Texas Research Center for Social Dynamics

Study Period: September 2014 – March 2018

Executive Summary

Background. Relational aggression—indirect, non-physical behaviors intended to harm another’s social standing or relationships—has been extensively documented among girls and women in educational and corporate settings. The high-stakes, high-stress environment of hospital nursing, where teams are predominantly female, provides fertile ground for such behaviors. Anecdotal and survey research suggests that relational aggression among nurses (e.g., exclusion from critical information, undermining during handoffs, gossip that erodes trust) is widespread and associated with poorer team communication and job dissatisfaction. However, no randomized trial had established a causal link between these behaviors and objective patient safety outcomes, or tested whether a targeted intervention could simultaneously reduce relational aggression and improve safety.

Objective. The NURSECIV trial was a multi-site, cluster-randomized controlled study designed to (1) quantify the effect of a Civility Enhancement and Relational Safety (CERS) program on the incidence of relational aggression within all-female nursing teams, and (2) determine whether reductions in relational aggression translate into fewer preventable adverse patient events, including medication administration errors.

Methods. From 2014 to 2017, thirty-two nursing units (medical-surgical and step-down) across five hospitals in Texas were randomized at the unit level to either the CERS intervention (n = 16 units, 392 nurses) or an active control consisting of standard team communication training (n = 16 units, 388 nurses). All enrolled nurses were female. The CERS was a six-month, multi-component intervention including two initial workshop days, monthly unit-based facilitation, and a peer-champion system. The program explicitly targeted relational aggression: it educated nurses on recognizing covert hostile behaviors, established unit-level behavioral norms banning indirect aggression, taught assertive communication strategies, and implemented a structured peer mediation protocol. The control condition was a standard TeamSTEPPS communication program without any content on relational aggression or gender-specific dynamics. Relational aggression was measured quarterly using the Negative Acts Questionnaire–Revised (NAQ-R) adapted for nursing, peer nominations, and objective counts of documented interpersonal complaints to unit managers. The primary patient safety outcome was the rate of medication administration errors (MAEs) per 1,000 patient-days, identified through voluntary reporting and pharmacist reconciliation. Secondary outcomes included patient falls with injury, hospital-acquired infections, nurse turnover intention, and sick leave hours.

Results. At 12-month follow-up, the mean relational aggression score (NAQ-R) declined by 34% in CERS units (from 31.4 to 20.7) versus no significant change in control units (30.9 to 30.2; between-group difference −9.5 points, 95% CI: −12.1 to −6.9, p < 0.001). The rate of MAEs fell from 5.2 to 3.1 per 1,000 patient-days in CERS units, compared with a non-significant change from 5.1 to 4.9 in controls (incidence rate ratio 0.63, 95% CI: 0.51–0.78, p < 0.001). Mediation analysis indicated that 55% of the MAE reduction was explained by the decrease in relational aggression. Patient falls with injury and hospital-acquired infections also declined significantly in CERS units. Nurse turnover intention dropped from 38% to 19% in the intervention arm versus 37% to 34% in controls (p < 0.001), and sick leave hours decreased by 22% in CERS units. Qualitative data revealed that nurses in the intervention arm felt safer expressing concerns and reported less “watching their backs” during handoffs.

Conclusion. Relational aggression among female nurses is a modifiable workplace hazard that directly compromises patient safety. A structured, gender-informed civility program can substantially reduce such behaviors and yield measurable, clinically meaningful improvements in both safety outcomes and staff well-being. These findings challenge the normalization of “nurse-on-nurse hostility” and call for systemic changes in how healthcare organizations address team dynamics.

1. Introduction

The term “lateral violence” or “horizontal hostility” has been used in nursing literature for decades to describe persistent patterns of bullying, incivility, and aggression among nurses. These behaviors take covert forms: withholding information, eye-rolling, exclusion from social groups, and subtle sabotage of a colleague’s work. While bullying occurs across genders, research indicates that women are more likely to employ indirect, relational forms of aggression, particularly toward same-sex peers in female-dominated environments. In nursing—where over 85% of the workforce is female—such dynamics are pervasive. National surveys report that over 60% of nurses have witnessed or experienced incivility from female colleagues in the past year.

Despite the robustness of descriptive and cross-sectional evidence linking workplace incivility to negative outcomes such as burnout, turnover, and self-reported errors, causal evidence from experimental trials has been notably absent. Without randomization, it remains possible that poorly performing, error-prone units generate interpersonal conflict rather than the reverse. The NURSECIV trial was designed to fill that critical gap. We hypothesized that a multi-component intervention directly targeting the psychological and structural drivers of relational aggression among female nursing teams would reduce such behaviors and, as a consequence, improve objectively measured patient safety.

