Minicozzi Pamela, Innos Kaire, Sánchez Maria-José, Trama Annalisa, Walsh Paul M, Marcos-Gragera Rafael, Dimitrova Nadya, Botta Laura, Visser Otto, Rossi Silvia, Tavilla Andrea, Sant Milena
Analytical Epidemiology and Health Impact Unit, Fondazione IRCCS Istituto Nazionale dei Tumori, Milan, Italy.
Department of Epidemiology and Biostatistics, National Institute for Health Development, Tallinn, Estonia.
Eur J Cancer. 2017 Oct;84:335-353. doi: 10.1016/j.ejca.2017.07.015. Epub 2017 Sep 1.
Cancer registries (CRs) are fundamental for estimating cancer burden, evaluating screening and monitoring health service performance. Stage at diagnosis-an essential information item collected by CRs-has been made available, for the first time, by CRs participating in EUROCARE-5. We analysed the quality of this information and estimated stage-specific survival across Europe for CRs with good data quality.
Sixty-two CRs sent stage (as TNM, condensed TNM or extent of disease) for 15 cancers diagnosed in 2000-2007. We assessed the quality, partly by comparing stage according to the three systems. We also developed procedures to reconstruct stage (categories: local, regional, metastatic and unknown) using information from all three systems, thus minimising the amount of missing information.
Moderate-to-excellent stage concordance was found for practically all 24 CRs, for which it was possible to compare at least two staging systems. However, since stage was often incorrectly assigned, and information on the presence/absence of metastases was often lacking, data on only 7/15 cancers from 34/62 CRs (15 countries) were of sufficient quality for further analysis. Cases diagnosed ≥70 years had more advanced (or lacking) stage- and worse stage-specific survival than those <70 years.
Many European CRs collect and record reasonably accurate stage information. Others have difficulties. Both the completeness of primary data and the accuracy of stage coding need to be improved in order for CRs to fulfil their expanding roles in cancer control. We propose our stage reconstruction/checking procedures as a means of fully exploiting the stage information provided by EUROCARE CRs. More advanced (or lacking) stage at diagnosis plus poorer stage-specific survival in the elderly are worrying.
癌症登记处（CRs）对于估计癌症负担、评估筛查以及监测卫生服务绩效至关重要。诊断分期是CRs收集的一项重要信息，参与EUROCARE - 5的CRs首次提供了该信息。我们分析了此信息的质量，并对数据质量良好的CRs在欧洲范围内按分期估计了生存率。
62个CRs发送了2000 - 2007年诊断的15种癌症的分期（如TNM、简化TNM或疾病范围）。我们通过根据三种系统比较分期来部分评估质量。我们还开发了程序，利用来自所有三种系统的信息重建分期（类别：局部、区域、转移和未知），从而将缺失信息的量降至最低。
几乎所有24个CRs都发现了中度至高度的分期一致性，对于这些CRs，可以比较至少两种分期系统。然而，由于分期经常被错误分配，并且关于是否存在转移的信息经常缺失，来自62个CRs（15个国家）中的34个的15种癌症中只有7种癌症的数据质量足以进行进一步分析。诊断时年龄≥70岁的病例比年龄<70岁的病例分期更晚（或分期信息缺失）且分期特异性生存率更差。
许多欧洲CRs收集并记录了合理准确的分期信息。其他CRs则存在困难。为了使CRs在癌症控制中发挥日益扩大的作用，需要提高原始数据的完整性和分期编码的准确性。我们提出我们的分期重建/检查程序，作为充分利用EUROCARE CRs提供的分期信息的一种手段。诊断时分期更晚（或分期信息缺失）以及老年人分期特异性生存率更低令人担忧。
