The Structured Interview for Violence Risk Assessment (SIVRA-35): Structure and Campus Use
Behavioral intervention teams on college campuses run into a version of a problem every prevention field shares. The person who first hears a worrying report, a resident advisor, a conduct officer, a professor, is usually not a clinician and has no formal training in violence risk. Research the U.S. Secret Service and Department of Education produced after studying school attacks found that attackers rarely erupt without warning. They tend to move along a pathway: developing a grievance, considering violence, then planning and preparing, often leaving behavioral signs along the way for someone to notice. The practical challenge for a campus team is turning that research into something a non-clinician can actually use in the room.
What SIVRA-35 Measures
The Structured Interview for Violence Risk Assessment, known as SIVRA-35, was developed by Dr. Brian Van Brunt at the NCHERM Group to meet that need. It is a 35 item structured interview built for behavioral intervention team members and campus staff. It is explicitly not a psychological test, and it is not designed to assess suicide risk on its own. An administrator works through the items in conversation with the person of concern, or reviews incident reports and other records, and determines whether each item applies. A separate numeric scoring key turns those determinations into points.
Each item reflects a specific warning behavior drawn from decades of threat assessment research: a direct communicated threat, fixation on a target, leakage of a plan to other people, access to weapons, a grievance the person cannot let go of, and more. The items sum to a score between 0 and 70, with ranges marking low, moderate, and high risk. The first twelve items are critical items that carry an additional scoring rule: if four or more of them are marked non-zero, the case is placed in the high category regardless of the overall total. That rule reflects something threat assessment researchers have documented repeatedly: certain warning behaviors, a direct threat paired with a concrete plan, for example, matter more than an average score would suggest.
A number on its own tells a team very little. What matters is which specific behaviors are present, and whether the picture as a whole points toward planning and preparation rather than distress alone.
Where a Structured First Look Fits
SIVRA-35 was built for one setting, a college campus behavioral intervention team, and it does that job well. Homicide Zero's Homicide Threat Screener, or HTS, occupies a similar role across a wider range of settings: schools, workplaces, faith communities, anywhere a non-clinician is often the first person to notice a concern. Both instruments exist to solve the same underlying problem: someone without violence risk training needs a structured way to organize what they are seeing before deciding whether the situation needs to go further. For the HTS, that structured look takes five to ten minutes.
Where an HTS screen or a SIVRA-35 interview lands in the moderate or high range, the next step is not for the same person to keep guessing. The Violence Safety and Risk Assessment, or VSRA, is a deeper 20 to 30 minute assessment built for a trained professional to take over from there. This is the same logic SIVRA-35 itself is built on: an initial structured look by whoever is present, then a deliberate handoff to someone trained when the picture is not clearly low.
The Prevention Point
No instrument, SIVRA-35, the HTS, or the VSRA, can predict who will become violent. None of them claim to. What a structured interview does is make sure the same categories of information get asked about and recorded every time, instead of depending on whether the staff member in the room that day happens to know what to look for. That consistency is the actual prevention value: a concern that gets structured and documented is a concern a team can act on, escalate, or reassess later. A concern that stays an informal impression usually just fades until the next one arrives.