Poole Charles
Department of Epidemiology, University of North Carolina, Chapel Hill, NC 27599-7435 USA.
Epidemiol Perspect Innov. 2010 Aug 24;7:6. doi: 10.1186/1742-5573-7-6.
A person can experience an effect on the occurrence of an outcome in a defined follow-up period without experiencing an effect on the risk of that outcome over the same period. Sufficient causes are sometimes used to deepen potential-outcome explanations of this phenomenon. In doing so, care should be taken to avoid tipping the balance between simplification and realism too far toward simplification. Death and other competing risks should not be assumed away. The time scale should be explicit, with specific times for the occurrence of specified component causes and for the completion of each sufficient cause. Component causes that affect risk should occur no later than the start of the risk period. Sufficient causes should be allowed to have component causes in common. When individuals experience all components of two or more sufficient causes, the outcome must be recurrent. In addition to effects on rates and risks, effects on incidence time itself should be considered.
一个人在规定的随访期内可能会经历对某个结局发生的影响，但在同一时期内却不会经历对该结局风险的影响。有时会使用充分病因来深化对这一现象的潜在结果解释。这样做时，应注意避免在简化与现实之间的平衡过度偏向简化。不应假定死亡和其他竞争风险不存在。时间尺度应明确，要给出特定组成病因发生的具体时间以及每个充分病因完成的时间。影响风险的组成病因不应晚于风险期开始时出现。应允许充分病因有共同的组成病因。当个体经历两个或更多充分病因的所有组成部分时，结局必然会复发。除了对发生率和风险的影响外，还应考虑对发病时间本身的影响。
