Workplace Violence and Harassment Against Emergency Medicine Residents

Benjamin H. Schnapp, MD; Benjamin H. Slovis, MD; Anar D. Shah, MD, MBA; Abra L. Fant, MD, MS; Michael A. Gisondi, MD; Kaushal H. Shah, MD; Christie A. Lech, MD


Western J Emerg Med. 2016;17(5):567-573. 

In This Article


The study was based on self-reported survey responses, so the results may be vulnerable to recall bias; as with all surveys, there is no way to objectively verify the accuracy of the reported incidents. This was a cross-sectional study and therefore represents only one moment in time. The time of year of the survey, as well as several well-publicized incidents of violence in the depertment in the months before the survey may have influenced the residents' recall. This study asked residents at all stages of their training to report their exposure to violence without controlling for the number of months spent in the ED (due to PGY year or off-service rotations), which may have affected reported levels of violence. This survey was provided to residents at urban, academic hospital centers and may not be representative of other hospital or clinical settings with a different patient populations and less alcohol/substance abuse or psychiatric disease. However, given the number and variety of New York hospitals surveyed, the results are likely representative of the experience in the city at large. The response rate for this study was 84%. We were unable to account for demographic information for those residents who did not participate in the survey, as we did not collect data from this group. However, we did attempt to compare the demographics of those who completed the survey with the residency populations as a whole based on each residency's respective website. While participation holds the potential for selection bias, we presume a lack of participation was most likely due to missing the ABEM ITE. Missing the ABEM ITE was most likely due to clinical responsibilities or vacation; therefore, those who did not respond to our survey were unlikely to be systematically different from our respondents, and our analysis of the demographics of the residency websites seems to confirm this. It is possible that residents who responded to the survey may have had experienced a greater or fewer number of episodes of violence than the population at large, but we consider systematic bias unlikely. While we attempted to include all of the most commonly cited reasons for ED violence in our survey, it is possible there were other reasons missed by our instrument. Additionally, our modified survey instrument has not been validated for use in the ED population, though modifications for our study were minimal.