A study of behavioural responses to an industrial disaster.

Based upon 7 response variables we categorized the impact behaviour of 123 industrial employees who had been exposed to a disastrous factory explosion and had survived without suffering severe injuries. The subjects were personally examined and the data controlled by cross-interviews shortly after the disaster. In the analysis the subjects were divided into a high stress (n = 64) and a medium stress exposure group (n = 59) based upon their closeness to the explosion centre. The 7 variables were cognitive function, inadequate behaviour, help received, leadership, cooperative activity, absolute and relative rescue efforts. While about 50% of the total number reported some disturbance in their cognitive control, and 34% of the high stress exposure group experienced a near total loss of cognitive control, severe inadequate behaviour did not occur. Modelling and corrective social interactions may have played an important role in this. Still, 20% of the high stress exposure group had some behavioural response that increased the risk to their life or that of others. The 7 scores were added to an index and cut-off points established to separate 3 different groups of behavioural responses: 29% of the high stress group displayed Maladaptive Disaster Behaviour; inhibited behaviour and uncontrolled flight behaviour dominated in this group. Adaptive Disaster Behaviour was rated in 34% and Optimal Disaster Behaviour in 37%. The following background variables correlated strongly to Optimal Disaster Behaviour: high level of disaster training/experience, male sex, age above 40, maritime occupational background, above average intellectual ability, a life history without mental health problems. A discriminant analysis with 8 variables predicted correctly whether the response would be Optimal or less than that in 84% of the 121 subjects tested in the analysis. A high level of disaster training/experience yielded an overall correct prediction rate of 63.6%, and a sensitivity of 81%, specificity of 85.9% and positive predictive power of 70.7% in predicting Optimal Disaster Behaviour.