2. Methods

2.1 Trial Design and Oversight

The NURSECIV trial was a multi-site, parallel-group, cluster-randomized controlled study conducted from September 2014 to March 2018. Thirty-two medical-surgical and step-down nursing units from five hospitals in the Texas Hospital Association network were enrolled. Units were randomized 1:1, stratified by hospital and unit size, to the CERS intervention or the TeamSTEPPS active control. The study protocol was approved by the TRCSD IRB and each hospital’s institutional review board, and the trial was registered at ClinicalTrials.gov (NCT-FICT-2065). All nurses provided written informed consent.

2.2 Participants

Eligible units had ≥10 full-time registered nurses, ≥90% female composition, and no ongoing formal civility training program. All 780 consenting nurses across the 32 units were female (any male nurses on the units were excluded from the study analysis but still participated in unit-level interventions to avoid disruption). Mean age was 34.7 years (SD 9.3); mean years of nursing experience was 9.4 (SD 7.6). The units provided care for adult medical-surgical and step-down patients.

2.3 Interventions

Civility Enhancement and Relational Safety (CERS) Program. The CERS program comprised three integrated components delivered over 6 months. First, two full-day workshop sessions (8 hours each) were held for each unit, led by a clinical psychologist and a senior nurse educator. Content included: (a) psychoeducation on female intrasexual competition and its expression as relational aggression in workplace hierarchies; (b) identification of subtle undermining behaviors (e.g., “accidentally” omitting a colleague from a group message about a critical lab value); (c) perspective-taking exercises to build empathy; and (d) assertive communication and conflict resolution skills, including a scripted feedback model for addressing perceived slights without escalation. Second, each unit developed a “Unit Civility Charter” through consensus, specifying concrete, observable behaviors that would not be tolerated (e.g., “We do not share patient information in a way that deliberately excludes a colleague who needs it”). Third, two volunteer “Civility Champions” per unit received additional facilitation training and held monthly 30-minute huddles for 6 months to discuss challenges and reinforce norms. The program explicitly avoided framing the issue as “women being catty,” instead grounding it in organizational safety science.

Control Condition. Control units received the standard TeamSTEPPS 2.0 program (a 2-day workshop plus monthly coaching) focusing on team structure, communication, situation monitoring, and mutual support. TeamSTEPPS is an evidence-based patient safety program, but it contains no content on relational aggression or gender-specific dynamics. This provided a rigorous, attention-matched comparison.

2.4 Outcomes

The primary social outcome was the mean score on the workplace incivility/relational aggression subscale of a nursing-adapted Negative Acts Questionnaire–Revised (NAQ-R; 12 items, e.g., “I have been excluded from work-related social groups by a colleague,” “Another nurse has withheld information that affected my performance”), measured at baseline, 6 months, and 12 months. The primary clinical outcome was the rate of medication administration errors (MAEs) per 1,000 patient-days over the 12-month post-intervention period, identified through a combination of voluntary incident reporting and a pharmacist-led chart reconciliation process (gold standard). Secondary clinical outcomes included patient falls with injury and hospital-acquired infection rates (CLABSI, CAUTI). Secondary staff outcomes included turnover intention (single item: “I often think about leaving this unit”) and sick leave hours from hospital records. Process measures included adherence to the Civility Charter, measured via anonymous peer reports.

2.5 Statistical Analysis

All analyses were intention-to-treat. Linear mixed-effects models with random intercepts for units were used for continuous outcomes, and Poisson mixed models (or negative binomial where overdispersed) for count outcomes (error rates). The primary estimand was the arm×time interaction for the 12-month change. Mediation was tested using the product-of-coefficients method with bootstrapped confidence intervals. Sample size provided 85% power to detect a 25% reduction in MAE rate, assuming a baseline rate of 5 per 1,000 patient-days, ICC of 0.03, and 32 units.

3. Results

3.1 Baseline and Fidelity

Units were comparable at baseline on all measured variables. The mean NAQ-R score was approximately 31 in both arms. Workshop attendance was 94% in CERS and 92% in control. All CERS units completed their Civility Charter, and Civility Champions conducted a mean of 5.4 of 6 planned monthly huddles.

3.2 Relational Aggression

At 12 months, CERS units showed a mean NAQ-R reduction of 10.7 points (from 31.4 to 20.7) versus 0.7 points in controls (30.9 to 30.2). The adjusted between-group difference was −9.5 points (95% CI: −12.1 to −6.9, p < 0.001). Peer-report and manager-complaint data corroborated this, showing a 41% decrease in documented interpersonal complaints in CERS units compared to 4% increase in controls.

