NOTE: This page is archived for historical purposes and is no longer being maintained or updated.

C5.1 Evidence on the Effectiveness of Measures Recommended to Prevent Workplace Homicide- Loomis D, Wolf S, Runyan CW, Marshall S, Butts JD

Government agencies have recommended that employers adopt measures to reduce the risk of homicide on the job. Their recommendations include both environmental design modifications and administrative actions. To investigate the value of recommended preventive measures, we analyzed data from a case-control study of homicide in North Carolina workplaces in 1994-98. Workplaces were the units of analysis: case workplaces were those where a worker's death resulted from homicide during the study period (n=105); control workplaces were an incidence-density sample of the study base, matched by industry sector (n=210). Data on safety measures and other workplace characteristics were collected by telephone interview. Conditional logistic regression was used to estimate the exposure odds ratio (OR) as an indicator of association. Among 13 environmental interventions examined, only the presence of a barrier between workers and the public (OR 0.5, 95% CI 0.2-1.2) was associated with a noteworthy reduction in risk. Keeping entrances closed during working hours (OR 0.5, 95% CI 0.2-1.1), special arrangements with a law enforcement agency (OR 0.4, 95% CI 0.2-0.8), pre-employment psychological screening (OR 0.5, 95% CI 0.2-1.5) and having more than one worker on duty (OR 0.4, 95% CI 0.2-0.7) were the only administrative measures associated with notably lower risk among the 13 considered. Some interventions, including installing surveillance cameras (OR 1.8) and mirrors (OR 1.7) posting cash-limit warnings (OR 4.2), employing security guards (OR 1.2), and training employees to respond to robbery (OR 1.6) were associated with higher homicide risk. Neither adjustment for community and employer characteristics nor restriction to the subset of killings that resulted from robbery substantially changed these results. These findings suggest that many measures recommended for reducing the risk of workplace homicide may not be providing adequate protection as they are currently implemented in industry.


C5.2 Risk Factors for Violence Among Nurses: Methods and Preliminary Results-Gerberich SG, Church T, McGovern P, Hansen H, Nachreiner N, Geisser M, Watt G

The purpose of this two-phase effort is to document the magnitude of work-related violence (phase 1) and identify relevant risk factors (phase 2) among the population of Minnesota nurses - one of the few readily accessible occupational groups. A conceptual model served as the basis for collecting exposure information and guiding the analysis. Pilot-testing was conducted prior to each phase. This paper presents preliminary results and the methods used to obtain quality violence incidence and exposure data.

To pilot-test phase one, 220 Registered Nurses and Licensed Practical Nurses were randomly selected from nursing license records (response rate 82%); phase two pilot testing involved 23 physical assault cases and 69 controls. Intensive follow-up methods (mail and telephone) were used to maximize the response rate; up to four follow-up efforts were incorporated in both phases. Pilot-testing revealed difficulty in quantifying the frequency of non-physical violence events. Thus, modifications were made to address ongoing events adequately.

Based on limited data from the literature, a conservative 5% assault rate had been estimated; however, the assault rate for eligible nurses was greater than expected - 17% in the pilot study, and 14% in the comprehensive study. Non-physical violence (threats, verbal abuse, and/or sexual harassment) was reported by 40% of nurses in the pilot study, and 38% in the comprehensive study. With the higher than expected assault rate, and the unexpected additional staff time necessary to conduct follow-up, the sample size was decreased from 12,600 to 6,300 for the comprehensive study, which still yielded an adequate size for the case-control study.

Work-related violence is a serious problem. This study is an important step in identification of the extent of the problem and relevant risk factors that will facilitate the development of relevant prevention and control strategies.


