NIOSHTIC-2 Publications Search

Reducing violence against caregivers in nursing homes.

Gates-D; Fitzwater-E; Succop-P; Sommers-MS
NIOSH 2003 Nov; :1-17
The worker most commonly assaulted in the workplace is the nursing assistant (NA) in long-term care and the perpetrator is most often the patient (resident). There are physical, emotional and economical costs related to violence in healthcare settings. Test the effectiveness of a violence prevention intervention, based on Social Cognitive Theory to increase knowledge and self-efficacy, and reduce assaults against NAs from residents. The following seven hypotheses were generated to compare the intervention group to comparison group one week and six months after the intervention: H1: Intervention NAs will have a significant increase in perceived knowledge of their violence prevention skills. H2: Intervention NAs will have a significant increase in self-efficacy of their violence prevention skills. H3: Intervention NAs will demonstrate an increase in violence prevention skills. H4: Intervention NAs will have a decreased number of assaults. H5: Age will be negatively related to the incidence of assaults. H6: Anger and stress will be positively related to the incidence of assaults. H7: Number of residents assigned will be positively related to incidence of assaults. Investigators conducted a quasi-experimental study; six nursing homes were randomly selected from all nursing homes with more than 100 beds in Hamilton County, Ohio. The six homes were randomly assigned to an intervention or comparison group. One hundred and thirty-eight NAs participated in three intervention and three 'comparison homes. Data was collected at baseline, 1 week after the intervention, and 6 months after the intervention. A baseline questionnaire was used to obtain information on demographics, employment, and past violence experience. At baseline, post-intervention and six months post-intervention the subjects completed the Knowledge and Self-efficacy survey, the State-Trait Anger Inventory (STAXI), Occupational Stress Inventory Revised (OSI-R), completed the Assault Log for 80 work hours and participated in a simulation exercise with a standardized actress. After baseline data collection was completed those in the intervention homes participated in nine, one hour weekly group sessions with a masters-prepared psychiatric nurse. The intervention was based on Social Cognitive Theory and included opportunities for role plays, modeling, problem-solving and discussion. T-tests and Poisson regressions were used to determine if significant changes in knowledge, self-efficacy and assaults occurred in the intervention group. ANOVAs were used to identify relationships between assaults and anger, stress, strain; and workload. On the baseline survey 59% of the NAs reported that they are usually assaulted by residents once a week and16% reported that they are assaulted by residents every day. Additional questionnaire findings regarding lifetime prevalence rates while working as an NA in a nursing home included: 1) 51 % reported that they have been injured by a resident, 2) 38% reported that they have received medical attention for an injury from a resident, 3) 10% reported that they have been assaulted by a co-worker, and 4) 4.3% reported that they have been assaulted by a visitor. The mean number of assaults for the homes ranged from 1.57 to 8.42. A total of 624 assaults were experienced by 94 NAs; 44 NAs (29%) did not encounter any assaults. The mean number of assaults for NAs was 4.52 and 6.64 for NAs who were assaulted at least once. The number of assaults per caregiver during the 80 hours ranged from zero to 64. Thirty one injuries resulted from assaults (5%) during the 80 hours of work. There was not a significant decrease in the assaults for the three intervention homes. However, data analysis excluding the third intervention home showed a significant decrease in the incidence of assaults (p<.05). The post-intervention results from the ANOVAs showed a significant increase in the intervention subjects' perceived knowledge (p<0.001), self-efficacy (p < 0.01), and violence prevention skills (p<0.05). For the measurement six months after intervention results the ANOV As again showed a significant increase in the intervention subjects' perceived knowledge (p<.05) However, there was no significant difference in the subjects' self-efficacy 6 months after the intervention nor increase in the use of violence prevention skills. At the second and third measurements, the intervention had no significant main effect on the number of assaults. However, an interaction effect between the intervention and number of pre-intervention assaults was found. The intervention had a significant effect on those NAs who had less than 8 assaults pre-intervention (p < 0.001) and no significant effect (p > 0.05) on those who had more than 7 assaults on pre-intervention. Results from the regression analysis also found that the following covariates predicted the incidence of assaults at the first, second and third measurements: age (p < .0001), number of assigned residents (p < .05), and state anger ( < .001). The rate of assaults from residents was not surprising and supported findings from previous studies. However, the lifetime prevalence rates of injury and medical care related to assaults from residents were unexpectedly high. The assault rates from co-workers and residents' family members were astounding when compared to other work settings. Although the intervention was significant in its ability to decrease assaults in the first 2 intervention homes, the inability to do so at study completion was probably due to two factors. First, the intervention homes had a much lower assault incidence at baseline making it more difficult to show a significant decrease. Secondly, during the intervention for the third nursing home, two major events occurred that affected the intervention: an attempt to unionize NAs and an unexpected state licensing inspection. More research is needed to study the incidence of all types of violence experience by NAs, as well as interventions to decrease the incidence of violence in this work setting. The interaction effect at both the second and third measurements with the number of pre-intervention assaults identifies implications for practice and policy in terms of workload. In addition, the covariates related to the incidence of assaults identify that the risk factors for assaults in nursing homes is multi-faceted and requires interventions that not only address skills but address workload, work environment, and worker characteristics.
Nursing; Nurses; Health-care; Health-care-facilities; Health-care-personnel; Injuries; Injury-prevention; Emotional-stress; Physical-stress; Safety-measures; Questionnaires; Demographic-characteristics; Medical-care; Medical-facilities; Medical-personnel
College of Nursing, Procter Hall, 3110 Vine Street, University of Cincinnati, Cincinnati, Ohio 45211-0038
Publication Date
Document Type
Final Grant Report
Funding Amount
Funding Type
Fiscal Year
Identifying No.
NIOSH Division
Source Name
National Institute for Occupational Safety and Health
Performing Organization
University of Cincinnati, Cincinnati, Ohio
Page last reviewed: March 11, 2019
Content source: National Institute for Occupational Safety and Health Education and Information Division