Acheson Louise S, Zyzanski Stephen J, Stange Kurt C, Deptowicz Amy, Wiesner Georgia L
Department of Family Medicine, Comprehensive Cancer Center, Case Western Reserve University and University Hospitals of Cleveland, Cleveland, OH 44106-5036, USA.
J Clin Oncol. 2006 Dec 1;24(34):5395-402. doi: 10.1200/JCO.2006.07.2462. Epub 2006 Nov 6.
A detailed family history is important for cancer risk assessment, but obtaining it is time consuming and infrequently accomplished in practice. The Genetic Risk Easy Assessment Tool (GREAT) conducts a computer-administered family history interview and immediately generates a pedigree diagram in digital form. The purpose of this study was to validate family cancer histories produced by patients using the computer tool in comparison with pedigrees made by genetic counselors.
Patients scheduled for genetics consultation recorded their family histories using the GREAT, separate from their genetic counseling session. The presence of each relative; presence, type, and age at diagnosis of cancers; and cancer geneticist's risk assessment were compared for 120 pairs of pedigrees produced by counselors versus computer tool.
The automated telephone interview took a mean of 33.5 minutes and was highly acceptable to respondents. Ninety-four percent of first-degree relatives, 67% of second-degree relatives, and 38% of third-degree relatives were identical on paired pedigrees; computer-generated pedigrees included additional relatives. Sixty-three percent of all cancers were identified by both family histories, with 90% agreement on the type of cancer. There was very good agreement (kappa = 0.70; correlation = 0.77) between the geneticist's breast cancer risk assessments based on computer versus counselors' pedigrees. In a subsample of 61 users, test-retest reliability for the computer-administered questionnaire was high (phi = 0.94 for cancers in first-degree and phi = 0.91 in second-degree relatives).
The GREAT computer-administered questionnaire provides an acceptable, reliable, and valid way of collecting an unverified but extensive family history of cancer and displaying it as a pedigree, in an entirely automated process.
详细的家族病史对于癌症风险评估很重要，但获取它既耗时又在实际操作中很少能完成。遗传风险简易评估工具（GREAT）进行计算机管理的家族病史访谈，并立即生成数字形式的系谱图。本研究的目的是将患者使用该计算机工具生成的家族癌症病史与遗传咨询师绘制的系谱图进行比较，以验证其准确性。
安排进行基因咨询的患者使用GREAT记录他们的家族病史，这与他们的遗传咨询过程分开。比较了由咨询师和计算机工具生成的120对系谱图中每个亲属的存在情况、癌症诊断时的存在情况、类型和年龄，以及癌症遗传学家的风险评估。
自动电话访谈平均耗时33.5分钟，受访者对其接受度很高。在配对的系谱图中，94%的一级亲属、67%的二级亲属和38%的三级亲属是相同的；计算机生成的系谱图包含了更多的亲属。两种家族病史都识别出了63%的所有癌症，在癌症类型上的一致性为90%。基于计算机生成的系谱图与咨询师绘制的系谱图，遗传学家对乳腺癌风险的评估之间有非常好的一致性（kappa = 0.70；相关性 = 0.77）。在61名用户的子样本中，计算机管理问卷的重测信度很高（一级亲属中癌症的phi = 0.94，二级亲属中癌症的phi = 0.91）。
GREAT计算机管理问卷提供了一种可接受、可靠且有效的方式，以完全自动化的过程收集未经核实但广泛的癌症家族病史，并将其显示为系谱图。