C5.3 Workplace-Level Risk Factors for Homicide on the Job-Loomis D, Wolf SH, Runyan CW, Butts JD, Marshall S

Homicide is the second leading cause of death on the job for all US workers and the leading cause for working women. Risk factors for workplace homicide have been identified primarily through descriptive epidemiologic studies based on routine surveillance data, which do not allow detailed analysis. To investigate predictors of workplace homicide more thoroughly, we conducted a case-control study of homicides in North Carolina workplaces in the years 1994-98. Workplaces were the units of analysis: case workplaces (n=105) were those where a worker was killed during the study period, while control workplaces (n=210) were an incidence-density sample of the study base, matched by industry sector. Data on potential risk and protective factors were collected by telephone interview. Strength of association was assessed by the exposure odds ratio (OR), estimated via conditional logistic regression. Industries at especially high risk were taxi services (OR 25.1, 95% CI 4.05- ), and grocery and convenience stores (OR 10.6, 95% CI 1.32- and OR 9.9, 95% CI 2.9-33.4, respectively). Employer characteristics associated with markedly higher risk included having been in business at the current location <12 months (OR 6.2, 95% CI 1.2-32), having only 1 worker (OR 2.9, 95% CI 1.9-7.2), and working at night (OR 4.9, 95% CI 2.7-8.8), or on Saturdays (OR 4.2, 95% CI 1.9-9.2). The occurrence of homicide was also associated with a history of robbery (OR 3.3, 95% CI 1.6-6.8) or violence against workers (OR 15.4, 95% CI 4.6-51.1) within the previous 2 years. In addition, workplaces with only male employees or with employees of only one race were more likely to experience a killing. While the preceding risk factors are not directly modifiable through classical workplace interventions, it is important to identify them and understand their interrelationships preliminary to developing or evaluating protective measures.


C5.4 Interventions for the Primary Prevention of Work-Related Carpal Tunnel Syndrome-Lincoln AE, Vernick JS, Ogaitis-Jones S, Smith GS, Mitchell CS, Agnew J

Objective: To evaluate interventions for the primary prevention of work-related carpal tunnel syndrome (CTS).

Selection Criteria: Studies included engineering, administrative, personal, or multiple component interventions with comparison data and applied to a working or working-age population. Outcome measures included the incidence, symptoms, or risk factors for CTS, or a work-related musculoskeletal disorder of the upper extremity that included CTS in the definition.

Results: Twenty-four studies met our inclusion criteria. Engineering interventions included alternative keyboards, computer mouse designs and wrist supports, keyboard support systems, and tool re-design. Personal interventions included ergonomics training, splint wearing, electromyographic biofeedback, and on-the-job exercise programs. Multiple component interventions (e.g., ergonomic programs) included workstation redesign, establishment of an ergonomics task force, job rotation, ergonomics training, and restricted duty provisions.

Multiple component programs were associated with reduced incidence rates of CTS, but the results are inconclusive because they did not adequately control for potential confounders. Several engineering interventions positively influenced risk factors associated with CTS, but the evaluations did not measure disease incidence. None of the personal interventions alone were associated with significant changes in symptoms or risk factors. All of the studies had important methodological limitations that may affect the validity of the results.

Conclusions: While results from several studies suggest that multiple component ergonomics programs, alternative keyboard supports, and mouse and tool re-design may be beneficial, none of the studies conclusively demonstrates that the interventions would result in the primary prevention of CTS in a working population. Given the lack of demonstrated effectiveness, intervention research should prioritize randomized controlled trials that include: 1) adequate sample size; 2) adjustment for relevant confounding variables; 3) isolation of specific program elements; and 4) measurement of long-term primary outcomes such as the incidence of CTS, and secondary outcomes such as employment status and cost.




NOIRS 2000 Menu Page
NIOSH home page  CDC home page  NORA Traumatic Injury home page 
NIOSH - National Institute for Occupational Safety and Health CDC - Centers for Disease Control and Prevention NORA - National Occupational Research Agenda
Page last updated: March 2001
Page last reviewed: March 2001
Content Source: National Institute for Occupational Safety and Health (NIOSH) Division of Safety Research