Targeting Institutional Risk Factors to Reduce Patient Violence

Jennifer Skeem, University of California, Berkeley

Barbara McDermott, University of California, Davis

David Cooke, University of Bergen

Research Grant, 2018

---

BACKGROUND: In forensic institutions, staff supervise and treat distressed adults who are involved in the criminal justice system, detained against their will, and restricted in their activities. Here, violence is common, costly, and serious. Current violence-prevention strategies focus on identifying and managing a small group of high-risk patients and are limited in their effectiveness.

According to leading theories and some evidence, strategies must be broadened to target situational or institutional risk factors (e.g., authoritarian relationships, crowding) to maximize violence prevention. Institutional risk factors—particularly those that engender a sense of injustice, disrespect, and deprivation—can provoke violence even among low- to moderate-risk patients.

This project shed light on situational risk factors featured in violence theories, and provided some empirical guidance on integrating organizational and individual approaches to maximize violence prevention.

AIMS: The goal was to assess whether targeting institutional risk factors (a population approach) adds value to violence-prevention efforts that target individual patients (a high-risk approach). Primary aims were (1) to identify institutional risk factors that most robustly predict violence rates (and are relevant both to violence theories and prevention); and (2) to test whether interventions that specifically target a unit’s institutional risk factors are feasible to implement and add value in reducing the unit’s risk level and violence rates. Given that we were unable to complete Aim 2 because of the COVID-19 pandemic, the remainder of this summary focuses primarily on Aim 1.

METHOD: University investigators partnered with the Napa site of the California Department of State Hospitals (DSH-Napa) to conduct two phases of study.

In the Phase 1 Assessment, researchers conducted multi-informant, multisource assessments of twenty-six units at DSH-Napa, using the PRISM risk-assessment instrument and other well-validated measures of institutional risk factors. Assessments included semistructured interviews with 262 patients and staff (ten per unit).

In the Phase 2 Follow-up, researchers obtained data on rates of violence on each unit over a three-month (“short-term”) and twelve-month (“long-term”) follow-up period. Analyses focused on testing the association between institutional risk factors assessed at baseline and future rates of violence on units.  

RESULTS:  First, institutional risk factors strongly predicted unit-level variation in violence. For example, units’ risk classifications on the PRISM instrument (low, moderate, or high) strongly predicted units’ violence rates one year later (r= .78). Second, specific factors that strongly predicted future violence included a unit’s historical rates of violence; limited violence-relevant interventions; problems with the physical environment; and limited staff training, competencies, and accessibility. Third, patient perspectives on risk factors were particularly predictive—notably patients’ perceptions of authoritarian staff-patient relationships, a nontherapeutic social climate, and unsatisfactory services. 

SIGNIFICANCE: Violence is an ongoing and serious problem in forensic institutions. Stakeholders need strategies that can better protect staff and patients from harm, improve morale, and enhance the therapeutic milieu. This project shed light on situational risk factors featured in violence theories, and provided some empirical guidance on integrating organizational and individual approaches to maximize violence prevention. DSH-Napa is now testing the effectiveness of an intervention approach that focuses on addressing institutional risk factors for violence in a pilot randomized controlled trial.