Abnormal Sodium is a Predictor for Respiratory Failure and Mortality in Hospitalized Patients With COVID-19

Abstract

Background: Hypernatremia and hyponatremia (serum sodium > 145 mmol/L and < 135 mmol/L, respectively) are independent risk factors for excess mortality in patients with bacterial pneumonia. We sought, for the first time, an association of sodium [Na] abnormalities with mortality, need for advanced respiratory support and Acute Kidney Injury (AKI) in hospitalized patients with coronavirus disease 19 (COVID-19). Methods: This retrospective, longitudinal, cohort study included 488 adults, 277 males and 211 females, with a median age of 68 years, who were hospitalized with COVID-19 to two hospitals in London over an 8-week period (February to May 2020). Results: The in-hospital mortality rate was 31.1% with a medial length of stay of 8 days. High [Na] levels at any timepoint during hospital stay were associated with significantly increased mortality rate (56.6% vs 21.1% in patients who remained constantly normonatremic; odds ratio 3.05, 95% CI 1.69-5.49; p<0.0001). On day 3 and on day 6, high [Na] values predicted mortality with an estimated odds ratio of 2.34 (95% CI 1.08 – 5.05, p=0.0014) and 2.40 (95% CI 1.18 - 4.85, p=0.001), respectively. Non-survivors had a significantly higher 5-day rise in serum [Na] when compared to survivors (3.60 mmol/L vs 1.14 mmol/L respectively, p<0.05). Patients with low serum [Na] levels on admission had a 2.18-fold increase (95% CI 1.34-3.46, p=0.001) in the likelihood of needing advanced ventilatory support compared to those with normal [Na] (31.7% vs 17.5%, respectively). However, exposure to hyponatremia at any timepoint, including at presentation or on day 3 or day 5, was not associated with excess risk of

death. AKI affected 37.1% of patients (21.3%, 7.4% and 8.4% stages 1, 2 and 3 respectively) but was not related to serum sodium values. Conclusions: In hospitalized COVD-19 patients, hypernatremia at any timepoint was associated with excess mortality, suggesting that [Na] concentration may facilitate risk stratification. In addition, whilst our data cannot prove causality, these findings highlight the significance of judicious rehydration in such patients.