Hemodialysis Access Outcomes for Patients with Unstable Housing

End-stage renal disease has traditionally been noted to disproportionately affect patients with fewer resources. Our goal was to assess practice patterns and outcomes in patients with unstable housing undergoing permanent hemodialysis access creation.

A retrospective, single-center review of patients with unstable housing was conducted. Perioperative and long-term outcomes were assessed. Univariable and multivariable analysis as well as Kaplan-Meier analysis were performed.

There were 144 patients (63% male) identified with a mean age of 57 years. Thirty-three percent of patients had no current permanent residence, and 81% had at least 3 addresses the year before operation. Access type included brachiocephalic (48%), brachiobasilic (19%), radiocephalic fistulas (11%), and prosthetic grafts (19%). Thirty-day readmission was seen in 27.8% of patients. Kaplan-Meier analysis showed that 1-year and 4-year patient survival were 94.4% and 80.6%, respectively. Multivariable analysis demonstrated that newly formed access (odds ratio [OR] 3.66, 95% CI 1.02 to 13.16, p = 0.05), absence of a permanent residence (OR 2.92, 95% CI 1.15 to 7.44, p = 0.03), and female gender (OR 2.86, 95% CI, 1.18 to 7.14, p = 0.02) were associated with 90-day readmission. Multivariable analysis of mortality revealed that previous stroke (hazard ratio [HR] 7.15, 95% CI 1.93 to 26.5, p = 0.003), history of alcohol use disorder (HR 4.55, 95% CI 1.22 to 16.99, p = 0.024), and age (HR 1.10, 95% CI, 1.02 to 1.18, p = 0.017) were associated with 4-year mortality; housing instability was not associated with decreased survival. Preoperative tunneled dialysis catheter (HR 1.63, 95% CI 1.02 to 2.61, p = 0.04) was associated with 4-year reintervention and frequent address change (HR 0.47, 95% CI 0.27 to 0.81, p = 0.01) was found to be a protective factor against long-term reintervention.

Poor outcomes in patients with unstable housing were primarily driven by comorbidities. Lack of permanent residence was significantly associated with readmission.