Swerdlow A J
Medical Statistics Division, Office of Population Censuses and Surveys, London, UK.
Br J Cancer. 1989 May;59(5):787-91. doi: 10.1038/bjc.1989.164.
For some death certificates in England and Wales the cause information coded and published in national data is not that initially submitted by the certifier, but instead derives from a subsequent enquiry to the certifier for further information. These enquiries can lead to substantial artefacts in secular mortality data, and also to substantial non-comparability between mortality data for special study groups, such as subjects in cohort studies, and published mortality data. A description of current enquiry policy relevant to cancers, and changes in this policy over recent years is given to aid interpretation of mortality data. The effects on secular data of changes in enquiry policy are illustrated. At 4-digit level of the ICD, changes in enquiry policy can alter published mortality rates by several hundred per cent. At 3-digit level the greatest effects of enquiries at present are to increase the number of deaths coded to cancer of the eye by 35% and cancer of the body of the uterus by 31%; cancers of the thymus, heart and mediastinum are increased by 18%, and pleural cancer by 17%, while decreases of more than 10% are caused for several 'other' and 'unspecified' rubrics, and a decrease of 6% for deaths coded to melanoma.
在英格兰和威尔士,对于某些死亡证明,国家数据中编码和公布的死因信息并非最初由证明人提交的信息,而是源自随后向证明人询问以获取更多信息。这些询问可能会在长期死亡率数据中导致大量人为因素,还会导致特殊研究群体(如队列研究中的受试者)的死亡率数据与公布的死亡率数据之间存在很大的不可比性。本文给出了与癌症相关的当前询问政策描述以及近年来该政策的变化,以帮助解释死亡率数据。文中说明了询问政策变化对长期数据的影响。在国际疾病分类(ICD)的4位数层面,询问政策的变化可使公布的死亡率改变数百个百分点。在3位数层面,目前询问的最大影响是将编码为眼癌的死亡人数增加35%,子宫体癌的死亡人数增加31%;胸腺、心脏和纵隔的癌症死亡人数增加18%,胸膜癌死亡人数增加17%,而对于几个“其他”和“未指定”的类别,死亡人数减少超过10%,编码为黑色素瘤的死亡人数减少6%。