ASN's Mission

ASN leads the fight to prevent, treat, and cure kidney diseases throughout the world by educating health professionals and scientists, advancing research and innovation, communicating new knowledge, and advocating for the highest quality care for patients.

learn more

Contact ASN

1401 H St, NW, Ste 900, Washington, DC 20005

email@asn-online.org

202-640-4660

The Latest on Twitter

Kidney Week

Abstract: PO0781

Association Between Kidney Dysfunction at Admission and Outcomes in Hospitalized Patients with COVID-19

Session Information

Category: Coronavirus (COVID-19)

  • 000 Coronavirus (COVID-19)

Authors

  • Chaudhri, Imran, Stony Brook University Hospital, Stony Brook, New York, United States
  • Koraishy, Farrukh M., Stony Brook University Hospital, Stony Brook, New York, United States
  • Bolotova, Olena, Stony Brook University Hospital, Stony Brook, New York, United States
  • Yoo, Jeanwoo, Stony Brook University Hospital, Stony Brook, New York, United States
  • Dhaliwal, Simrat, Stony Brook University Renaissance School of Medicine, Stony Brook, New York, United States
  • Annadi, Raji Reddy, Stony Brook University, Stony Brook, New York, United States
  • Nguyen, Minh Hoai, Stony Brook University, Stony Brook, New York, United States
  • Mallipattu, Sandeep K., Stony Brook University Hospital, Stony Brook, New York, United States
Background

AKI is a major predictor of mortality in patients with coronavirus disease 2019 (COVID-19). Data regarding association of renal dysfunction (AKI, hematuria and proteinuria) at the time of admission with hospital outcomes is limited.

Methods

In this retrospective single-center study, we analyzed electronic medical record data on 300 patients admitted with COVID-19. Data collection included history of comorbidities, medications, vital signs, and admission and peak laboratory values. Outcomes included inflammatory burden (calculated using composite scores for multiple markers of inflammation), AKI during hospitalization, admission to the intensive care unit (ICU), need for invasive mechanical ventilation, mortality and length of stay. For multivariate analyses, generalized linear model (continuous outcomes) and logistic regression (dichotomous outcomes) were used. Machine learning algorithms (XGBoost classifier with 3-fold cross-validation) were performed to develop a predictive model for in-hospital AKI.

Results

No significant associations between admission AKI and hospital outcomes were observed. Admission proteinuria was associated with increases in in-hospital AKI, ICU admission, death, peak inflammation score, and length of stay on descriptive analysis; however, on multivariate analysis (after adjusting for multiple covariates), only in-hospital AKI remained statistically significant (OR=4.71, 1.28–17.38, p=0.02). Admission hematuria was associated with increases in in-hospital AKI, ICU admission, invasive mechanical ventilation, and death on descriptive analysis; and on multivariate analysis it still predicted increased rates of ICU admissions (OR=4.56, 1.12-18.64, p=0.03), invasive mechanical ventilation (OR=8.79, 2.09-37.00, p=0.003), and death (OR=18.03, 2.84-114.57, p=0.002). Using machine learning algorithms, an area under the receiver operating curve (AUROC) of 87.4% with an accuracy of 87.6% was obtained for predicting in-hospital AKI using only admission data.

Conclusion

In patients with COVID-19, admission hematuria and proteinuria are associated with adverse hospital outcomes, and admission data can be used to predict AKI during hospitalization.

Funding

  • NIDDK Support