Supplementary Material for: Dialysis Modality and Mortality in Heart Failure: A Retrospective Study of Incident Dialysis Patients

<b><i>Introduction:</i></b> Prior studies reported lower mortality with hemodialysis (HD) compared to peritoneal dialysis (PD) in patients with heart failure (HF). We examined mortality rate by initial dialysis modality in incident dialysis patients with a history of HF using contemporary data and methods that ensure comparable HD and PD groups. <b><i>Methods:</i></b> Retrospective cohort study using administrative databases in Ontario, Canada. Adults (age 50–80) with a history of HF who initiated maintenance dialysis between April 1, 2007 and March 31, 2016 were included. We excluded patients typically ineligible for PD as an initial modality (dialysis start in hospital, dementia, long-term care facility residency). We determined the cause-specific hazard ratio (transplant as a competing event) between initial dialysis modality (HD vs. PD) and all-cause mortality using an intention-to-treat approach. <b><i>Results:</i></b> We included 2,199 patients with HF who initiated maintenance dialysis (77% HD and 23% PD). There were 1,152 (67.8%) and 340 (68.1%) mortality events over a median follow-up of 2.4 and 2.5 years in the HD and PD groups, respectively. Patients initiating HD versus PD was not associated with the mortality rate (adjusted hazard ratio 1.0, 95% CI 0.9–1.1). Similar results were seen in analyses censoring at modality switches and treating modality as time-varying. <b><i>Conclusions:</i></b> We found no difference in mortality by initial dialysis modality. Our data support the current practice of selecting dialysis modality based on patient preference for patients with pre-existing HF.

<b><i>引言：</i></b> 既往研究显示，心力衰竭（heart failure, HF）患者接受血液透析（hemodialysis, HD）的死亡率低于腹膜透析（peritoneal dialysis, PD）。本研究依托当代研究数据与分析方法，确保血液透析与腹膜透析组具有可比性，旨在探讨有心力衰竭病史的新发透析患者中，初始透析方式与死亡率的关联。<b><i>方法：</i></b> 本研究为回顾性队列研究，采用加拿大安大略省的行政数据库开展。纳入2007年4月1日至2016年3月31日期间启动维持性透析、年龄50~80岁且有心力衰竭病史的成年患者。排除初始透析阶段通常无法接受腹膜透析的患者，包括透析起始于住院期间、患有痴呆、长期入住疗养机构的患者。本研究采用意向性治疗分析方法，计算初始透析方式（血液透析与腹膜透析）与全因死亡率之间的病因特异性风险比（以肾移植作为竞争事件）。<b><i>结果：</i></b> 本研究共纳入2199例启动维持性透析的心力衰竭患者，其中77%接受血液透析，23%接受腹膜透析。血液透析组与腹膜透析组的中位随访时间分别为2.4年与2.5年，两组分别发生1152例（占比67.8%）与340例（占比68.1%）死亡事件。初始接受血液透析与腹膜透析的患者，其死亡率无显著差异（校正后风险比为1.0，95%置信区间0.9~1.1）。在对透析方式转换进行删失处理以及将透析方式视为时变协变量的分析中，同样得到了相似结果。<b><i>结论：</i></b> 本研究未发现初始透析方式与死亡率存在显著差异。本研究数据支持当前临床实践：对于合并基础心力衰竭的患者，可根据患者意愿选择透析方式。