Hakulinen T, Teppo L
Int J Cancer. 1976 Apr 15;17(4):429-35. doi: 10.1002/ijc.2910170403.
The relative significance of various forms of cancer in terms of causing death is analysed by estimation of the increase in person-years of working age (20-64 years) following elimination of the disease. Methods based upon the theory of competing risks are applied to the statistics on causes of death in Finland during the years 1966-70. It is estimated that if there were no lung cancer (the commonest type of cancer in both morbidity and mortality statistics in males in Finland) the annual deaths saved would yield 5,900 working years (both sexes combined). Leukaemia and cancer of the stomach would be next in rank order, with figures of respectively 4,000 and 3,900 working years more. Female cancer with the highest incidence, that of the breast, would be characterized by 2,900 additional working years. The significance of types of cancer that affect young people is stressed in these calculations: leukaemia, brain tumours and lymphomas (both sexes combined) are 6th, 12th and 11th respectively in the statistics of cancer causes of death, but 2nd, 4th and 6th respectively in the list of additional working years to be gained by elimination of the disease. On the other hand, cancer of the prostate, 3rd in males according to the annual numbers of cancer deaths, would take the 15th position for the increase in working years in males. If no risk of cancer existed, the annual deaths saved would produce 36,000 working years, a figure exceeded only by those for cardiovascular diseases (55,000 working years) and accidents (51,000 working years). The results indicate that no practical differences exist between the results derived under the assumptions of various models for competing risks, but that the exclusion of competing risks may result in considerable degrees of bias in estimation if the population has a high general mortality.
通过估计消除某种疾病后工作年龄（20 - 64岁）人群的人年数增加情况，分析了各种癌症在导致死亡方面的相对重要性。基于竞争风险理论的方法应用于芬兰1966 - 1970年期间的死亡原因统计数据。据估计，如果没有肺癌（芬兰男性发病率和死亡率统计中最常见的癌症类型），每年挽救的死亡人数将产生5900个工作年（男女合计）。白血病和胃癌位列其次，分别多产生4000和3900个工作年。发病率最高的女性癌症——乳腺癌，将额外产生2900个工作年。这些计算强调了影响年轻人的癌症类型的重要性：白血病、脑肿瘤和淋巴瘤（男女合计）在癌症死亡原因统计中分别位列第6、第12和第11，但在消除这些疾病后可获得的额外工作年数列表中分别位列第2、第4和第6。另一方面，按每年癌症死亡人数计算在男性中位列第3的前列腺癌，在男性工作年增加方面将位列第15。如果不存在癌症风险，每年挽救的死亡人数将产生36000个工作年，这一数字仅低于心血管疾病（55000个工作年）和事故（51000个工作年）。结果表明，在各种竞争风险模型假设下得出的结果之间不存在实际差异，但如果人群总体死亡率较高，排除竞争风险可能会导致估计中出现相当程度的偏差。
