Cost Implications of Treatment Non-Completion in a Forensic Personality Disorder Service

Abstract

Background: A high proportion of individuals admitted to specialist secure hospital services for treatment of personality disorder do not complete treatment. Non-completion has been associated with poorer treatment outcomes and increased rates of recidivism and hospital readmission, when compared with individuals who do complete treatment or who do not receive treatment at all.

Aims: In this study, we sought to determine the economic consequences of non-completion of treatment, using case study data from a secure hospital sample. Both health and criminal justice service perspectives were taken into account.

Methods: Data were collected from a medium secure hospital personality disorder unit. A probabilistic decision-analytic model was constructed, using a Markov cohort simulation with 10,000 iterations. The expected cost differential between those who do and those who do not complete treatment was estimated, as was the probability of a cost differential over a 10-year post-admission time horizon.

Results: On average, in the first 10 years following admission, those who do not complete treatment go on to incur £52,000 more in costs to the National Health Service and criminal justice system than those who complete treatment. The model estimates that the probability that non-completers incur greater costs than completers is 78%.

Conclusion: It is possible that an improvement in treatment completion rates in secure hospital personality disorder units would lead to some cost savings. This might be achievable through better selection into treatment or improved strategies for engagement and retention. Our study highlights a financial cost to society of individuals discharged from secure hospital care when incompletely treated. We suggest that it could, therefore, be useful for secure hospitals to introduce routine monitoring of treatment completion.

Keywords: personality disorder, treatment completion, Markov model

JEL Classification: I19, K49