The Joint Commission’s new National Performance Goals chapter for hospitals and critical access hospitals became effective January 1, 2026, and it carries forward National Performance Goal 2a on preventing workplace violence. The chapter replaces the former National Patient Safety Goals chapter and reorganizes 14 measurable topics, but the workplace violence requirements are not new. They are the four standards the Joint Commission made effective January 1, 2022: EC.02.01.01 (annual worksite analysis of violence risk), EC.04.01.01 (continuous incident data collection and reporting), HR.01.05.08 (training at hire and annually), and LD.03.01.01 (a prevention program led by a designated individual and a multidisciplinary team). The 2026 chapter renumbers and re-elevates them. For self-insured hospitals, that re-elevation is the reserving event, because a mature reporting and data-collection apparatus changes what shows up in the workers compensation loss run before it changes the underlying risk.
The standard requires hospitals to capture, code, and trend violence events and to report them to the governing body. The Joint Commission’s workplace violence page notes that hospitals have received more than 100 related requirements for improvement since January 2022, with correction required within 60 days of survey activity. That enforcement pressure pushes events out of security logs, patient safety systems, and human resources files and into a counted, reportable stream. None of that creates new assaults. It creates new visibility, and visibility is what drives reserves.
Why the exposure is large to begin with
Healthcare carries the highest workplace violence burden of any sector. BLS Survey of Occupational Injuries and Illnesses data put the rate of intentional injury by another person at 10.4 per 10,000 full-time workers in health care and social assistance, against 2.1 for all workers, roughly five times the all-worker rate, with general hospital employees higher still. Healthcare workers absorbed roughly three-quarters of all nonfatal workplace violence injuries requiring days away from work. The trend is the harder part. A 10-year analysis of BLS data found violence rates at general medical and surgical hospitals rose 158% from 2011 to 2021-2022, and that psychiatric and substance-abuse hospitals run at 110.4 per 10,000 FTE, an order of magnitude above the hospital average. A separate Press Ganey analysis of its nursing-quality database found more than two nursing personnel assaulted every hour in the second quarter of 2022, about 57 per day, concentrated in psychiatric units, emergency departments, and, notably, pediatrics. The NCCI’s workplace assault series frames the same trend on the claims side: assault rates per 10,000 FTE up 62% from 2011 to 2022, with healthcare carrying roughly ten times the exposure of the next-highest sector.
Who It Affects
Self-insured health systems, public hospitals, academic medical centers, county hospitals, behavioral health facilities, and hospital captives that retain workers compensation risk. The highest-priority loss-run review is not the whole hospital payroll class. It is nurses, technicians, security staff, emergency department employees, behavioral health units, transport teams, and inpatient units with frequent patient or visitor escalation, the same units the facility data flags as outliers.
OSHA’s workplace violence page still states there are no specific OSHA standards for workplace violence, while pointing employers to worksite assessment, prevention programs, engineering and administrative controls, and training. That split matters for reserving. Accreditation files now develop faster than OSHA rulemaking, so a severe assault claim may carry both workers compensation exposure and a written record of prior hazard identification, leadership review, or delayed corrective action, the kind of documentation that can raise both indemnity and defense reserves.
Reserve mechanism
The first lever is frequency, and it points up for a measurement reason before any safety reason. Better capture raises reported counts in the short run, especially where prior events sat outside the workers compensation coding stream. A 2026 frequency increase at an accredited hospital may be a coding-completeness change layered on top of a sector trend already running near 5% to 6% annually, not a sudden deterioration in safety. The two are easy to confuse and expensive to confuse.
The second lever is severity and case adequacy. Assault claims do not develop like strain, slip, or needle-stick claims. They carry a psychological trauma component, PTSD, anxiety, and delayed or failed return to work, that lengthens treatment and widens the gap between the first case reserve and the ultimate. Lost-time claims already drive the overwhelming majority of WC loss dollars, and assault claims skew toward lost-time. Where the file also contains accreditation findings or prior similar incidents, the claim warrants a stronger case reserve earlier. State benefit changes amplify this: Connecticut’s Public Act 26-12 lifts wage replacement to 100% for assaulted healthcare workers, roughly a one-third indemnity bump on a claim type already trending up.
The third lever is development pattern, and it is where the IBNR error hides. If new capture surfaces many lower-severity events, reported counts rise while average severity falls, and naive application of historical link ratios over-reserves. If instead it surfaces a backlog of serious underreported events, both reported loss and incurred development rise, and the same link ratios under-reserve. Either way, factors calibrated on a period when assaults were a smaller, less completely coded share of the mix will misstate the tail on 2024 through 2026 accident years. For the mechanics, see workers compensation IBNR for self-insured employers; the parallel signal on the hospital professional liability side is that the same incident files now feed two reserve lines, and staffing-driven liability exposure sits next to it on the same units.
The call
The directional read is that hospital WC assault frequency rises in 2026 reporting at accredited facilities, and that the rise is part measurement and part real, in proportions the loss run alone cannot settle. Reserves should move, but selectively. Treating a coding-completeness jump as pure trend over-reserves and overstates the rate need; ignoring it under-reserves a line where severity development is genuinely lengthening. The most exposed are self-insured behavioral health, emergency, and psychiatric operations, where base rates already run an order of magnitude above the hospital average and where the new data apparatus will surface the most previously uncounted events. The defensible position is to carve assault claims out of the blended triangle, give emergency and behavioral health units their own frequency and severity assumptions, and separate coding cleanup from true trend before moving the selected ultimate. The practical test is unchanged: if the safety dashboard tells one story and the workers compensation triangle tells another, reconcile the systems before booking the difference as loss trend.
Sources
- Joint Commission, National Performance Goal 2a: Preventing Workplace Violence
- Joint Commission, National Performance Goals
- The Joint Commission’s New and Revised Workplace Violence Prevention Standards for Hospitals (PMC, 2022)
- Trends in Workplace Violence for Health Care Occupations and Facilities Over the Last 10 Years (PMC, 2024)
- BLS, Workplace Violence in Healthcare, 2018
- Press Ganey: On Average, Two Nurses Are Assaulted Every Hour (2022)
- OSHA, Workplace Violence
- OSHA, Guidelines for Preventing Workplace Violence for Healthcare and Social Service Workers