Is non‐completion of treatment related to security need?

Abstract

Background Treatment completion difficulties are common in forensic mental health settings and may have a profound impact on recidivism rates. Aims To test for associations between measures of risk and of security needs on the one hand and treatment non‐completion on the other among male offender‐patients in one medium security hospital. Methods We conducted a retrospective file study in a Flemish medium security hospital. A random sample of 25 treatment non‐completers was compared to a random same‐size sample of completers, each rated, blind to outcome, on the DUNDRUM‐1 security needs scale from data recorded at the time of admission to the unit. ‘Non‐completion’ was defined as any failure to complete treatment, whether staff‐terminated or self‐terminated; in Flanders, failure to comply with the judicial conditions of placement can result in re‐imprisonment. We used binary logistic regression to test relationships between treatment completion/non‐completion and security need, measured with the DUNDRUM‐1, together with a range of possible confounding variables. Results Most patients had psychosis and/or personality disorder and often substance use disorders also. Treatment non‐completion was invariably staff ordered because of security breaches. DUNDRUM‐1 and PCL‐R Facet 4 scores at the time of admission and HCR‐20 scores during admission were significantly higher among non‐completers than completers, but after binary logistic regression, only the DUNDRUM‐1 rating was independently associated with non‐completion. Conclusions Our study showed that an admission DUNDRUM‐1 rating, indicating levels of security need, co‐varies only to a small extent with the historical items of the HCR‐20, so may be regarded as measuring complementary domains. While conditions in Flanders at the time of the study complicated it in that medium security hospital units offered the highest level of hospital security available, the finding that non‐completion of treatment was particularly likely when the DUNDRUM‐1 indicated a higher security need than facilities could provide may have implications for all secure hospital services.

Excellence in forensic psychiatry services: international survey of qualities and correlates

“…This suggests that critical mass has a beneficial effect on breadth and depth of skills, experience and expertisesize matters, smaller is not better, although it is again noted that no service in this survey had more than 250 beds, in keeping with other modern surveys. 11 Having several security levels within a forensic mental health service is an expected feature of effective forensic psychiatry services, [44][45][46][47][48] as are linked patient pathways in all health services, 49 in forensic services 50 and in comprehensive mental health services that include forensic services and forensic pathways. [51][52][53][54] It may also be that larger services with well-defined models of care 4 inherently require or generate better governance structures for consistency and quality.…”

Italian Evaluation and Excellence in REMS (ITAL-EE-REMS): Appropriate placement of forensic patients in REMS forensic facilities

“…it is necessary also for the protection of the public and for the safe working environment of care staff and clinicians that mentally disordered offenders should not represent a level of dangerousness and risk that exceeds the capacity of the centre to which they are sent by the court. There is also some evidence that allocating a patient to a lower level of security than is safe may lead to failure to complete treatment successfully, mainly due to breaches of therapeutic security with disruptive behaviour (Jeandarme et al, 2021).…”

“…In other jurisdictions models of care for forensic psychiatry have been tested and can be shown broadly to allocate forensic patients to the correct level of care required in Ireland for a forensic hospital using the DUNDRUM Toolkit as a guide (Flynn et al, 2011b, Flynn et al, 2011a or from a remand prison to a range of levels of therapeutic security (Flynn et al, 2011b, O'Neill et al, 2016 in an English medium secure unit (Freestone et al, 2015) or an English high secure hospital (Williams et al, 2020b) and in other countries (Jeandarme et al, 2019, Adams et al, 2018, Jones et al, 2019. The value of correctly allocating to an appropriate level of therapeutic security arises from the possibility that treatment in an inappropriately low level of therapeutic security may result in failure to complete treatment (Jeandarme et al, 2021) and may arise from heuristic bias in the absence of a structured professional judgement approach (Lawrence et al, 2018). Allocating a patient with a history of very dangerous behaviour (serios violence) to a low level of therapeutic security may endanger other patients, staff and the patient themselves.…”

“…A further progression in this method has been the combination of static measures of need for therapeutic security (DUNDRUM-1) and dynamic measures of progress in need for therapeutic security (DUNDRUM-3 therapeutic programme completion and DUNDRUM-4 forensic recovery) to provide service-wide estimates of the extent to which current placements match current needs (Davoren et al, 2012, Eckert et al, 2017, Adams et al, 2018. It has been shown that placing a forensic patient in a setting of higher therapeutic security than they need may delay treatment responses, delay recovery and lead to prolonged lengths of stay (Williams et al, 2020a, Sharma et al, 2015, Davoren et al, 2013, Eckert et al, 2017 while placing a patient in a setting less therapeutically secure than needed may lead to failure to complete treatment with shortened lengths of stay (Jeandarme et al, 2021) and greater risk of disruptive behaviour, restraint and seclusion (Jewell et al, 2023).…”