中危主动脉瓣狭窄患者SAVR与TAVR的院间及院内变异:PARTNER-2A分析
PARTNER-2A试验中,中危主动脉瓣狭窄患者随机接受SAVR(n=1017)或TAVR(n=1011),随访2年。混合效应模型显示,SAVR与TAVR的死亡院间变异相似,但SAVR的卒中院间变异大于TAVR;院内两种术式事件相似。
要点
- SAVR与TAVR的死亡院间变异相似,但SAVR卒中院间变异更大。
- 病例构成和中心试验量低解释了部分死亡变异,卒中变异解释较少。
- 除1个中心外,各中心内SAVR与TAVR死亡风险比95%CI均跨1.0。
核心结论
- SAVR与TAVR的死亡院间变异相似,但SAVR的卒中院间变异大于TAVR。
- 院内SAVR与TAVR事件相似。
- 基于性能的试验应分析中心变异及其来源。
研究设计与样本
- PARTNER-2A随机试验的事后分析,中危主动脉瓣狭窄患者随机接受SAVR(n=1017)或TAVR(n=1011),在54个中心使用SAPIEN XT装置,随访2年。
- 采用混合效应模型量化院间和院内变异。
主要发现
- 共336例死亡(SAVR 170例,TAVR 166例)和176例卒中(SAVR 85例,TAVR 91例)。
- SAVR死亡风险比院间范围0.52-1.93,TAVR为0.49-2.03。
- SAVR卒中风险比院间范围0.44-2.26,TAVR未检测到变异。
- 病例构成和中心试验量低解释了SAVR死亡院间变异的37%、TAVR的73%,但仅解释SAVR卒中变异的14%。
- 除1个中心外,各中心内SAVR与TAVR死亡风险比95%CI均跨1.0。
临床怎么用
- 解读基于性能的试验结果时需考虑中心变异及其来源。
- 中危主动脉瓣狭窄患者选择SAVR或TAVR时应关注中心层面的卒中变异。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
OBJECTIVES: Multisite procedure-based randomized trials may be confounded by performance variability and variability among sites. Therefore, we studied variability in mortality and stroke after patients were randomized to surgical (SAVR) or transcatheter aortic valve replacement (TAVR) in the Placement of Aortic Transcatheter Valves-2A (PARTNER-2A) randomized trial. METHODS: Patients at intermediate risk for SAVR were randomized to SAVR (n = 1017) or TAVR (n = 1011) with a SAPIEN XT device (Edwards Lifesciences, Irvine, Calif) at 54 sites. Patients were followed to 2 years. A mixed-effect model quantified variability at intersite and intrasite levels. RESULTS: There were 336 deaths (SAVR 170, TAVR 166) and 176 strokes (SAVR 85, TAVR 91). Intersite variability for mortality was similar across sites for SAVR (hazard ratios ranging from 0.52-1.93 among sites) and TAVR (hazard ratios ranging from 0.49-2.03), but intersite variability for stroke was greater for SAVR (hazard ratios ranging from 0.44-2.26) than for TAVR (no detectable variability). Case mix and lower site trial volume accounted for 37% of mortality intersite variability for SAVR and 73% for TAVR, but only 14% for stroke for SAVR. Intrasite mortality hazard ratios demonstrated all but 1 site's 95% confidence interval overlapped 1.0, indicating generally similar SAVR and TAVR mortalities within sites. CONCLUSIONS: Intersite variability was similar for mortality in SAVR and TAVR, but variability for stroke was greater for SAVR than for TAVR. Intrasite events were similar for both SAVR and TAVR. These findings suggest that in performance-based trials, site variability and its sources should be taken into account in analyzing and interpreting trial results.
引用
- Greason KL, Blackstone EH, Rajeswaran J, et al. Inter- and intrasite variability of mortality and stroke for sites performing both surgical and transcatheter aortic valve replacement for aortic valve stenosis in intermediate-risk patients. J Thorac Cardiovasc Surg. 2019. PMID: 31350027. doi:10.1016/j.jtcvs.2019.04.112
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。