Accuracy of Point-of-Care Blood Glucose Level Measurements in Critically Ill Patients with Sepsis Receiving High-Dose Intravenous Vitamin C.

Document Type


Publication Date



Pathology, Critical Care Medicine, Pharmacy, Medical Education

Journal Title


MeSH Headings

Administration, Intravenous, Antioxidants, Ascorbic Acid, Blood Glucose, Critical Illness, Female, Humans, Intensive Care Units, Male, Middle Aged, Pilot Projects, Point-of-Care Systems, Prospective Studies, Reference Standards, Reproducibility of Results, Sepsis


STUDY OBJECTIVE: High-dose intravenous vitamin C is a potential treatment option for patients with sepsis and may interfere with point-of-care (POC) blood glucose (BG) testing. This study aimed to determine if vitamin C dosing used for sepsis affected POC BG level results.

DESIGN: Prospective observational pilot study.

SETTING: Intensive care unit in a large academic tertiary care medical center.

PATIENTS: Five consecutive critically ill adults hospitalized between April 1 and June 1, 2017, who received two or more doses of intravenous vitamin C 1500 mg for the treatment of sepsis and had at least two paired POC BG levels and laboratory venous BG levels measured within 1 hour of each other during vitamin C therapy.

MEASUREMENTS AND MAIN RESULTS: The performance of POC BG level measurement was compared with the reference method of laboratory BG level measurement. The concordance to minimum accuracy criteria for BG meters set forth by the International Organization for Standardization (ISO) 15197:2013, the measurement of agreement between POC BG level and laboratory BG level using the Bland-Altman method, and the clinical accuracy through Parkes error grid analysis were assessed. A total of 16 paired POC and laboratory BG level measurements from the five patients were included. The accuracy of POC BG with laboratory BG level measurements during vitamin C administration according to ISO 15197:2013 criteria was 81.3%, which did not meet the minimum accuracy criteria of 95%. The Bland-Altman analysis showed a mean difference between POC and laboratory BG levels of 8.9 mg/dl, and the Parkes error grid analysis showed that the differences between POC and laboratory BG level measurements would not have resulted in a change in clinical action.

CONCLUSION: The accuracy and agreement of POC and laboratory BG level measurements in critically ill patients receiving vitamin C were consistent with previously published reports in critically ill patients not receiving vitamin C and did not demonstrate clinically significant interference due to vitamin C dosing for sepsis.



First Page


Last Page