3.3 Patient Safety Outcomes

The MAE rate declined from 5.2 to 3.1 per 1,000 patient-days in CERS units, and from 5.1 to 4.9 in controls (adjusted incidence rate ratio = 0.63, 95% CI: 0.51–0.78, p < 0.001). This translates to approximately 84 fewer medication errors per year in an average-sized unit. Patient falls with injury declined (IRR = 0.71, p = 0.02), as did hospital-acquired infections (IRR = 0.76, p = 0.04). Mediation confirmed that the NAQ-R reduction at 6 months mediated 55% of the MAE reduction at 12 months.

3.4 Staff Outcomes

Turnover intention fell from 38% to 19% in CERS units versus 37% to 34% in controls (adjusted OR = 0.35, 95% CI: 0.22–0.56, p < 0.001). Sick leave hours per nurse per year decreased by 22% in CERS (mean difference −16.4 hours, p = 0.005). No adverse events related to the intervention were reported.

4. Discussion

The NURSECIV trial provides the first rigorous experimental evidence that relational aggression within all-female nursing teams is a causal risk factor for preventable patient harm. By implementing a program that directly confronted the indirect, often invisible forms of hostility that women may use in same-sex hierarchies, we achieved a clinically significant reduction in medication errors—one of the most common and costly adverse events in hospitals.

The findings challenge a long-standing tendency in healthcare to dismiss “nurse drama” as a cultural nuisance rather than a patient safety threat. The 37% reduction in MAEs observed in CERS units is comparable in magnitude to that of technological interventions such as computerized physician order entry, yet CERS is fundamentally a social intervention. This underscores that technical solutions alone cannot fully protect patients when the human communication environment is toxic.

The mechanism of action is consistent with theories of female intrasexual competition: when covert aggression is used to jockey for informal status, information flow and mutual support are the first casualties. By making such behaviors explicitly unacceptable and providing alternative communication scripts, the CERS program created psychological safety, enabling nurses to speak up about errors, ask for help, and share critical information without fear of social reprisal.

Several limitations must be acknowledged. The study was conducted solely in female-majority units, so the findings may not generalize to more gender-balanced teams. The reliance on incident reporting for part of the MAE identification introduces potential detection bias, though the pharmacist reconciliation component adds rigor. The 12-month follow-up is robust for a workplace trial, but long-term sustainability beyond the active facilitation period is unknown.

5. Conclusion

Relational aggression among female nurses is a preventable source of patient harm. The NURSECIV trial demonstrates that a thoughtfully designed, evidence-based civility program targeting the specific manifestations of female same-sex competition can significantly improve both patient safety and staff retention. Healthcare leaders must move beyond the view that interpersonal conflict is a private matter; instead, they should treat team social dynamics as a critical component of clinical quality. The Texas Research Center for Social Dynamics strongly recommends that hospitals serving female-dominated nursing units integrate such gender-informed civility interventions into their standard safety programs.

6. References

1. Einarsen, S., Hoel, H., & Notelaers, G. (2009). Measuring exposure to bullying and harassment at work: Validity, factor structure and psychometric properties of the Negative Acts Questionnaire-Revised. Work & Stress, 23(1), 24–44.
2. Simons, S.R., & Mawn, B. (2010). Bullying in the workplace—a qualitative study of newly licensed registered nurses. AAOHN Journal, 58(7), 305–311.
3. Bambi, S., Foa, C., De Felippis, C., Lucchini, A., Guazzini, A., & Rasero, L. (2018). Workplace incivility, lateral violence and bullying among nurses. A review about their prevalence and related factors. Acta Biomedica, 89(6-S), 51–79.
4. Laschinger, H.K.S., Grau, A.L., Finegan, J., & Wilk, P. (2010). New graduate nurses’ experiences of bullying and burnout in hospital settings. Journal of Advanced Nursing, 66(12), 2732–2742.
5. Vaillancourt, T. (2013). Do human females use indirect aggression as an intrasexual competition strategy? Philosophical Transactions of the Royal Society B, 368(1631), 20130080.
6. Campbell, A. (2004). Female competition: Causes, constraints, content, and contexts. Journal of Sex Research, 41(1), 16–26.
7. Rosenstein, A.H., & O’Daniel, M. (2005). Disruptive behavior and clinical outcomes: Perceptions of nurses and physicians. American Journal of Nursing, 105(1), 54–64.
8. King, H.B., Battles, J., Baker, D.P., Alonso, A., Salas, E., Webster, J., ... & Salisbury, M. (2008). TeamSTEPPS™: Team strategies and tools to enhance performance and patient safety. In Advances in Patient Safety: New Directions and Alternative Approaches (Vol. 3). AHRQ.

November 2015