Respondents were also asked several questions about how they personally have been affected by HV over the last 12 months at their current workplace. Overall personal effects were reported to occur from a few times to monthly in that period (N = 951, mean 3.67, SD 1.3) (Cronbach's alpha 0.88). The most frequently reported personal effect was: “I've felt discouraged because of lack of positive feedback” (N = 944, mean 4.35, SD 1.5). (See Mean score for frequency of personal effects of HV.)
Perpetrators of HV. The most frequent perpetrators of HV were reported to be nurse peers (RNs or LPNs) with a mean of 4.67 (SD 1.7). Supervisors scored next highest (mean 4.2, SD 1.5) and then unlicensed assistive personnel (mean 3.84, SD 1.7). Physicians scored a mean of 3.4 (SD 1.6). Other professionals (mean 2.7, SD 1.5) and workers such as housekeeping, security, and maintenance (2.83, SD 1.7) came in with scores that could be interpreted as once to a few times. (See Mean Score for frequency by perpetrator of HV behaviors.)
Analysis of Relationships. The following paragraphs discuss relationships between nurse characteristics and frequency of HV and its effects. This sample included some statistically significant findings; however, as you'll see, no strong relationships emerged.
- Overall HV and personal effects to years in nursing. No relationships were found between years in nursing and frequency of personal effects of HV. The data were entered into SPSS 20. Correlation statistics were run on continuous data and chi-square tests were run on nominal data. In this sample, we found a slight inverse relationship between years in nursing and frequency of overall HV experienced/witnessed (N = 891, Pearson's r = -0.084, P = 0.012). However, Pearson's r is a correlation statistic and correlations go from +1 to 0 to -1. Any r value of less than ±0.3 is generally considered so small that there is practically no relationship. In other words, even though there was statistical significance, it may not be meaningful clinically. These data were also analyzed with Spearman's rho, and the results were the same.
- Relationships to gender. Many more females (879) were included in the study than males (50). However, the variances in the data were equal. T-Tests were run on differences between frequency of HV experienced or witnessed and frequency of personal effects of HV by gender. We found that compared to females, male nurses experienced higher frequencies of both the behaviors of HV (4.8, SD 0.9 versus 4.5, SD 1.1, P = 0.04) and the personal effects (4.1, SD 1.2 versus 3.6, SD 1.3, P = 0.009).
Chi-square tests were run, revealing that 84% of male nurses stated they experienced at least one personal effect from HV more than monthly, compared to 63% of females. This was statistically significant (Pearson chi-square 8.786, P = 0.003).
The two most common questions male nurses answered with high frequencies were “left work feeling discouraged because of lack of positive feedback” (mean 4.92, SD 1) and “left work feeling bad about myself because of interactions with coworkers” (mean 4.3, SD 1.4).
- Relationships to age of nurse. Age data were collected as age ranges, not actual age. In general, nurses in all age ranges reported experiencing/witnessing overall HV an average of between monthly and weekly (4.5, SD 1.1). Nurses in the 41 to 50 age range reported the highest frequency (4.6, SD 1.2). All age ranges also reported similar frequencies for personal effects of HV (3.7, SD 1.3) that can be interpreted as between a few times a year and less than monthly. Interestingly, the youngest nurses reported the lowest frequency of personal effects (3.2, SD 1.3).
These data can be considered categorical data. Analysis of variance (ANOVA) tests were run to determine differences in frequency of experiencing/witnessing HV and in frequency of personal effects of HV based on age category. (See Differences in reported frequencies of HV and personal effects of HV by age group.) Bonferroni post hoc tests identified that the only statistically significant difference in frequency of experiencing/witnessing HV was between the age 41 to 50 group (N = 245) and the over-60 group (N = 117), with the older group experiencing/witnessing HV less frequently (mean of 4.61, SD 1.2 versus 4.21, SD 1.2, respectively, P = 0.015). For frequency of personal effects of HV, we found statistically significant differences between the youngest nurses (21 to 30) having a lower reported frequency than two of the older age ranges of nurses (41 to 50 and 51 to 60), with a mean of 3.18, SD 1.3 versus 4.61, SD 3.87, P < 0.001, and 4.53, SD 1.4, P = 0.004.
