Verheul et al's article (Reference Verheul, van den Bosch and Koeter2003) states that dialectical behaviour therapy is an efficacious treatment for high-risk behaviours in patients with borderline personality disorder and suggests that this occurs via four core features (Reference LinehanLinehan, 1993): routine monitoring; modification of high-risk behaviours; encouragement of patients to consult therapists before carrying out these behaviours; and prevention of therapist burnout.

We propose a management strategy for these patients delivered via a systemic approach that incorporates these principles and is especially relevant for services without the capacity to provide the skills base or intensity required for effective dialectical behaviour therapy. Such an approach has been developed by our service and is currently the principal method of working with clients with borderline personality disorder in the lower North Island of New Zealand. It is a service-wide intervention with a long-term perspective, providing stabilisation and containment for both patient and staff. It is encapsulated in a management plan – a behavioural intervention to minimise reinforcement of hazardous behaviours and promote self-responsibility.

The plan defines the treatment system (e.g. psychiatric team, family, police, accident and emergency department staff), contains an acceptance of risk and explains the dangers of risk-averse responses from the service (Reference MaltsbergerMaltsberger, 1994). This breaks the cycle of assuming responsibility for the client and replaying a traumatising parent–child dynamic, with subsequent regression, increased risk and institutionalisation. We found that this is achieved through the process of writing and implementing the plan and it enables patients to move towards autonomous functioning. It must be agreed to by all involved and regular review meetings provide a forum for staff to own and manage their differences. Each plan should be an individualised document written by the case manager in consultation with the client; however, we have designed a template for ease of use. This work grew from the ideas of Krawitz & Watson (Reference Krawitz and Watson1999) around the use of brief admissions as a successful part of long-term management, and the observation that the majority of work by out-of-hours services involved these ‘revolving door’ patients. As yet, our approach has been validated only by empirical evidence. A paper is currently in preparation.