Outcomes of Patients on Maintenance Dialysis Hospitalized with COVID-19

• chronic dialysis
• COVID-19
• end-stage kidney disease

Coronavirus disease 2019 (COVID-19) due to severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has had devastating effects worldwide. Patients with kidney failure on dialysis may have a higher risk of worse outcomes. Reports from China found that these patients with SARS-CoV-2 had fewer symptoms and required less intensive care than expected (1). A recent observational study of hospitalized patients with kidney failure and COVID-19 reported 31% mortality (2). However, this study lacked a comparator group, and thus, it is unclear if this high mortality would be found in patients without kidney failure with a similarly high comorbidity burden. Therefore, we conducted this retrospective cohort study of patients with kidney failure hospitalized with COVID-19 in the Mount Sinai Health Care System (MSHS) and compared it with a propensity-matched cohort without kidney failure.

Only patients age ≥18 years admitted between March 15 and June 7, 2020, with laboratory-confirmed SARS-CoV-2 within 48 hours of admission were included. Patients with kidney failure were identified by a combination of kidney failure diagnosis and dialysis procedure International Classificaton of Diseases codes. Patients with previous kidney transplants were not excluded if they had kidney failure at the time of study. The Mount Sinai Institutional Review Board approved this research.

We propensity matched patients with kidney failure to those without kidney failure (1:5) without use of a caliper by age, sex, race/ethnicity, comorbidities (atrial fibrillation, coronary artery disease, cancer, congestive heart failure, diabetes, hypertension, chronic obstructive pulmonary disease, asthma, peripheral vascular disease, stroke, and liver disease), body mass index (kilograms per meter²), admission facility, and admission week using nearest neighbor matching. Despite propensity matching, significant differences in patient characteristics remained between kidney failure and non-kidney failure cohorts. Therefore, we performed logistic regression analysis after controlling for age, diabetes, hypertension, stroke, coronary artery disease, and congestive heart failure to determine the association between kidney failure and mechanical ventilation, intensive care unit (ICU) admission, and mortality. All analyses were performed using R 4.0.0 (R Foundation, Vienna, Austria).

In total, 122 patients with kidney failure were admitted at MSHS with COVID-19 and matched to 610 patients without kidney failure with COVID-19 from a total of 3270 patients without kidney failure. Patients with kidney failure had a higher prevalence of diabetes (55% versus 43%, P=0.02) and hypertension (66% versus 55%, P=0.03). There were no significant differences in admission symptoms between patients with kidney failure and patients without kidney failure. Patients with kidney failure had higher systolic BP, lower heart rates, and higher oxygen saturation at admission. Patients with kidney failure had lower white blood cell counts (5.5 versus 7.1 10³/μl, P<0.001) and higher inflammatory markers of ferritin (3174 versus 1041 ng/ml, P<0.001), erythrocyte sedimentation rate (84 versus 64 mm/h, P=0.001), and procalcitonin (3.4 versus 0.3 ng/ml, P<0.001) (Table 1).

Patients with kidney failure were less likely to receive mechanical ventilation (11% versus 17%, P=0.11) or be admitted to the ICU (9% versus 21%, P=0.003) and had similar in-hospital mortality (9% versus 12%, P=0.31). There was no difference in do not resuscitate/intubate orders (16% in patients with kidney failure versus 17% in patients without kidney failure, P=0.70). After further adjustments, kidney failure status was associated with lower odds of ICU admission (adjusted odds ratio [aOR], 0.37; 95% confidence interval [95% CI], 0.19 to 0.71), with similar direction of association for mechanical ventilation (aOR, 0.6; 95% CI, 0.32 to 1.11) and mortality (aOR, 0.67; 95% CI, 0.33 to 1.38).

Patients with kidney failure have the highest hospital admission rates out of all Medicare beneficiaries. In this propensity-matched study, we found that although patients with kidney failure had higher comorbidity burden and higher inflammatory markers than patients without kidney failure, there were no significant differences in presenting symptoms and vital signs between patients with kidney failure and patients without kidney failure. Additionally, need for intensive care was lower in patients with kidney failure, and there was no difference in mechanical ventilation or in-hospital mortality.

We report lower in-hospital mortality of patients with kidney failure than other institutions (27%–32%) (2(⇓–4). This is likely due to differences in baseline characteristics; patients with kidney failure who were admitted to MSHS had less diabetes and hypertension than other studies. Additionally, our study time had a longer follow-up period (3 months versus 1 month), and patients who were admitted later in the course of the pandemic may have received different treatments than patients who were admitted earlier.

We found no statistically significant difference in in-hospital mortality between patients with kidney failure and propensity-matched patients without kidney failure who were matched on age and comorbidities. Additionally, there has been speculation that uremia-associated immune dysfunction may blunt the COVID-19 cytokine storm; however, this remains to be proven.

Despite propensity matching, patients with kidney failure had more comorbidities and higher inflammatory markers of ferritin and procalcitonin, but they had lower odds of ICU transfer and similar risk for mechanical ventilation. This is in contrast to previous literature that documents the association of kidney failure status with adverse outcomes (5). Patients with kidney failure had a less severe COVID-19 phenotype with lower heart rate, lower white blood cell count, and higher oxygen saturation than patients without kidney failure. Differences in mechanical ventilation and ICU transfer were unlikely to be driven by code status differences given that the proportion of patients with do not resuscitate/do not intubate status was similar.

Limitations include lack of data on the proportion of patients with kidney failure and COVID-19 who died out of hospital. Additionally, not all patients had inflammatory markers. Also, the use of propensity scores has inherent drawbacks, including the inability to adjust for unmeasured variables such as smoking status, which may bias patient selection and subsequent outcomes.

In conclusion, this is the first study to compare in-hospital outcomes for COVID-19 in patients with kidney failure and matched patients without kidney failure. Although patients with kidney failure had higher inflammatory markers and higher comorbidity burden versus patients without, they had lower odds of mechanical ventilation and similar odds of intensive care use and in-hospital mortality.

• Copyright © 2021 by the American Society of Nephrology