To the Editor Kitz et al¹ evaluated the prognostic value of total mesorectal excision (TME) plane in patients with rectal cancer. Total mesorectal excision plane quality was grouped into 3 categories: mesorectal, intramesorectal, and muscularis propria. One end point was disease-free survival (DFS). The authors quantified the between-group differences using 3-year event-free rates and hazard ratios (HRs). Three-year DFS rate estimates for mesorectal, intramesorectal, and muscularis propria TME were 75.9% (95% CI, 73.1-78.8), 68.4% (95% CI, 61.6-76.0), and 67.2% (95% CI, 55.6-81.3), respectively. Because these confidence intervals overlap, it is unclear whether there is a true difference between certain pairs, eg, between intramesorectal and muscularis propria TME. Moreover, in Figure 2A,¹ the DFS curves extend up to 60 months. Thus, the 3-year event rate provides a local profile of DFS only. To use data after 3 years, the authors reported HRs for DFS (intramesorectal vs mesorectal TME: HR, 1.35; 95% CI, 1.01-1.80; muscularis propria vs mesorectal TME: HR, 1.73; 95% CI, 1.13-2.66). However, it is difficult to interpret HRs in the clinical context. The hazard is not a chance or probability measure and therefore is not equivalent to risk. Thus, an HR of 1.35 cannot be translated into a 35% risk increase. Also, no reference hazard value from mesorectal TME was provided. If mesorectal TME hazard is low, a 35% increase in hazard may not be clinically significant. Other issues and concerns in using HRs to quantify between-group differences have been discussed extensively.^2-4

致编辑Kitz等[ ^1]评估了直肠癌患者全直肠系膜切除（TME）平面的预后价值。总直肠系膜切除平面质量分为三类：直肠系膜，直肠内膜和固有肌层。终点是无病生存期（DFS）。作者使用3年无事件发生率和危险比（HRs）量化了组间差异。直肠，肠系膜内和固有肌层TME的三年DFS率估计分别为75.9％（95％CI，73.1-78.8），68.4％（95％CI，61.6-76.0）和67.2％（95％CI，55.6- 81.3）。由于这些置信区间重叠，因此尚不清楚某些对之间（例如，大肠直肠内和固有肌层TME之间）是否存在真正的差异。此外，在图2A中，¹ DFS曲线长达60个月。因此，三年事件发生率仅提供了DFS的本地资料。为了使用3年后的数据，作者报告了DFS的心率（直肠内与直肠中膜TME：HR，1.35； 95％CI，1.01-1.80；固有肌与直肠中膜TME：HR，1.73； 95％CI，1.13-2.66）。但是，很难在临床背景下解释HR。危害不是机会或概率度量，因此不等同于风险。因此，HR为1.35不能转化为35％的风险增加。同样，没有提供来自直肠系膜TME的参考危险值。如果直肠系膜TME危险性较低，则危险性增加35％在临床上可能并不重要。关于使用人力资源来量化组间差异的其他问题和关注，已经进行了广泛的讨论。^{2 -4}