Response to Letter to the Editor,“Measurement of Workplace Violence Reporting”

We are grateful for the thoughtful comments on our recent article, “Underreporting of Workplace Violence: Comparison of Self-Report and Actual Documentation of Hospital Incidents” (Arnetz, Hamblin, Ager, et al., 2015). The writer raises issues that are well worthy of discussion. However, on a number of points, there has been some misunderstanding that we hope this response will clarify.

Conclusion. The writer states that our article concluded “ . . . that hospital employees under-reported incidents of workplace violence mainly because non-victims (witnesses) tended to report incidents far less often than victims” (Huang & Glenn, 2016). This was not the main conclusion of our article and is factually incorrect. A closer look at Table 3 (p. 206) reveals no statistically significant difference between reporters and under-reporters with regard to being a target or witness of violence. Our main conclusions in this article were (a) the majority of hospital workers who experienced violence at work (88%) did not record the incident in the hospital system’s electronic database, (b) nearly half of hospital workers who experienced violence (45%) did report the incident to their supervisors, (c) hospital workers who were injured in a violent incident and/or lost time from work due to a violent incident were significantly more likely to report using the electronic system (Table 4), and (d) workers with less than 5 years job tenure and security staff were significantly less likely to report a violent incident to supervisors (Table 5).

Unclear definitions of violence. Our survey did, in fact, provide the following general definition: “In this survey, ‘violence’ includes acts or threats of physical or verbal aggression.” We then asked, “Have you been a target of violence or aggression at work during the past year?” Thus, we clearly delineated between physical violence and verbal aggression. We believed that the juxtaposition of “physical” and “verbal” would suffice as indication to the questionnaire respondents that we were using a broad definition of “violence.” Moreover, in the item asking “What type(s) of violence/aggression did you experience?” the first response alternative was “Verbal aggression (shouting, swearing).” The writer posits that “the two forms of this question would be expected to produce different responses,” but in truth, we could not be sure that responses would differ. Workplace violence is subjective; what may be perceived by one worker as “violence” may be perceived by another as lower-level aggression (Arnetz, Arnetz, & Petterson, 1996). We recognize that perception may influence reporting, which is why we specifically combined physical violence and aggression in a single question.

Low reliability for the participants’ determination of whether they were a target of workplace violence. The question on whether the individual had been a target of violence or aggression during the past year was followed by two additional questions: “Who was violent or aggressive towards you?” and “What type(s) of violence/aggression did you experience?” Both items had “I wasn’t a target of violence” as a possible response alternative. However, questions regarding why employees did not report a violent incident used the response alternative, “I wasn’t a target or witness of violence.” We believe that this may be the source of the writer’s confusion. In other words, when asking about experience of violence at work, we asked only whether employees had themselves been the target of violence. When asking why they had not reported an incident, we included the option that they had not even witnessed such an incident. The reason the questions were structured this way has to do with hospital system policy, as indicated on page 202 of our article. Hospital system policy actually mandates that employees report any known incidents of violence, either through the electronic system or to a supervisor. Policy does not specify that an employee must be a target of violence to report the violent episode; rather, any “known incident of violence” should be reported. In previous research on this hospital system, we found that incident reporting by third parties not directly involved in the violent incident does occur (Arnetz, Hamblin, Essenmacher, et al., 2015).

Another possible explanation is faulty recall due to the lengthy recall period of 1 year. We agree with the writer that recall bias is a possible confounder of this study, a point we raised ourselves in the “Strengths and Limitations” section, page 208. As justification, the survey, part of a large, randomized controlled intervention study, was administered pre-intervention and 1-year post-intervention. Thus, asking about experience with workplace violence over the course of the previous 12-month period was a deliberate effort on our part to compare self-report with documentation of incidents over the same period of time.

The estimate of 88% is likely exaggerated. The writer contends that “an incident could have been witnessed by multiple employees but reported only once by one person in the electronic reporting system. If an employee knows that someone else reported it, they may decide that this is sufficient” (Huang & Glenn, 2016). There is nothing in our data to support this statement. The writer further posits that the statement is supported “by the study conclusion that employees tended not to report incidents if they were not the victims of violence themselves.” As indicated in the point above regarding low reliability, this is a misunderstanding that we hope we have now clarified.

It is probably unreasonable to conclude a failure-to-report problem exists when reporting is not mandated. This explanation would also be expected to inflate the failure-to-report percentage. As clarified in the low-reliability point above, reporting acts of known violence is, in fact, mandated by the hospital system. The writer is correct in that reporting via the electronic system is encouraged but not required. However, the mandate states that an employee should report either via the system or directly to one’s supervisor. As we point out on page 202, unit supervisors also have a mandate to report all violent events of which they are notified into the electronic system within 24 hours from the end of the shift during which they received the notification. Thus, theoretically, incidents should ultimately be entered into the electronic database. Our study indicates that electronic database reporting occurs in only 12% of cases.

. . . the value of the electronic reporting system is low among hospital employees, particularly nurses. We can agree that the electronic reporting system is underused for reporting incidents of workplace violence, and among nurses, there were more under-reporters (62.6%) than reporters (40.6%, Table 1). However, other, more far-reaching implications of this study’s findings exist. As we stated on page 208, although employees who report verbally to their supervisors may be fulfilling their responsibility for reporting, the informal reports may not be reaching upper management for policy decisions. A hospital system or other health care organization can only develop prevention strategies based on available data. By implementing the electronic system, this particular hospital system has made a concerted effort to facilitate the employee reporting process. Our data indicate that the system is underused.

The purpose of our article was to increase understanding of underreporting by investigating differences between self-report and actual documentation practices, and exploring the characteristics and reporting patterns of health care workers who underreport. To the best of our knowledge, our study is the first to accomplish these aims by linking documentation behavior and self-report to individuals via de-identified ID numbers. Hospital employees in this study did drastically underreport workplace violence, both electronically and by other, less formal means. Our article identified certain worker characteristics associated with both electronic and informal reporting of violent incidents. It is our hope that knowledge of this phenomenon will help occupational health nurses and health care organizations improve incident reporting, as such data are the first step toward preventing workplace violence.