Treatment adherence in adolescent psychiatric inpatients with severe disruptive behaviour

Abstract

Objective

To examine pre‐treatment characteristics associated with dropout in 224 adolescent psychiatric inpatients (mean age 17.1 years) with severe disruptive behaviour. To date, little is known about the factors predicting dropout among adolescents treated for severe disruptive behaviour. This is surprising, as dropout is a major problem in this specific group.

Method

Dropouts (n = 77) and completers (n = 147) were compared on known risk factors for dropout, such as severity of externalising problems and disorders, ethnic minority status, male gender, and lower academic functioning, as well as on other factors considered relevant: behavioural characteristics, including age of onset and different types of disruptive behaviour. Within dropout, withdrawal (termination against the advice of the therapist; n = 40) and pushout (termination against the wish of the client; n = 37) were distinguished.

Results

Two characteristics significantly predicted dropout: early onset of disruptive behaviour and cannabis usage prior to treatment. Within the dropout group no differences were found between withdrawals and pushouts.

Conclusions

Predicting dropout among adolescent psychiatric inpatients with severe disruptive behaviour is difficult. The two predictors found were already present at the time of admission and are therefore considered unalterable. However, they can be used to pinpoint individuals with higher chance at dropout, who in turn can be offered tailored interventions aimed at improving the therapeutic relationship.

Abstract

Key words:

• adolescent psychiatry
• antisocial behaviour
• dropout
• pushout

Acknowledgements

The authors gratefully acknowledge the contribution of the adolescents who participated in this study, the colleagues of De Fjord (Emmeline Ravestijn, Hinke Hiemstra, Igor Dudevsky, Joes Hagendoorn, Martine van der Laan, Peter van der Sanden, Petra Bakarbessy, Patricia Oosterom, Sylvia de Raat, Wijnand Mulder, Winston Leeflang) for their contribution and interest, in particular Zita Haijer, who enabled the study. We also wish to thank Danielle Braspenning for her helpful comments, and Mathijs Deen for statistical aid. This research did not receive any specific grant from funding agencies in the public, commercial, or not‐for‐profit sectors.