Renal outcomes in the ONTARGET study
Renal outcomes in the ONTARGET study
复制标题
ONTARGET 研究中的肾脏结果
DOI：
10.1016/s0140-6736(08)61862-0
复制
发表时间：
2008
期刊：
影响因子：
--
通讯作者：
J. Wetzels
中科院分区：
文献类型：
--
作者：
J. Wetzels
我们对ONTARGET试验1的主要担忧是，与单药治疗组相比，联合治疗组的脱落率较高。这种较高的脱落率与双阻滞诱导的低血压直接相关，导致留在研究组中的受试者之间存在系统差异（即偏倚）。引入的偏倚方向是不可预测的，但可能对结果有贡献。2当存在差异脱落时，应用意向治疗分析可能会无意中导致进一步的偏倚。3联合治疗组脱落的患者可能未接受任何血管紧张素转换酶抑制剂或血管紧张素受体阻滞剂治疗。因此，他们将更容易受到不利的肾脏终点的影响，结果被计入联合治疗组。这种偏见不利于联合用药组。联合治疗组的脱落率（约4.7%）与治疗的效应量不成比例。双阻滞组和单阻滞组之间主要终点的风险只有1%的绝对差异。敏感性分析（即，假设所有脱落者都经历了有利或相反的不利结局）表明，1%的效应量缺乏统计学稳健性。此外，这项研究对一般糖尿病肾病人群的一般能力有限，认为只有13%的人有微量蛋白尿，31%的人血压正常。
Our main concern with the ONTARGET trial1 is the substantial drop out rate in the combination therapy group compared with the mono therapy group. This higher dropout rate is directly related to dual-blockadeinduced hypo tension, resulting in syste matic differ ence (ie, bias) between groups in the participants who remained. The direction of bias introduced is un predictable but probably contributed to the outcome. 2 Applying intention-to-treat analysis when there is differential dropout can inadvertently lead to further bias. 3 Pat ients who dropped out of the com bination therapy group might sub sequently not have received any treat ment with angiotensin-con verting-enzyme inhibitors or angiotensin-receptor blockers. Thus they will have been more sus ceptible to ad verse renal end points, with out comes being counted towards the com bination therapy group. This bias weighs un favourably for the combination group. The magnitude of the dropout rate in the combination therapy group (about 4· 7%) is dis proportionately large com pared with the effect size attri buted to the treat ment. There was only about a 1% absolute difference in the risk of primary endpoints between the dualblockade and single-blockade groups. Sensitivity analysis (ie, assuming all dropouts experience a favour able or, conversely, adverse outcome) suggests that the effect size of 1% lacks statistical robustness. Furthermore, the study has limited generalis ability to the general diabetic nephro pathy population, con sidering that only 13% had micro albuminuria and 31% were normotensive.