Lower short-term mortality in ICU patients on chronic dialysis than in those requiring acute dialysis.

Background: In ICU patients, we aimed to describe the outcomes of those with end-stage renal disease (ESRD) as compared to those requiring acute renal replacement therapy (RRT).

Methods: Retrospective analysis of all adult patients admitted to a general, university hospital ICU from 2005 to 2012. ESRD was defined as use of chronic RRT >90 days prior to admission.

Results: We included 5927 patients of whom 1004 (17%) received acute RRT and 161 (3%) had pre-existing ESRD requiring RRT. Thirty-day mortality was 42% vs 28% for acute RRT vs ESRD patients (adjusted hazard ratio (aHR) 0.90 (0.61-1.34)), and 16% for those not requiring RRT (aHR 0.91 (0.60-1.38) compared to ESRD patients). Ninety-day mortality was 55% vs 45% for acute RRT vs ESRD patients (aHR 0.96 (0.70-1.31)), and 22% for those not requiring RRT (aHR 1.19 (0.84-1.67) compared to ESRD patients). Ninety-day ESRD survivors were younger, less severely ill and needed less vasopressor treatment than 90-day ESRD non-survivors. Five-year mortality was 68% vs 69% for acute RRT vs ESRD patients (aHR 1.06 (0.81-1.39)), and 38% for those not requiring RRT (aHR 1.31 (0.99-1.74) compared to ESRD patients).

Conclusions: The crude mortality for patients with pre-existing ESRD was high. Short-term mortality was within range of those not receiving RRT when adjusted for confounders. The severity of acute illness and the burden of comorbidities may be more important than the lack of kidney function per se for the short-term prognosis of RRT patients in the ICU.