摘要
[目的]结合实例探讨APC泊松对数线性模型的结构及应用条件；阐明年龄、观察时期和出生队列对乌鲁木齐地区恶性肿瘤死亡率的影响，为恶性肿瘤的预防提供科学的依据。[方法]以乌鲁木齐地区恶性肿瘤死亡病例为研究对象，系统收集该地区1978-1997年恶性肿瘤死亡登记资料，采用SAS软件分析资料。[结果]30岁以上各年龄组患恶性肿瘤的危险性快速增大。30岁以上各组分别与15-19岁年龄组相比差别明显（P〈0．05），与1978-1982年相比，各观察时期的作用不明显（P〉0．05），在所有出生队列中，1914—1942年出生的人群患恶性肿瘤的危险性最大。[结论]乌鲁木齐地区恶性肿瘤死亡率受年龄和出生队列的影响；泊松对数线性模型在应用中为了获得可靠的专业结论需采用合理的数据压缩方法。
[Objective ] To explore the structure of and applicative limit to APC Poisson＇s loglinear model by an example and make clear the influence of age, period and cohortupon the death rate of Urumqi of Urumchi, thus provide proofs for the control of malignant tumors. [Methods] The patients of Urumqi city that died from malignant tumors were recruited as subjects. The data were collected from routine records of death and analyzed by the SAS system. [ Results] Persons over 30 years old had high risk for affecting malignant tumors than those at the age of 15-19 （P 〈 0.05）. Compared with the period of 1978-1982, the other periods had a slight influence on the death rate of carcinoma （P 〉 0.05）. The 1914-1942 birth cohort ranked in the highest risk of carcinoma. [ Conclusion] The death rate of malignant tumor of the area of Urumqi is associated with age and birth cohort. A reasonable method tot data-constraint is prerequisite in order to draw the pertinent conclusions that are interpretable by bio-chemical specialty.
出处
《现代预防医学》
CAS
北大核心
2008年第20期3907-3910,共4页
Modern Preventive Medicine
关键词
泊松对数线性模型
出生队列
乌鲁木齐
Poisson＇s logarithmically linear model
Birth cohort
Urumqi
