中国城市居民死因统计资料准确性评价研究

Assessment on Cause of Death Statistics in China Urban Area

目的评估中国城市居民常规死因登记报告资料的准确性。方法以专家组核查医院死亡病例医学记录后作出的死因诊断为金标准,通过常规死因登记资料与金标准比较,判断常规死因登记资料的准确性。结果常规死因登记报告对高血压、糖尿病、肺炎的错报比较严重,对脑血管病、几种主要的癌症(肺、胃、肝、乳腺、结肠直肠癌)的报告比较准确,其PPV和敏感度都超过了80%;但细分死因后,敏感度降低,如脑血管病敏感度为82%,而脑出血仅为69%,脑梗死仅为66%。专家组诊断和常规死因登记报告的死因矩阵分布结果显示,疾病和伤害死因的错分结果是相互补偿的。结论中国城市常规死因登记资料死因错报比较普遍,但由于死因错分的补偿效应,还是基本能正确地反应真实的死因模式。

Objective Assessing the accuracy of routine vital registration data on cause of death in China urban areas. Methods Comparing the diagnosis from routine Vital Registration （VR） with the Gold Standard, which is based on experts＇ reviewing on cause of death for medical case records in hospitals, to assess the accuracy. Results Misdiagnosing Hypertension, Diabetes and Pneumonia in VR is serious problem, and the degree of accuracy on cerebrovascular diseases and some types of cancer （e. g. lung cancer, liver cancer, stomach cancer, colon cancer, breast cancer, rectum and anus） is high with over 80% on PPV and sensitivity. However, Sensitivity is deceasing after exploring cause of death in depth. For instance, the sensitivity for cardiovascular diseases, intracerebral haemorrhage and cerebral infarction, is 82%, 69% and 66% respectively. The result of Matrices analysis for diagnosing cause of death by the experts and routine vital registration shows that the result of misreporting about diseases and injuries is complementary for each other. Conclusion Misreporting cause of death in Routine Vital Registration in China urban areas is common. It can, however, basically reflect the real pattern of cause of death due to complementary effect of misdiagnosis.