Hyperglycemia is an independent predictor of in-hospital mortality in critically ill patients with acute kidney injury-a cohort study

Maristela Bohlke, Laura Madeira, Tulio Reichert, Ana Carolina Brochado Geist, Pedro Funari Pereira, Helena Rotta Pereira, Gustavo Tragnago, Paulo Caruso, Franklin Correa Barcellos


Introduction: The association of hyperglycemia with poor outcomes has been described in several settings, including in generalintensive care unit (ICU) patients. However, it is not clear whether this relationship is consistent for all critically ill patients. Ourstudy assessed the association of blood glucose (BG) with in-hospital mortality in critically ill patients with acute kidney injury(AKI).

Methods: A cohort of critical care patients with AKI was followed up until death or hospital discharge. The associationof BG level with in-hospital mortality was analyzed with multivariate logistic regression analysis adjusted for demographic,socioeconomic, laboratory and clinical variables. Receiver-operating characteristics (ROC) analysis was used to assess the abilityof various levels of BG to predict in-hospital mortality.

Results: One hundred patients were followed, with a mean age of 62.2 years, 49 male, 41 surgical, 34 diabetics and 63 withsepsis. Nineteen patients needed renal replacement therapy and 67 died during hospital stay. In the final multivariate model, age,glucose level and sepsis had an independent association with the outcome death. The threshold level of BG that maximized thecombined sensitivity and specificity for the prediction of in-hospital mortality by ROC analysis was 109 mg/dl. In the stratifiedanalysis, BG was an independent predictor of death only among non-diabetic patients.

Conclusions: To the best of our knowledge, this is the first study to describe an association between hyperglycemia and in-hospitalmortality in critically ill patients with AKI. Further studies are needed to confirm this finding and to assess the potential impact oftighter glucose control in this subpopulation.

Full Text:


DOI: https://doi.org/10.5430/jer.v2n1p102


  • There are currently no refbacks.

Journal of Epidemiological Research

ISSN 2377-9306(Print)  ISSN 2377-9330(Online)

Copyright © Sciedu Press

To make sure that you can receive messages from us, please add the 'sciedu.ca' and ‘sciedupress.com’ domains to your e-mail 'safe list'. If you do not receive e-mail in your 'inbox', please check your 'spam' or 'junk' folder.