Relationship between insight and violence in psychosis: a systematic review & The predictive validity of the HCR20v3 within Scottish forensic inpatient facilities: a closer look at key dynamic variables
Smith, Kerry Johanna
MetadataShow full item record
Introduction: Poor insight is included as a risk factor for violence in risk assessment tools such as the Historical Clinical Risk-Management-20 version 3 (HCR-20v3) yet there is a lack of consensus around the relationship between poor insight and violence in individuals with psychosis. A systematic literature review was therefore carried out to clarify this relationship. Relatedly, a research project aimed to outline the predictive validity of the HCR-20v3 total and sub-scale scores to violence in forensic inpatients. A secondary aim was to understand the predictive ability of 2 dynamic risk factors within the HCR-20v3 clinical sub-scale; insight and positive symptoms, alongside age and history of violence in relation to violence in psychosis. Method: A systematic search of studies investigating insight and violence in patients with psychosis, published between 1980 and 2016 was carried out on relevant databases.17 articles from combined search results of 5694, met the inclusion criteria. These were selected for full-text review and quality grading which was subject to inter-rater reliability. In the research project, the predictive validity of the HCR-20v3 to violence was assessed in N=167 forensic inpatients. A sub-sample of N=135 was then used to investigate insight, positive symptoms, age and history of violence in relation to violence. Data was extracted from case files, with the exception of violence data which was collected prospectively from date of HCR-20v3 publication via DATIX. Results: The systematic review found 8 studies in support of a positive relationship between poor insight and violence, whilst 9 studies did not support this relationship. The majority of better quality studies measured the clinical insight dimension which tended to demonstrate a positive relationship between poor insight and violence. Methodological limitations were apparent across studies. The research project found HCR-20v3 total and clinical and risk-management sub-scale scores to predict violence. The clinical sub-scale was the strongest predictor of violence and physical violence specifically. Sub-sample analysis found positive symptoms and history of violence to significantly predict violence generally whilst only positive symptoms demonstrated prediction of physical violence. Insight and age were not significantly associated with either violence type. Discussion: The systematic review found partial support for a positive relationship between poor insight and violence in psychosis. Future good quality research is required to develop a fuller understanding of this issue. Research project results support the use of the HCR-20v3 in the risk assessment and management of forensic inpatients. They reinforce the usefulness of dynamic risk factors within the clinical sub-scale in particular. In line with the majority of studies within the systematic review however, a relationship between insight and violence in a sub-sample of patients with psychosis was not found. Recommendations are made for the regular re-assessment of dynamic risk factors within the HCR-20v3 clinical sub-scale in order to support patients to reduce their level of risk, with the caveat that future research is still required to support a relationship between insight and violence in this patient group.