Modeling of toxicological effects of fire effluents: prediction of toxicity and evaluation of animal model.

Methodology for the prediction of the toxic effects of fire effluents has made considerable progress [1,2]. As emphasized by Hartzell [2] a limiting factor has often been the availability and quality of analytical input data. The Finney model [3] was used to predict the lethal potency and would appear to have utility as a tool in reducing the number of experimental animals used in material testing. However, pilot bioassay data are indispensable to validate the prediction and to categorize fire effluents into narcosis inducing, irritant or of causing unusual or unexpected toxicity. The comparison of predicted and actually observed carboxy-hemoglobin levels is considered to be a sensitive but indirect tool to assess whether major effects on respiration occurred. All laboratory combustion toxicity methods suffer from several types of limitations. However, they might be expected to be relatable to at least some stages of actual fires. Due to a lack of a clearly defined 'generation process'--if compared with conventional inhalation toxicity studies--the classification into broad categories of relative toxic potency seemed to be more appropriate than an absolute classification scheme.