- Relationships to education and certification. The means for experiencing/witnessing and personal effects of HV were similar across all educational levels. The highest means for both experiencing/witnessing and personal effects of HV were from the PhD or other doctoral level at 4.79, SD 0.9 and 4.29, SD 1.5, respectively. However, these results should be interpreted with caution as the sample included only 14 respondents.
The lowest reported means were from the MSN group, 4.19, SD 1.2 for experiencing/witnessing and 3.35, SD 1.4 for personal effects. Using ANOVA and Bonferroni post hoc tests, this was found to be statistically significantly lower (P < 0.01) than the frequencies reported by the LPN/LVN group, RN diploma group, and the AD group. (See Frequency of reported HV by educational level.)
Forty-one percent (395 partici-pants) reported being certified in their specialty area. No differences in reported frequencies of HV or personal effects based on certification were found.
- Considerations with statistical analyses. One must always be careful in making conclusions from statistical analyses. The mean frequency of experiencing/witnessing HV for the 31 to 40 age range was also higher at 4.56, but it didn't reach statistical significance because of the smaller sample size. Consideration must also be given to the fact that the data were nonparametric, meaning they don't follow the rules of mathematics and normal distribution. The Likert scale in which the data were collected on frequency of HV doesn't provide interval or ratio data. Distances between each point on the scale aren't equal. Even though the difference between a mean of 4.61 and 4.21 is statistically significant, in light of what those numbers represent, you must ask, does it have any practical significance?
- Qualitative data analysis. The editors received 14 letters, and 385 respondents provided comments. Analyzing these data, we found three major themes. The first is the stress and complexity of care in which nurses work. Many participants commented that working in these conditions causes nurses to feel powerless, demeaned, and victimized, and that this contributes to unproductive or negative coping behaviors.
The second theme was the relationship of management to an environment that fosters bullying or HV. Two basic mechanisms were described. One was that formal leaders (managers, charge nurses, directors of nursing, administrators, and physicians) use their power or position to bully and intimidate subordinates. The second was that these leaders often turn a blind eye to what's happening in terms of both HV perpetrated around them and HV they themselves may perpetrate. Participants stated that when they reported the behaviors, nothing was done about it. The bully was often stated to be “friends” with the leader. A trickle-down effect was described in which the leadership set the tone: “bullying is commonplace and tolerated by upper management” and is “allowed to occur.” Major contributing factors respondents identified included a lack of respect, support, and positive recognition from management.
The third major theme was fear of retaliation. Throughout the letters and comments, participants expressed passionate feelings about HV but also pleaded anonymity. Some even stated that if their hospital found out what they'd written, they'd lose their job. This thread of fear is disconcerting. We are a profession that advocates for and protects others, but we're afraid to advocate for ourselves.
There are over 3 million RNs and over 750,000 LPNs in the United States alone.6 Only a small portion of those nurses (955 participants) responded to this survey. Although most participants identified themselves as nurses, 18 identified themselves as certified nursing assistants, 4 as students, and 55 didn't provide information on their primary position. The sample was a self-selected convenience sample, meaning that those who responded to the survey may represent different opinions from the rest of the population of nurses. A large component of nurses may not think HV is a problem at all.
In general, the results of this survey tell us that many nurses are concerned and affected by HV or bullying in their workplace. Despite statistical significance found among some of the relationships of nurse characteristics and experience with HV, the strength of those relationships—and thus, interpretation of clinical significance—appears to be weak. We conclude HV isn't isolated to a particular age, gender, educational background, work setting, or tenure in nursing. No one is exempt, and the behaviors associated with HV are destructive to the individual and to the healthcare system. As a profession, nurses need to gather around this problem and make sure nurses and nursing leaders have the skills needed to effectively deal with issues of interpersonal conflict.
What would it take for us to show our appreciation of each other more often and what difference would it make? The results of this survey provide a mandate to the nursing profession for leadership that fosters a healthy environment. Research validating measurement tools for frequency and impact of HV and interventions for decreasing HV are needed.
Learning to resolve a conflict directly with the person involved instead of complaining to others requires maturity and communication skills, both of which can be learned and fostered in a workplace environment. In our organization, we're currently conducting an interventional research study. The purpose is to determine if developing a support group of nurse champions along with providing housewide education will create an environment that decreases the frequency of HV. Our first step was to create a measurement tool with reliability and validity. This tool will be described in a future article and will be made available for others to use.
