Survival trends of US dialysis patients with heart failure: 1995 to 2005

Abstract

Background and objectives: Congestive heart failure (CHF) is a major risk factor for death in end-stage kidney disease; however, data on prevalence and survival trends are limited. The objective of this study was to determine the prevalence and mortality effect of CHF in successive incident dialysis cohorts.

Design, setting, participants, & measurements: This was a population-based cohort of incident US dialysis patients (n = 926,298) from 1995 to 2005. Age- and gender-specific prevalence of CHF was determined by incident year, whereas temporal trends in mortality were compared using multivariable Cox regression.

Results: The prevalence of CHF was significantly higher in women than men and in older than younger patients, but it did not change over time in men (range 28% to 33%) or women (range 33% to 36%). From 1995 to 2005, incident death rates decreased for younger men (≤70 years) and increased for older men (>70 years). For women, the pattern was similar but less impressive. During this period, the adjusted mortality risks (relative risk [RR]) from CHF decreased in men (from RR = 1.06 95% Confidence intervals (CI) 1.02-1.11 in 1995 to 0.91 95% CI 0.87-0.96 in 2005) and women (from RR = 1.06 95% CI 1.01-1.10 in 1995 to 0.90 95% CI 0.85-0.95 in 2005 compared with referent year 2000; RR = 1.00). The reduction in mortality over time was greater for younger than older patients (20% to 30% versus 5% to 10% decrease per decade).

Conclusions: Although CHF remains a common condition at dialysis initiation, mortality risks in US patients have declined from 1995 to 2005.

Figure 1.

Adjusted Cox survival curves for men with congestive heart failure (CHF) by grouped years of incidence (1995 to 1997, 1998 to 2000, 2001 to 2003, 2004 to 2005). Adjusted for age at study start, race, diabetes, hypertension, coronary artery disease, peripheral vascular and cerebrovascular disease, current tobacco use, chronic lung disease, neoplasm, alcohol dependence, body mass index, serum albumin, hematocrit, estimated GFR, and predialysis erythropoietin use. P < 0.001 for all group comparisons.

Figure 2.

Adjusted Cox survival curves for women with CHF by grouped years of incidence (1995 to 1997, 1998 to 2000, 2001 to 2003, 2004 to 2005). Adjusted for age at study start, race, diabetes, hypertension, coronary artery disease, peripheral vascular and cerebrovascular disease, current tobacco use, chronic lung disease, neoplasm, alcohol dependence, body mass index, serum albumin, hematocrit, estimated GFR, and predialysis erythropoietin use. P < 0.001 for all group comparisons.