INTRODUCTION: The statistics related to labor accidents as with any other notification system ought to be the basis for programs and policies with a view to the adoption of preventive measures. In order to establish preventive norms, however, the health system needs data from researchers focussing on the dynamics of and the pitfalls revealed by specific events. Within this context the main objective of this study is to proceed with an in-depth analysis of the labor accidents verified in Barcelona (Spain) using for this purpose a descriptive statistics model to test variables such as type of accident, economic sector, economic enterprise and type of labor contract. METHOD: The data source utilized was the notification system for labor accidents with grave consequences such as death of the victim registered in Barcelona during the period 1992-1993. Labor accidents registered for male workers numbered 848. A log-linear model was applied to this data base. RESULTS AND CONCLUSIONS: The results show a positive association between traumatic accidents with the construction, traffic and services sectors. A positive association was also found between traumatic accidents and the size of the company concerved the small ones being the worse type in terms of worker's injuries. Regarding the non-traumatic accidents, the study showed a positive correlation between large-sized enterprizes and type of temporary worker and the civil construction sector as compared to workers with long term work contracts within industry and services. There was some evidence, also, of a positive association between small and medium sized companies and temporary work and the ocurrence of work accidents.
Accidents; Log-linear models
