Metrics & Leadership
Classifying WVP Incident Severity Defensibly
A consistent severity scale turns a raw incident log into trendable, board-grade data. Here is how to classify workplace violence severity so it survives a survey and a deposition.
A raw incident count answers "how many" but not "how bad," and leadership needs both. Classifying every workplace violence report against a written, consistently applied severity scale turns a flat log into data you can trend, target, and defend — letting a board see whether incidents are growing more or less serious, not just more or less frequent. The scale matters more than the labels.
This article supports our pillar, the metrics every hospital board should see. It is written for the risk manager, quality lead, and safety director who own the incident taxonomy and keep getting asked whether a "spike" reflects more violence or just more verbal complaints.
#Why an unclassified count misleads
Two facilities can report the same number of incidents while living in completely different realities. One logged forty verbal complaints; the other logged forty assaults that put staff in the ED. A count that flattens those into a single number hides exactly the signal leadership needs to allocate resources and that a surveyor expects you to be tracking.
Severity classification solves three problems at once:
- It makes the trend meaningful. A rising count of near-misses while injuries fall is a successful program surfacing early signals. Without severity, that good news reads as bad news.
- It targets controls. You cannot prioritize the units and event types that actually injure staff if every event weighs the same.
- It connects to recordkeeping. Your top severity tier should map cleanly to OSHA 300 Log recordability, so your internal data and your regulatory log reconcile rather than diverge. See the OSHA 300 Log and workplace violence.
#A defensible four-tier severity scale
Most healthcare WVP programs converge on four tiers. The exact labels are yours to set; the discipline is that the definitions are written, communicated, and applied the same way by every reporter. Anchor the classification to observable facts — what happened — not to the reporter's emotional read of it.
| Tier | Definition | Typical examples | Recordkeeping link |
|---|---|---|---|
| 1 — Near-miss / precursor | An act or condition that could have caused harm but did not | Thrown object that missed, an averted physical confrontation, a credible warning behavior | Internal log only |
| 2 — Verbal / non-contact | Verbal threat, harassment, intimidation, menacing without physical contact | Threats to a clerk, intimidation at triage, verbal abuse during discharge | Internal log only |
| 3 — Physical, no injury | Physical contact or assault that did not result in a recordable injury | Grabbed, pushed, struck without injury beyond first aid | Internal log; review against 1904 criteria |
| 4 — Physical, with injury | Assault meeting OSHA recording criteria | Injury with days away, restricted duty, or treatment beyond first aid | OSHA 300 Log under 29 CFR 1904 |
The boundary that demands the most rigor is the Tier 3 / Tier 4 line, because Tier 4 is where your internal data must reconcile with the OSHA 300 Log. Define Tier 4 by the 1904 recording triggers — days away, restricted duty or transfer, medical treatment beyond first aid, loss of consciousness, or a significant diagnosis — so a Tier 4 classification and a 300 Log entry rise and fall together. A Tier 4 incident with no matching 300 Log entry is a recordkeeping gap an inspector will find.
#Classify on facts, not feelings
The most common reason a severity scale produces inconsistent data is that different reporters grade the same event differently based on how shaken they were. Two controls fix this:
- Define each tier by observable criteria — contact or no contact, injury or no injury, recordable or not — so two reporters reach the same tier for the same facts.
- Add a brief secondary review. A single reviewer (the program leader or a designee) confirms or adjusts the assigned tier before the incident closes, which standardizes the line calls that matter most. This is also where you capture intent and type — the four standard workplace-violence types — alongside severity, since type and severity are different axes.
Pairing severity with type, unit, and shift is what produces the cross-tabs a board and a surveyor find useful: Tier 4 events concentrate on the night shift in the ED, for instance, is an actionable finding that a single number can never surface. For how these definitions and their denominators hold up under scrutiny, see defining WVP KPIs with denominators that hold up.
#Reporting severity to leadership
On a board scorecard, severity is most legible as a stacked trend — the count by tier over time — paired with a short narrative. That view answers the question a count cannot: are we seeing more events, or more serious events?
- A rising Tier 1–2 share with flat or falling Tier 4 often signals an improving reporting culture combined with effective intervention — staff are surfacing precursors, and fewer are becoming injuries.
- A rising Tier 4 share is an unambiguous escalation signal that warrants immediate corrective action, regardless of total volume.
Always read severity trends alongside your reporting-capture context, because a shift in the tier mix can reflect changed reporting behavior rather than changed risk. For the broader trend discipline this feeds, see trending incident data the way Joint Commission expects.
#Common mistakes
- No written definitions — "severe" means whatever the reporter felt, and the data cannot be trended.
- Logging only injuries — discarding near-misses and verbal events throws away the early warning the program needs most.
- A Tier 4 / 300 Log mismatch — internal severity and OSHA recordkeeping that disagree is a credibility and recordkeeping gap.
- Confusing severity with type — they are different axes; capture both, never one in place of the other.
- Grading on emotion — without observable criteria, the same event lands in three different tiers across three reporters.
#How VIGILO helps
VIGILO builds a written, four-tier severity taxonomy into the incident report form and the trending system, aligns the top tier to OSHA 300 Log recordability, and folds the severity trend into a board-ready report. This is compliance and survey-readiness assistance, not a guarantee of any safety outcome, and VIGILO operates strictly as a compliance, training, and consulting firm.
To put a defensible severity scale in place, start with a flat-fee survey-readiness audit, or maintain it through an annual program review. To see who relies on this data, visit who we serve.
Sources: Texas Health & Safety Code Chapter 331 (SB 240, 88th Leg., 2023); OSHA General Duty Clause §5(a)(1), Publication 3148, and recordkeeping rule 29 CFR 1904; The Joint Commission Workplace Violence Prevention requirements (effective Jan. 1, 2022 for hospitals); BLS 2018 incidence data via NIOSH/CDC. This article is general compliance information, not legal advice.