RESUMEN
INTRODUCTION: The prevalence of heart failure (HF) is more common in people with advanced non-dialysis chronic kidney disease (ND-CKD) compared to the general population. It is well known that HF with reduced ejection fraction (HFrEF) is associated with a higher risk of mortality in people with ND-CKD. However, the impact of HFrEF on progression into end-stage kidney disease (ESKD) is not well studied. Our study aimed to examine the independent association of HFrEF on progression to ESKD after correcting for confounding factors using two methods of propensity scoring. METHODS: This study used data from the Salford Kidney Study, a longitudinal study which has recruited more than 3,000 patients with ND-CKD since 2002. Patients without a history of HF during the recruitment questionnaire were included in the control group. Patients with a reported history of HF and echo showing left ventricular ejection fraction <40% at enrolment were included in the HFrEF group. Two propensity score methods were used to attenuate the effects of confounding factors between the two groups - propensity score matching (PSM) and inverse probability weighting (IPW). Univariate and multivariate Cox-regression analyses were performed. RESULTS: A total of 2,383 patients were included in the analysis. Patients with HFrEF had significantly higher median age and a higher percentage of male gender compared to patients with no HF (72.5 vs. 66.6 years and 71.8 vs. 61.1%, respectively). Univariate and 5 models of multivariate Cox-regression analysis showed that HFrEF in people with CKD was a strong predictor for a higher incidence of ESKD (model 5: hazard ratio 1.38; 95% confidence interval = 1.01-1.90; p = 0.044). The association between HFrEF and the risk of ESKD remained significant after using the PSM and the IPW methods. CONCLUSION: Patients with concomitant advanced ND-CKD and prevalent HFrEF were found to have a higher risk of ESKD when compared to patients with no HF. This risk persists despite the adjustment of confounding factors using PSM and IPW.