Morabia A, Rey P, Bousquet A, Giezendanner D, Junod A, Liechti B
Soz Praventivmed. 1986;31(4-5):257-9. doi: 10.1007/BF02083480.
In this paper, two oral presentations are combined. The first described the broad aspects of decision analysis and the second mentioned those medical data which need to be gathered in order to apply the model to industrial fluorosis. For this purpose, biological and medical observations were collected, both from workers of the aluminium industry and from controls. The five successive steps for building up a decision tree are then demonstrated, the aim of which being to evaluate the fitness of a screening strategy. A computer programme has been developed which may be applied both to occupational or non occupational diseases. Referring to industrial fluorosis, the computerized decision tree showed that screening with preshift urinary fluor, clinical and radiological signs, plus bone fluor rate is required, as soon as the risk corresponds to a 7% prevalence of the disease.
在本文中，两个口头报告合并在一起。第一个报告描述了决策分析的广泛方面，第二个报告提到了为将该模型应用于工业性氟病而需要收集的医学数据。为此，从铝行业工人和对照组中收集了生物学和医学观察数据。然后展示了构建决策树的五个连续步骤，其目的是评估筛查策略的适用性。已经开发了一种计算机程序，该程序可应用于职业性或非职业性疾病。关于工业性氟病，计算机化决策树表明，一旦风险对应于该疾病7%的患病率，就需要进行班前尿氟、临床和放射学体征以及骨氟率的筛查。