Although respondents to this survey represented a wide spectrum of practice settings and roles (from technician in acute care, to nurses in correctional facilities, to academic faculty), most were staff nurses (53%) working in acute care (83%). Most were between ages 40 and 60, and 92% were female. The mean years in nursing were reported to be 21.8 (SD 12.7, minimum 1 and maximum 42 years).
All levels of nursing education were represented in survey responses, with the most frequently reported being baccalaureate of science in nursing (BSN) (36%) and associate degree in nursing (AD) (20%). Forty-one percent reported being certified in their specialty.
Statistical definitions made easy
ANOVA. Analysis of variance is a statistical test for differences among groups when you have more than two groups and the dependent variable is measured as continuous data. Post hoc tests such as Bonferroni are used with ANOVA to figure out which of the groups are really different.
Bonferroni test. See ANOVA above.
Chi-square. A statistical test used to determine if statistical differences exist between two or more categorical variables (such as male and female, or categories of education or age).
Cronbach's alpha. A statistical test used to measure internal reliability of a survey tool, meaning that each question in a subscale reliably represents the true opinions of the respondents on the concept measured by the subscale. The higher the result, the better; greater than 0.8 is very good.
P value. The probability that a difference in outcomes is related to the intervention being tested, not chance. If P is 0.05 or less, it's more likely that the difference in outcomes is related to the intervention being tested rather than to chance.
Parametric data. Data that follow the rules of mathematics and normal distribution (the bell curve), interval or ratio data as opposed to rank order or categorical data.
Pearson's r. A correlation statistic used to test relationships (parametric).
Post hoc test. See ANOVA above.
Spearman's rho. The nonparametric correlation test.
Standard deviation (SD). A statistical number that shows the variation from the average or mean. A low SD indicates that the data points tend to be very close to the mean; a high SD indicates that the data points are spread out over a large range of values.
T-Test. A statistic test used to determine differences between two groups when the dependent variable is continuous data.
About the survey tool
Last spring, nurses taking this survey responded to the following queries. Responses were measured with a 6-point Likert scale, as follows: 1 = never, 2 = once, 3 = a few times, 4 = monthly, 5 = weekly, 6 = daily. To see the complete survey tool, including answer choices, see Dumont C, Riggleman K, Meisinger S, Lein A. Nursing2011 horizontal violence survey. Nursing. 2011;41(4):9–10.
Within the last 12 months, how often have you personally experienced or witnessed the following:
- Harshly criticizing someone without having heard both sides of the story.
- Belittling or making hurtful remarks to or about coworkers in front of others.
- Complaining about a coworker to others instead of attempting to resolve a conflict directly by discussing it with that person.
- Raising eyebrows or rolling eyes at another coworker.
- Pretending not to notice a coworker struggling with his or her workload.
Answer these questions from the perspective of how you personally have been affected within the last 12 months at your current workplace.
- I've felt discouraged because of lack of positive feedback.
- I haven't spoken up about something I thought was wrong because of fear of retaliation.
- I've hesitated to ask questions for fear I'd be ridiculed.
- I've left work feeling bad about myself because of interactions with coworkers.
- I've had physical symptoms such as inability to sleep, headaches, and abdominal pain because of poor interactions with certain coworkers.
How often have you observed the following categories of healthcare workers exhibiting the negative behaviors described in this survey?
- unlicensed assistive personnel (nurses' aides)
- nurse peers (RNs and LPNs)
- supervisors (directors, managers, educators, charge nurses)
- other professionals (such as respiratory therapists, lab technicians, physical therapists, radiology technologists, pharmacists)
- other staff (such as housekeeping, security, secretarial, maintenance).
1. International Council of Nurses. Guidelines on coping with violence in the workplace. Geneva, Switzerland; 2004.
2. Pontus C, Scherrer D. Is it lateral violence, bullying or workplace harassment? Often, it is one and the same. Massachusetts Nurse Newsletter
. April 16–17, 2011.
3. Brunt B. Breaking the cycle of horizontal violence. ISNA Bulletin
. February-April 2011;37(2):6–11.
4. Cleary M, Hunt GE, Horsfall J. Identifying and addressing bullying in nursing. Issues Ment Health Nurs
© 2012 Lippincott Williams & Wilkins, Inc.
5. Vessey JA, Demarco R, DiFazio R. Bullying, hararssment, and horizontal violence in the nursing workforce: the state of the science. Ann Rev Nurs Res