Predicting length of stay in a male medium secure psychiatric hospital

This thesis examines factors associated with length of hospital stay for mentally disordered offenders, detained within the medium secure psychiatric estate. Following an introduction, Chapter two presents a systematic literature review examining the current literature on factors that predict length of stay for patients detained in medium secure hospitals. Mixed results were found. There was limited convergence across clinical and forensic variables investigated, but greater consensus on what is not associated with length of stay. The limited research available and inconsistencies found indicates the need for further research. Chapter three comprises an empirical research study, investigating which variables within a population of male mentally disordered offenders predict length of stay within a regional, medium secure psychiatric hospital. Preliminary analyses revealed statistically significant relationships between length of stay and nine variables. Effect sizes were small to medium. Logistic regression revealed a statistically significant relationship between length of stay of two years or more and having a diagnosis of schizophrenic disorder. Chapter four presents a critical review of the Historical, Clinical, Risk–20 Violence Risk Assessment (HCR-20) (Version 2), a widely adopted risk assessment framework utilised within forensic psychiatry and the standardised measure used within the empirical study. The review explores the literature on the reliability and validity of the HCR-20, and considers its strengths and limitations. A discussion of the work presented concludes the thesis