Reporting of skin cancer risks in the house-building industry: alternative approaches to the analysis of categorical data.

BACKGROUND

Increasing incidence of skin cancer is of concern to public health. Working predominantly outside, construction workers are at increased risk of sunburn and certain forms of skin cancer. The objective of this paper was to explore these concerns via use of alternative approaches to categorical data analysis by considering the relationship between the size of house-building companies, use of risk assessments and the system for reporting sunburn.

METHODS

The survey population was speculative house-building companies. A self-administered questionnaire was sent to health and safety advisors in house-building companies. Questions related to the number of house units built, inclusion of adverse weather in the risk assessment, and the presence or absence of a system for reporting sunburn were analysed. Alternative approaches to conventional chi(2) tests were used in order to investigate a multiway data structure. A loglinear model was employed using Statistical Package for Social Sciences to explain data in three-way tables.

RESULTS

The questionnaire achieved a response rate of 31% (n=94). A significant association identified in a two-way study of the mechanisms of reporting sunburn also appeared at the sublevels introduced by the inclusion of a third method of categorization. Finally, correspondence analysis was used to present a graphical insight into the study.

CONCLUSIONS

In the two-way analysis, the size of house-building company had a significant impact on the presence of a system for reporting sunburn, although this was not translated into the three-way analysis. The establishment of systems within risk assessments to incorporate exposure of construction workers to ultraviolet radiation, and reporting of incidents of sunburn, is essential if this public health issue is to be addressed.