以医院为基础的全国多中心女性原发性乳腺癌临床流行病学调查

A Nationwide Hospital-based Multicenter Clinical Epidemiological Investigation for Female Primary Breast Cancer in China

[目的]利用现有的医院病案资料,开展全国多中心的女性原发性乳腺癌临床流行病学调查,提供肿瘤的临床分期、病理特征以及临床诊治现状等全国性临床流行病学资料,推广简单易行的研究设计,为制定具体防治措施提供临床依据。[方法]按传统地理区域划分中国大陆七个大区,使用便利抽样的方法从每个大区选定一家三级甲等医院;以随机抽样的方式从1999年到2008年每年抽取除1、2月份以外一个月的所有病例进行回顾调查。将研究估算的病例数与医院实际的病例数进行比较;将各大区的女性人口分布与各大区10年累计实际与估算病例数分布进行比较。[结果]所有医院都按统一标准和要求完成了病案摘录,各协作中心的病案摘录合格率都在95.00%以上,总体资料合格率为98.69%;研究估算病例数与实际收治病例数10年间都呈逐年上升的趋势;华北、华南、华东和西北四个大区所选医院对该大区具有较好的代表性。研究估算的病例数与实际收治的病例数在七个大区之间的分布相似。[结论]按大区划分,使用便利抽样与随机抽样相结合的方法抽取的病案资料能够在一定程度上代表总体。样本资料能够反映全国范围内女性原发乳腺癌的临床特征和治疗现状,在现有的人群肿瘤登记资料的基础上为肿瘤的进一步防治提供临床依据。

[Purpose] To explore the methodology on designing a study which is easy to conduct, nationwide representative and reliable to illustrate the disease profile such as staging,diagnosis, treatment and prognosis based on female primary breast cancer in the existed hospital clinical records. [Methods] According to the geographic distribution, China's Mainland was divided into 7 regions; from each region,one tertiary hospital was selected by using convenience sampling. With the exception of January and February,one month was randomly selected to represent each year from 1999 to 2008 at each hospital. The distribution of female population and the actual number of visits were compared region by region. The estimated data were compared with the actual num- ber of visits. [Results] Investigation had been completed with the same standardized procedures by all sites,up to the requirement of quality control. More than 95.00%of questionnaires had met the criteria of quality control in each area. Overall,98.69%of the collected data was valid. Annual increasing trend was seen both in the sample cases collected and the actual visits in 10 years. Hospitals selected from North ,South ,East and North-west China were to some extent representa- tive of its region. Similar distributions were observed between the actual visits and the estimated cases of the 7 regions. [Conclusions] Based on the 7 geographical regions,the combined use of convenience and random sampling is financially and scientifically feasible in designing a nation- wide multi-center clinical epidemiologic investigation.