**1. Introduction**

In certain occupational contexts, such as knowledge-based work, some level of employee turnover is considered potentially healthy for an organization (e.g., by providing opportunities to replace poor performers) [1,2]. More broadly, however, it is widely believed that a high level of turnover is a marker for one or more undesirable characteristics of the employment circumstance, including safety and health hazards [3–5]. Increased levels of employee turnover are also commonly associated with decreased organizational performance and profitability [3,5,6]. Thus, understanding and mitigating factors leading to employee turnover is an important business management strategy.

An increased risk of occupational injury during the earliest periods of employment has been observed in numerous studies spanning many decades [7,8]. Examinations of early data compiled by the US Bureau of Labor Statistics observed that both the proportion and incidence rates of occupational injuries were greatest during the first three months of employment [9,10]. Many subsequent studies also observed greater injury frequencies, prevalence, or incidence rates among employees with the lowest job tenure. For example, from 1995 to 2004, 28% of >86,000 injuries among mining workers

occurred among those in the first year of employment, while 33% occurred among those employed 1–5 years [11]. Among logging workers, the rate of workers' compensation claims (from 1999 to 2003) was more than double among those with two or fewer months of job tenure (vs. those with >2 months of job tenure) [12]. In addition, substantially greater incidence rates and relative risks of first-time workers' compensation claims involving lost work days were observed among those employed (across a range of industrial sectors) for less than one month (vs. those employed for at least 13 months) [13]. This effect was consistent across injury event classifications, including those typical of acute, traumatic injuries (e.g., contact with objects/equipment and falls) and those typical of more chronic, musculoskeletal outcomes (e.g., repetitive motion). Recent meta-analyses also suggest that negative impacts of turnover on safety outcomes (e.g., occupational injury rates) are among the strongest drivers of reduced organizational financial performance [14]. Thus, if turnover levels are high, then an employer will find itself continuously (i) replenishing its workforce with new employees at greatest risk of injury and associated costs (e.g., workers' compensation) and (ii) incurring potentially avoidable costs associated with hiring and training new employees.

While employment duration or worker experience has frequently been examined as a risk factor for occupational injury, occupational injury has less often been examined as a risk factor for employment duration (or termination). The relationship between occupational injury and employee turnover, if any, is likely complex. Cree and Kelloway [15] suggested a model whereby employees' turnover intentions are influenced by the perception of risk associated with the work, which itself is a function of (i) employees' "accident history," either direct (i.e., actual injury experience) or indirect (i.e., knowledge of others' injury experiences) and (ii) the attitudes of coworkers, supervisors, and management regarding workplace safety and health. In this context, it is important to note that turnover intentions relate to voluntary employment termination, which is also influenced by factors not directly related to workplace safety and health, such the availability of alternative employment options [16]. Associations between occupational injury and subsequent employment termination (both voluntary and involuntary) have been recently observed among health care workers [17,18]. It is not known, however, if the associations observed among health care workers also apply to manufacturing settings, in which entry into the workforce may not require specific knowledge and skills, and in which workers may not be afforded the same level of autonomy [19].

The objective of this study was to estimate the association between occupational injury and subsequent employment termination among newly hired manufacturing workers. Specifically, we merged human resources information with information from an on-site occupational health center to create a time history for each worker in relation to both occupational injury events and employment duration (and/or termination). We also restricted our analyses to the first 60 days from the date of hire, and hypothesized that those experiencing an occupational injury during this time were more likely to terminate employment (for both voluntary and involuntary reasons) prior to 60 days from the date of hire.
