SHORT- AND LONG-TERM SURVIVAL AND CAUSE-SPECIFIC MORTALITY IN PATIENTS WITH UNPLANNED VS. PLANNED HEMODIALYSIS INITIATION

Planned hemodialysis (HD) initiation with a matured vascular access is recommended in several international guidelines because it is associated with better survival than unplanned HD initiation using a temporary vascular catheter. Previous studies have shown higher short-term mortality within one year among patients with unplanned HD initiation, while long-term survival differences remain unclear. However, few studies have evaluated cause-specific mortality patterns, particularly in long-term observational cohorts.

We consecutively assessed newly declared end-stage kidney disease (ESKD) patients who started hemodialysis (HD) between January 1, 2007 and December 31, 2014 at Nara Medical University Hospital. We excluded patients who had dialysis-dependent acute kidney injury (AKI) leading to ESKD, recovered from AKI, underwent continuous kidney replacement therapy as the only dialysis modality, or had HD for extra-renal indications. Patients were stratified into two groups (unplanned vs. planned HD initiation). Survival was compared using the Kaplan–Meier method and multivariable Cox regression analysis adjusted for age, sex, and Charlson Comorbidity Index. Landmark analyses were performed for three time windows (0–1 year, 1–5 years, and 5–15 years). The primary outcome was all-cause mortality until December 31, 2020. Causes of death were categorized as heart failure, infection, cancer, cerebrovascular disease, ischemic heart disease, other, or unknown, and compared between the two groups.

Of the 460 newly declared ESKD patients who were assessed for eligibility, 345 patients (172 unplanned and 173 planned HD initiation) were included in this study. The median follow-up duration of the entire cohort was 4.6 years (range, 1.5–7.8 years). In the landmark analysis, unplanned HD initiation was significantly associated with higher mortality during the first year (HR 1.89, 95% CI 1.06–3.39, P = 0.03), but not during 1–5 years (HR 1.31, 95% CI 0.85–2.00, P = 0.2) or 5–15 years (HR 0.70, 95% CI 0.42–1.19, P = 0.2). After adjustment for age, sex, and Charlson Comorbidity Index, the association remained significant only in the first year (adjusted HR 2.32, 95% CI 1.28–4.21, P = 0.005). In the 0–1-year period, the leading causes of death were infection (32.3% in unplanned vs. 33.3% in planned) and cancer (6.5% vs. 22.2%). During 1–5 years, infection (22.2% vs. 7.5%) and cancer (22.2% vs. 15.0%) were predominant. In the 5–15-year period, infection (29.2% vs. 25.0%) remained the most frequent cause, followed by cerebrovascular disease (12.5% vs. 11.1%). The distribution of causes of death did not differ significantly between the two groups across all time windows (χ² test, P = 0.2, 0.5, and 0.5 for 0–1, 1–5, and 5–15 years, respectively). Figure 1A shows unadjusted Kaplan–Meier survival curves, while Figure 1B presents adjusted survival curves derived from multivariable Cox regression analyses for the three landmark time windows.

Unplanned HD initiation was independently associated with higher short-term mortality within the first year, but not with long-term survival. The distribution of causes of death did not differ, implying that the higher early mortality in unplanned HD patients was not associated with specific causes. Timely nephrology referral and planned HD initiation may improve early survival.