressive episode (n = 1,846); depression severity was measured by the Montgomery-Åsberg Depression Rating Scale (MADRS). Patients were assessed for suicidal ideation/planning, self-harm, and recent suicide attempts on admission. Mean follow-up time was 5.5 years (minimum/maximum: 0/10.6 years). We used Cox regression analyses and Kaplan-Meier analyses to explore potential predictors and time to suicide.Results: During the follow-up period, 46 patients died by suicide, 30 (65%) of these within the year following admission. Psychotic depression (P = .014), admission MADRS score (P = .006), suicide attempts (P = .021), and male sex (P = .043) significantly predicted suicide. Suicidal ideation and self-harm did not predict suicide. The cumulative suicide risk in psychotic depression was 1.7% after 12 weeks and 3.0% after 52 weeks.Conclusions: Depression severity as measured with the MADRS or a diagnosis of psychotic depression independently predicted suicide. More suicides may be prevented by implementing intensive treatment and post-discharge follow-up for patients who present to psychiatric acute wards with severe depressive episodes and recent suicide attempts, regardless of self-reported suicidal ideation, suicide plans, and self-harm.

"Everyone Needs a Friend Sometimes" - Social Predictors of Long-Term Remission In First Episode Psychosis.

Background: Predictors of long-term symptomatic remission are crucial to the successful tailoring of treatment in first episode psychosis. There is lack of studies distinguishing the predictive effects of different social factors. This prevents a valid evaluating of their independent effects. Objectives: To test specific social baseline predictors of long-term remission. We hypothesized that first, satisfaction with social relations predicts remission; second, that frequency of social interaction predicts remission; and third, that the effect of friend relationship satisfaction and frequency will be greater than that of family relations satisfaction and frequency. Material and Methods: A sample of first episode psychosis (n = 186) completed baseline measures of social functioning, as well as clinical assessments. We compared groups of remitted and non-remitted individuals using generalized estimating equations analyses. Results: Frequency of social interaction with friends was a significant positive predictor of remission over a two-year period. Neither global perceived social satisfaction nor frequency of family interaction showed significant effects. Conclusions: The study findings are of particular clinical importance since frequency of friendship interaction is a possibly malleable factor. Frequency of interaction could be affected through behavioral modification and therapy already from an early stage in the course, and thus increase remission rates.