An Organizational and Community-Based Approach to Deliberate Self Harm

DEAR SIR,

We wish to describe the changes in response of a mental health service to deliberate self harm (DSH). In 1993, the former Grampians Psychiatric Services (GPS) was confronted with the increasingly difficult task of managing people with deliberate self harm. Five single women with a mean age of 26.6 years and who met an average of 8.4 of 9 DSM-IV criteria for borderline personality disorder were identified. On average, they had previously received 5.6 different diagnoses, had been admitted voluntarily on 12.5 occasions to 6.4 different hospitals but had become involuntary within 6.4 days, with threats to leave or of further DSH. They had been hospitalized for a mean total of 1015 days, had 1162 hours of seclusion and had received a mean of 14 different psychotropic agents. Four had received electroconvulsive therapy (ECT). All had significant medical histories including meningitis, cardiac murmurs, deep vein thrombosis, pelvic inflammatory disease and hepatitis C. All had been sexually abused from a mean age of 7.25 years, three had comorbid substance abuse and four had a forensic history. DSH included threatened or actual immolation, mutilation of genitals, ingestion of foreign bodies, jumping from buildings or walking in front of cars and trains. Threats to staff were personal, with a forensic history including arson, siege, threats to kill, stalking and armed robbery. Two patients had been imprisoned.

To put these figures in context, GPS was one of 22 Victorian Area Mental Health Services and had a population of 204 208 spread across 48 112 square kilometres. There was 10.6% unemployment, 16% of families were ‘single parent’ and 20% of households received less than $25 000 per annum. Available resources included a fixed workforce derived from two large regional asylums in the process of closure, a planned 40% budget reduction over 2 years and six psychiatrists. The political backdrop included local, multi-agency concern, a focus on psychotic disorders and de-institutionalization, and a trend towards police shootings.

In 1994, GPS began an organizational, community-based model of care, which resulted in these same five women receiving a different service. GPS-initiated phonecalls increased from 2.5 to 79.5, home visits from 4.0 to 151.3 and crisis contacts from 0.8 to 17 per person. The mean number of admissions, length of stay and time in seclusion fell to 7, 103.2 days and 21.6 hours respectively. The mean number of psychotropics dropped to 2.8 per person with one receiving ECT. Following the use of this model, no new patient became as severely disturbed as these five women had become.

For health-giving professionals, the development of a humane understanding of DSH was regarded as central, GPS using the conceptual framework of abnormal illness behaviour whereby DSH was seen as ‘a calling card’ by which people who see themselves as fundamentally despicable could make contact, at least with professionals who had a duty to respond. Management revolved around seeing a person regularly rather than in crisis – a “You don't need to harm yourself, I will see you anyway” social contact became uncoupled from DSH. That is, people with DSH were embraced rather than alienated as service users. Realistic expectations including an acknowledgement of the iatrogenic effects of protracted admissions and the inappropriate use of medication led to a model whereby people were neither admitted nor evicted onto the street. Rather, people could be cared for in a different way. While ultimately revised, the statement that hospital and medication were unhelpful gave clinicians permission to respond in a different way. The guidelines eventually included the flexibility of brief, up to 48-hour admission for diagnostic clarification or to defuse an exhausted community. Regular respite admissions and a patient-based voucher system in which the patient chose when to use a predetermined number of inpatient days per year were also explored. 1 As no sustainable change was noted despite multiple psychotropic use and because of the potential for the medicalization of psychosocial disadvantage and for acting out, a decision to not prescribe to new patients and to gradually invite medication reduction in existing patients was made. For those patients unable to tolerate this, medication was kept constant. Crises were managed by two competent adults, the patient and the case manager. The task was to discover what had happened and what could be done differently with live problems from a community existence. Tiny changes were reinforced, phoning before cutting acknowledged as a major achievement after weeks of work. Regular contact was a fundamental principle, each person in the relationship keeping the same time each week with each developing an ability to contain their distress until the next appointment. A ‘sliding scale of insulin’ approach to exceptional need followed by a return to regular contact allowed for flexibility. The diagnosis, and risk and personal responsibility of suicide was discussed with the patient and, with consent, with other family.

A specialist team was not established as it was thought that to do so may create the possibility of failure, should key individuals leave or become overwhelmed by the responsibility. Instead, each clinician became part of an intersecting network of service responses that together could support the person in the community. The patient was encouraged to develop a relationship with the service as a whole, as well as with a single case manager. Seconds were nominated and introduced for roster change, holiday and sick leave and for handover at pre-agreed times. Realistic caseloads were important.

A medico-legally safe environment within which staff could work in the community was discussed with management, hospital solicitors and subsequently the coroner. Having been appraised of the dilemmas, their support was both crucial and surprisingly forthcoming. Training over a 12-month period included a questionnaire and a 1-page summary of the model. These were distributed to the acute, community and inpatient components of the psychiatric service, the mainstreamed general hospital and to those agencies with whom the patient continued to have contact. The latter included the ambulance service, police, general practitioners and non-government agencies. Rather than requiring agencies to agree, the presentations focused on understanding how and why care was being delivered in a different way. In addition, a 2-day in-service was available for interested clinicians. The voluntary nature was an acknowledgement that, for some clinicians, DSH was not their interest area. A workbook 2 accompanied the course with four essential references addressing relationship management, 3 medico-legal dilemmas, 4 splitting 5 and malignant alienation. 6 Subsequently, a short, practical text has been published 7 that describes the development of specific management plans for each person that can be consistently administered across services and two useful papers on professional risk taking. 8 , 9 These same principles of care were later introduced and accepted by the child and adolescent service.

Caring in a different way offered a clinical approach and softened previous therapeutic nihilism and antipathy. There were difficulties including the practicalities of ensuring regular referral to readily available, consistently applied management plans and for the quiet patience of the confident clinician. For some, threats were too frightening and a conservatism remained. Managing this split was important, modelling as it does an ability to accept different opinions and to avoid polarization. While successful in reducing hospitalization and inappropriate prescribing, managing long-term dsyphoria and clinician fatigue was underestimated.

For the service wishing to work with people with DSH, morale is important. Peer support and review, prospectively collecting data with diagnostic clarity rather than pejorative labelling, and knowing the results of that data are important. Unique observations, the unexpected finding of significant medical histories unrelated to DSH, the clinical triad of multiple diagnoses, multiple admissions and multiple medications as a diagnostic marker, and the differing length of stays between hospitals and psychiatrists were intriguing. Contact with specialist providers and access to journals and conferences also maintains enthusiasm.