By Gary Evans, Medical Writer

Although there are cautions and caveats about generalizing the data, occupational health researchers are tracking a disturbing increase in violence in a network of surveillance hospitals.

Researchers with the National Institute for Occupational Safety and Health (NIOSH) analyzed workplace violence injury data from hospitals participating in the Occupational Health Safety Network (OHSN) from 2012 to 2015.1 Overall, 106 participating hospitals reported a 72% increase in workplace violence injuries. The rate went from 4.4 injuries per 1,000 full-time equivalent (FTE) workers in 2012 to a high of 7.6 per 1,000 FTE. The rate declined to 7.2 in 2015.

The study did not assess the reasons for violence, but cited patient factors such as the increasing prevalence of substance abuse, mental illness, dementia, and other conditions as likely contributing causes. In addition, workplace factors like understaffing, high turnover, and long patient wait times can exacerbate the situation.

Applying a different statistical measure called an “adjusted workplace violence injury rate,” the NIOSH researchers reported that injuries due to violence increased by an average of 23% each year. That is a rather staggering finding, and there are caveats to be cited, but it bears noting at the outset that the researchers think the increase is real.

“The caveats are important, but even given them, we are relatively confident that what we are seeing among this group of hospitals is something that is also being seen nationally,” says Matthew Groenewold, PhD, MSPH, lead author of the NIOSH study. “Workplace violence incidents do seem to be increasing in hospitals generally in the U.S., and we definitely see it in the data from our OHSN hospitals.”

Caveats include that the participating OHSN hospitals represent a “self-selected” group, and not a randomized sample representative of hospitals nationally, he says.

“But one reason that we think we are seeing a true trend — an increase — is that if you look at nationally representative data from the Bureau of Labor Statistics, you see a similar trend,” he says.

Nurses and nursing assistants were the primary victims of violence, with the majority of assaults coming from patients. The latter is based on the finding that the recorded assailants were almost always patients, but the problem is that the incidents were not always completely documented by the hospitals.

“The OHSN hospitals have tools for data collection, with sort of a minimum core set of variables that ought to be collected,” Groenewold says. “But we found that data the participating hospitals submitted to us often had either missing or unknown values for a number of these variables.”

Furthermore, it was not clear from the survey whether the missing information was simply unknown or was known but not reported.

“In order to do an effective job of prevention, you really have to have a good handle on what is going on,” he says. “Collecting and routinely analyzing surveillance data in order to monitor how frequently these workplace violence incidents happen — and also the characteristics associated with them — is really important to prevention.”

NIOSH is recommending just that to its OHSN hospitals, asking for further investigations of incidents to provide the specifics if possible.

“The occupational health department or whoever else is investigating should actually go out and collect the rest of the data,” Groenewold says. “The flip side of that may be that in a lot of cases where these variables were left blank or marked unknown, the information was known and it just wasn’t recorded.”

The recommendation in that case is to make improvements in reporting procedures or informatics to remove barriers and facilitate the complete entry of the known data.

As cited in the paper, “OSHA defines a work-related injury or illness as recordable if it results in death, unconsciousness, days away from work, restricted work, transfer to another job, or requires medical treatment beyond first aid.”

“From the OHSN perspective, we only analyze data for OSHA recordable incidents,” Groenewold says. “We don’t mean to say by that, the [other incidents] are not important. Hospitals certainly can record those data. We would not discourage them from recording those data and analyzing it.”

A total of 3,263 violent events were reported by the hospitals. By occupation, nurses in OHSN-participating hospitals sustained most workplace violence injuries (40%), followed by nonpatient-care personnel (33%) and nursing assistants (20%).

Within outpatient care in OHSN-participating hospitals, the most common location of workplace violence injuries was the ED (19%). Beyond the ED, most workplace violence events occurred in patient rooms (53%), followed by corridors, elevators, and stairwells (10%), and examination rooms (6%).

“Among the 540 injuries where the severity was reported, 261 (48%) resulted in lost work days, job restrictions, or transfers,” NIOSH reported.

Nurse Aides at Risk

Pharmacists and physicians had the lowest rates of workplace violence events. Compared to all other healthcare workers, nursing assistants and nurses experienced higher rates of workplace violence. The study also suggests the possibility of an “underappreciated disparity” in workplace violence injury rates between nurses and nursing assistants who work in hospitals.

The excess risk for nurses for workplace violence injuries is well-documented in the literature, Groenewold says. Similarly, reports clearly indicate that nurse assistants in long-term care settings are at heightened risk of violence.

“Other than a couple of studies that we cite in the paper, there has not been a lot in the literature describing the excess risk to nursing assistants working in hospitals,” he says.“We found that, although nurses account for most workplace violence injuries, nursing assistants in hospitals have a much higher rate of injury. Their risk really seems to be higher.”

One explanation is that hospital nurses work in administrative, educational, and other duties that are not directly involved in patient care. Such is not usually the case with nursing assistants, who spend much of their time in patient-care duties.

“The fundamental mechanism underlying both nurses’ and nursing assistants’ high workplace violence injury rates is likely to be their more frequent, prolonged, and direct exposure to patients compared with other hospital workers,” the authors concluded.


1. Groenewold MR, Sarmiento RFR, Vanoli K, et al. Workplace violence injury in 106 US hospitals participating in the Occupational Health Safety Network (OHSN), 2012-2015. Am J Ind Med 2018;61:157–166.