P1343does “Fistula First” Fits to Elderly?

Abstract

Background and Aims The incident patients on hemodialysis (HD) are becoming older. However, the optimal type of initial permanent vascular access (VA) among the elderly is controversial. Patient comorbidities and life expectancy are important considerations in whether to place an arteriovenous fistula (AVF) or graft (AVG). We design an observational study to compare clinical outcomes of elderly (≥65 year old) versus younger patients, who underwent for first VA placement before initiation of renal replacement therapy, between January 2014 and December 2018. Method We evaluated successful use of VA, requirement of surgical interventions before successful use, VA in use after the first and third months on HD and clinical outcomes, until December 2019. The comorbidity burden was calculated through age-adjusted Charlson Comorbidity Index (aCCI). We also evaluated the impact of comorbidity burden on the VA type on HD start and mortality after HD initiation. Results We identified 252 predialysis patients who underwent for VA placement in our center. We created two groups based on age at the time of VA placement: there were 199 (79,0%) with age ≥ 65 years (the elderly group), and 53 (21,0%) younger patients. The elderly group presented a mean age of 76,3 ± 6,4 (maximum of 92) years on first VA placement; in the younger group, the mean age was 54,5 ± 9,1 (minimum of 26) years. The following analysis are presented for elderly versus younger group. On both groups there were a predominance of male gender (66,8%; 73,6%; p=0,498) and caucasian race (95,0%; 88,7%; p=0,193). At time of referral for AV placement, both groups presented similar mean estimated glomerular filtration rate by CKD-EPI equation (11,7 ± 3,2; 11,2 ± 3,2 mL/Kg/1,72m2; p=0,391). Elderly group presented a significant higher aCCI (7,3 ± 1,74; 9,0 ± 1,9; p<0,001). The groups were also different in smoking status (6,0%; 30,8%; p<0,001). There were no differences on kidney disease etiology between groups, with diabetes being the most prevalent (23,1%; 24,5%; p=0,856). For all patients, the first VA placed was AVF. Only two patient placed an AVG on second and third vascular accesses. The median number of VA placed were similar between the two groups [1,0 (1 to 4); 1,0 (1 to 2); p= 0,811], likewise the occurrence of early complications (9,5%; 5,7%; p=0,583) and the need for surgical interventions (46,7%; 47,2%; p=1,000). In both groups, the majority of patients started HD (80,4%; 90,6%; p=0,103), with similar successful use of the VA (68,1%; 75,0%; p=0,474). In multivariate logistic regression, proteinuria (measured at time of referral for AV placement) and heart failure (HF) were predictors to HD initiation through a central venous catheter (CVC). This model classified correctly 74,9% of cases, with an HF odds ratio (OR) of 4,149 [confident interval (CI) of 1,721 to 10,000] and a proteinuria OR of 1,148 (CI: 1,047 to 1,259). After the first month on HD, 34,8% of elderly patients needed a CVC, a number significantly different from the younger group (15,9%; p=0,023). The same result was observed after the third month (22,2%; 7,1%; p=0,028). During the time of follow-up, the mortality rate was higher in the elderly group who started HD (log Rank test = 0,004), with a median survival of 29,3 (0,1 to 89,8) months, when compared to the younger group [median survival of 38,3 (0,1 to 76,9) months]. Conclusion There were no difference in the kind of VA on HD start (definitive VA versus CVC) between the two groups. However, elderly patients presented more fistula failure in the first three months after HD initiation. The need of CVC due to nonfunctioning AVF on the first and three months after HD initiation was higher in the elderly. The analysis of the patients who started HD showed that the elderly group presented a significant reduced survival when compared to the youngest patients.