Piecewise Exponential Models to Assess the Influence of Job-Specific Experience on the Hazard of Acute Injury for Hourly Factory Workers

• Activity Number: 403
• Type: Contributed
• Date/Time: Tuesday, July 31, 2012 : 2:00 PM to 3:50 PM
• Sponsor: Biometrics Section
• Author(s): Jessica Kubo*+ and Mark Cullen and Linda Cantley and Martin Slade and Sally Vegso and Baylah Tessier-Sherman and Oyebode Taiwo and Manisha Desai
• Companies: Stanford University and Stanford University and Yale University and Yale University and Yale University and Yale University and Yale University and Stanford University
• Keywords: piecewise-exponential ; Weibull ; job experience ; injury risk ; baseline hazard ; occupational health

Abstract

An inverse relationship between experience and injury risk has been observed in many occupations. Methodological challenges in quantifying the reduction in risk include repeated events, varying lengths of follow-up and censored data. Cox proportional hazards models are not applicable because they ignore explicit modeling of the baseline hazard, the main parameter of interest. We investigated the novel use of parametric survival models for the baseline hazard to assess whether experience affected the hazard of injury for hourly workers of a U.S. factory at hire and after later job changes. Specifically, the comparison of competing models including the exponential, Weibull, and piecewise exponential models was proposed to formally test the null hypothesis that experience impacts the hazard of injury. The selected model was a two-piece exponential model that allowed the baseline hazard of injury to change with experience. Our findings suggested a 30% decrease in the hazard after 1 year of experience. Piece-wise exponential models may be particularly useful in modeling injury risk and have the additional benefit of interpretability over other similarly flexible models.