Is the Use of Dialysis Associated With an Increased Risk of Death in COVID-19-Related Acute Kidney Injury?

introductionAcute kidney injury (AKI) is frequently reported in the setting of severe acute respiratory syndrome coronavirus 2 (SARS‑CoV‑2) infection. The aim of our work is to evaluate the impact of acute dialysis use on mortality in patients with AKI during the coronavirus disease 2019 (COVID-19) pandemic.

methodsThis is a retrospective study conducted in the Hassan II University Hospital of Fez, Morocco. From July 2020 to December 2021, we included all patients admitted to a COVID-19 unit with acute kidney injury defined according to Kidney Disease Improvement Global Outcomes 2012 (KDIGO 2012) criteria. Our patients were older than 18 years, and SARS-CoV-2 infection was confirmed by a positive RT-PCR test or thoracic CT scan imaging. Patients with end-stage renal disease (ESRD) and pregnant women were excluded from our study.

resultsThe total number of patients hospitalized in the COVID-19 unit during the study period was 2560, including 206 in an intensive care setting. We included 61 patients with AKI, with an incidence in the intensive care unit (ICU) setting of 15.5%. Eighty percent of patients had respiratory distress on admission, which was the main reason for consultation. Stage 1 AKI was found in 1.6% of patients, 25.8% had stage II AKI, and 72.6% had KDIGO stage 3 AKI. The main etiology of AKI was acute tubular necrosis. Lung involvement secondary to infection was severe in 18 patients; 21 had moderate involvement. In our study, twenty-one of our patients (34.4%) were hospitalized in an ICU. Thirteen of our patients were intubated (21.1%). Twenty-one (34.4%) patients were hemodynamically unstable and were put on vasoactive drugs. Twenty-three (37.7%) of our patients received at least one session of conventional acute hemodialysis with an average duration of 2.1 hours ± 0.9 (1-3.5). The indication was overload (27%), severe metabolic acidosis (1.6%), threatening hyperkalemia (1.6%), and symptomatic hyperuremia (62%). The evolution was marked by a return to baseline renal function in two patients, partial improvement in 35 of them at discharge, and no improvement in 24 patients. We recorded a death rate of 34.4% (n=21). In a univariate analysis, we compared the demographic, clinical, paraclinical, and dialytic characteristics of the dialysis and non-dialysis groups. There was a significant difference between unstable, intubated patients and those hospitalized in the ICU in the dialysis group, with respective p-values of p=0.0001, p=0.0001, and p=0.01. We noticed there were more deaths in the dialysis group than in the non-dialysis group; this difference was statistically significant with a p-value of 0.005. In multivariate analysis, a logistic regression model was performed to test the relationship between dialysis and COVID-19 mortality while adjusting for other co-factors. The final model did not show a significant association between dialysis and mortality (p = 0.150, OR: 2.578 [0.710-9.364]). The only factor that remained independently significant was admission to the intensive care unit (p = 0.004, OR: 6.732 [1.847-24.540]).

conclusionAKI is a frequently encountered complication in patients with COVID-19, especially those hospitalized in the ICU. In the context of the SARS-CoV-2 infection, the use of at least one dialysis session seems to represent an excess risk of mortality related to AKI.