Icodextrin is Associated with a Lower Mortality Rate in Peritoneal Dialysis Patients. Issue 3 (May 2019)

• Record Type: Journal Article
• Authors: Yang, Ju-Yeh; Chen, Likwang; Peng, Yu-Sen; Chen, Yun-Yi; Huang, Jenq-Wen; Hung, Kuan-Yu
• Is Part Of: Peritoneal dialysis international. Volume 39:Issue 3(2019)
• Page Start: 252
• Page End: 260
• Publication Date: 2019-05
• Subjects: Hospitalization -- major adverse cardiovascular events; Peritoneal dialysis -- Periodicals; Continuous ambulatory peritoneal dialysis -- Periodicals; 617.461059
• DOI: 10.3747/pdi.2018.00217
• Languages: English
• ISSNs: 0896-8608

Abstract

Background: Icodextrin (ICO) improves fluid removal in peritoneal dialysis (PD) patients. However, whether physiological benefits of ICO translate into patient survival remains unclear. We examine the association of ICO and clinical outcomes. Methods: We identified patients who initiated long-term PD from the National Health Insurance Research Database of Taiwan. We matched ICO users with non-users according to propensity score and survival status when ICO was prescribed. We utilized time-dependent analyses to avoid immortal time bias. Additional competing risk models were utilized for the outcomes except for death. The outcomes of interest were time to death, technique failure, peritonitis, major adverse cardiovascular events (MACE), and hospitalization. Results: A total of 4, 914 PD patients were enrolled and 2, 836 PD patients (57.7%) were identified as ICO users. The ICO users had significantly better overall survival (hazard ratio [HR] 0.74; 95% confidence interval [CI] 0.63 – 0.86), especially among early ICO users (HR 0.64; 95% CI 0.54 – 0.77, p value for interaction: 0.007). The ICO users were associated with higher risk of peritonitis (subdistribution HR 1.22, 95% CI 1.06 – 1.14) and hospitalization (subdistribution HR 1.14, 95% CI 1.05 – 1.24), considering competing risk of death. However, when considering ICO use as a time-varying covariate, ICO users shared similar risks for technique failure, peritonitis, MACE, and hospitalization as non-users. The effect of ICO on mortality was especially prominent among those early users. Conclusions: After adjustments for immortal time biases, ICO users were significantly associated with approximately 20% reduction in mortality, especially among early users.