Meta-analysis provides evidence-based interpretation guidelines for the clinical significance of mean differences for the FACT-G, a cancer-specific quality of life questionnaire
Meta-analysis provides evidence-based interpretation guidelines for the clinical significance of mean differences for the FACT-G, a cancer-specific quality of life questionnaire
复制标题
DOI：
10.2147/prom.s10621
复制
发表时间：
2010-01-01
影响因子：
2.1
通讯作者：
Eisenstein, Amy
中科院分区：
文献类型：
--
作者：
King, Madeleine T.;Cella, David;Eisenstein, Amy
我们的目的是从一系列临床相关锚点出发，结合专家对临床意义的判断，为癌症治疗功能评估（FACT-G）制定循证解释指南，FACT-G是一种癌症特异性健康相关生活质量（HRQOL）工具。三位具有多年癌症患者管理经验并在临床研究中使用HRQOL结果的临床医生回顾了71篇论文。对FACT-G结果不了解，他们考虑与每个FACT-G平均差异相关的临床锚点，预测HRQOL的哪些维度会受到影响，以及影响是微不足道、小、中等还是大。这些大小类别是根据临床相关性来定义的。然后将专家的判断与FACT-G平均差异联系起来，并计算每个规模类的反方差加权平均差异。1118个横断面比较的小、中、大差异（95%置信区间）如下：身体健康1.9（0.6-3.2）、4.1（2.7-5.5）、8.7 (5.2-12)；功能性健康2.0 (0.5 - -3.5),3.8 (2.0 - -5.5),8.8 (4.3 -13);情绪幸福感1.0（0.1-2.6）、1.9(0.3-3.5)，差异不显著；社会幸福感0.7(-0.7 ~ 2.1),0.8(-2.9 ~ 4.5)，差异不大。436个纵向比较的结果往往小于相应的横断面结果。这些结果增加了FACT-G的其他解释指南，包括样本量、功率计算和使用FACT-G的癌症临床试验的解释。
Our aim was to develop evidence-based interpretation guidelines for the Functional Assessment of Cancer Therapy-General (FACT-G), a cancer-specific health-related quality of life (HRQOL) instrument, from a range of clinically relevant anchors, incorporating expert judgment about clinical significance. Three clinicians with many years' experience managing cancer patients and using HRQOL outcomes in clinical research reviewed 71 papers. Blinded to the FACT-G results, they considered the clinical anchors associated with each FACT-G mean difference, predicted which dimensions of HRQOL would be affected, and whether the effects would be trivial, small, moderate, or large. These size classes were defined in terms of clinical relevance. The experts' judgments were then linked with FACT-G mean differences, and inverse-variance weighted mean differences were calculated for each size class. Small, medium, and large differences (95% confidence interval) from 1,118 cross-sectional comparisons were as follows: physical well-being 1.9 (0.6-3.2), 4.1 (2.7-5.5), 8.7 (5.2-12); functional well-being 2.0 (0.5-3.5), 3.8 (2.0-5.5), 8.8 (4.3-13); emotional well-being 1.0 (0.1-2.6), 1.9 (0.3-3.5), no large differences; social well-being 0.7 (-0.7 to 2.1), 0.8 (-2.9 to 4.5), no large differences. Results from 436 longitudinal comparisons tended to be smaller than the corresponding cross-sectional results. These results augment other interpretation guidelines for FACT-G with information on sample size, power calculations, and interpretation of cancer clinical trials that use FACT-G.