Predictors of outcomes of assertive outreach teams: a 3-year follow-up study in North East England

Introduction Assertive outreach (AO) is a required component of services for people with severe mental illness in England. However, the claims to its effectiveness have been contested and the relationships between team organisation, including model fidelity, the use of mental health interventions and outcomes for service users remain unclear. Method Three-year follow up of 33 AO teams was conducted using standardised measures of model fidelity and mental health interventions, and of current location and a range of outcomes for service users (n = 628). Predictors of the number of hospital admissions, mental health and social functioning at T2, and discharge from the team as ‘improved’ were modelled using multivariate regression analyses. Results Teams had moderate mean ratings of fidelity to the AO model. All rated highly on the core intervention modalities of engagement, assessment and care co-ordination, but ratings for psychosocial interventions were comparatively low. Two-thirds (462) of service users were still in AO and data were returned on 400 (87%). There was evidence of small improvements in mental health and social functioning and a reduction in the mean number of hospital admissions in the previous 2 years (from 2.09 to 1.39). Poor outcomes were predicted variously by service users’ characteristics, previous psychiatric history, poor collaboration with services, homelessness and dual diagnosis. Fidelity to the AO model did not emerge as a predictor of outcome, but the team working for extended hours was associated with more frequent in-patient admissions and less likelihood of discharge from AO. Supportive interventions in daily living, together with the team’s use of family and psychological interventions were also associated with poorer outcomes. Possible explanations for these unexpected findings are considered. Conclusion AO appears to have been quite successful in keeping users engaged over a substantial period and to have an impact in supporting many people to live in the community and to avoid the necessity of psychiatric hospital admission. However, teams should focus on those with a history of hospital admissions, who do not engage well with services and for whom outcomes are less good. Psychosocial interventions should be applied. The relationship between model fidelity, team organisation, mental health interventions and outcomes is not straightforward and deserves further study.