科研之友
微信
新浪微博
Facebook
分享链接
摘要
Cancer registries often receive inquiries on possible cancer clusters. In 2010, the Cancer Registry of Lower Saxony reported on a spatial leukemia accumulation in the community of Asse. As a result, the Cancer Registry was engaged in developing an approach for a regional cancer monitoring program by the federal government. The modular approach involves a two-stage process. First, all regional monitoring areas are under surveillance, and in the second phase, noticeable areas only are observed further. A positive confirmation test is regarded as an initial concern and may lead to further investigations.
The standardized incidence ratio (SIR) is used in the first stage. For confirmation, the Sequential Probability Ratio Test (SPRT) is applied in the second stage. All new observations are tested at a given time. The SPRT tests the null hypothesis (no increase) and an alternative hypothesis (increase) leading to three possible conclusions: warning, all-clear, or further surveillance necessary.
A high false alarm rate due to clustering by chance is to be considered with multiple statistical testing. The number of false alarms depends on the number of tested areas, diagnoses, subgroup analyses, and time periods. Preliminary considerations comprise 400 spatial areas, three diagnosis groups, no subgroup analyses, and a biannual first-stage testing. Guidelines for the level of significance and for detecting an increase in incidence will be necessary for the assessment of the monitoring parameters.
Before running the monitoring program, a number of questions have to handled politically; e.g., how many false alarms can be handled, the impact on the population involved, and existing risk communication structures.
- 出版日期2014-1
