Dynamic relationship between multiple START assessments and violent incidents over time: a prospective cohort study

Abstract

Background: Dynamic risk factors need to be assessed repeatedly over time rather than at a single time point to examine the relationship with violence. This predictive validity study sought to examine the degree of dynamic change in risk assessed in a group of mentally disordered offenders and the relationship between change and the occurrence of violence. Methods: Routine structured assessments of Strengths and Vulnerabilities on the Short-Term Assessment of Risk and Treatability (START) instrument (n = 475) were linked prospectively with 275 violent incidents using logistic regression in a sample of 50 patients. Results: Stability within patients estimated using the intra-class correlation coefficient was high (>.80) for both Strengths and Vulnerabilities. In the overall sample, a 10 point increase in START Vulnerabilities score was associated with a three-fold increased risk of violence (OR = 3.1; 95% CI, 1.47-7.46) but there was no association for Strengths score (OR = 0.91, 95% CI, 0.34-2.47). When examined within patients, both Vulnerabilities (OR = 1.77, 95% CI, 0.56-5.54) and Strengths (OR = 2.26, 95% CI, 0.38-13.42) were associated with an increased risk of violence but in both cases precision was low due to reduced sample sizes. Conclusions: Risk factors which are considered to have the capacity to fluctuate dynamically did not do so substantially in this group of mentally disordered offenders. When fluctuations did occur there was some tentative evidence that they are associated with violent outcomes and could guide the use of prevention measures.