Abstract
Since 1993, health authorities have had a responsibility for the co-ordination of the health aspects of response to chemical incidents. The public health team within health authorities currently takes the lead in complying with this duty. However, the majority of public health professionals have had minimal training and experience in this area. The focus of this investigation has therefore been the provision of management tools to improve the public health response to chemical incidents, with the aim of minimising adverse health and environmental impacts. A detailed investigation of the current state of knowledge was conducted, and the research engineer was involved in the management of 70 chemical incidents resulting in water pollution, as reported to the Chemical Incident Response Service. A best practice model for the public health response to chemical incidents was developed, based on lessons learnt from the management of chemical incidents, involvement in incident exercises and training days, and relevant literature on chemical incident management. The model was then used to guide the development of a chemical incident management guidance manual providing detailed information and guidance on the issues raised in the model. A quantitative evaluation method was also devised for assessing the performance of public health professionals in managing chemical incidents. Validation testing of the guidance manual was conducted to assess its effectiveness in improving the public health response to chemical incidents. A group of 79 public health professionals took part in the testing, which concluded that the manual was effective in improving performance in the response to a chemical incident exercise. Statistical analysis further demonstrated that this improvement would be expected across the general public health population. The value of previous training on chemical incident management was also confirmed. However, previous experience in managing chemical incidents was not shown to influence performance levels.