COVID-19: novel blood test is highly predictive of severe outcomes

  • McElvaney OJ & al.
  • EBioMedicine
  • 08.10.2020

  • von Liz Scherer
  • Clinical Essentials
Der Zugang zum gesamten Inhalt dieser Seite ist nur Angehörigen medizinischer Fachkreise vorbehalten. Der Zugang zum gesamten Inhalt dieser Seite ist nur Angehörigen medizinischer Fachkreise vorbehalten.

Takeaway

  • A novel prognostic tool dubbed the Dublin-Boston score significantly outperforms tracking IL-6 changes alone in determining risk for severe clinical outcomes in COVID-19.
  • The score is based in part on the 4-day change in the IL-6:IL-10 ratio.

Why this matters

  • Longitudinal measures of cytokine balance alterations may be more accurate than absolute IL-6 levels for predicting clinical decline in hospitalized COVID-19 patients.

Key results

  • 80 hospitalized patients; mean (±standard deviation) age, 58 (±17) years; mean symptom duration, 2 (±2) days.
  • Baseline: 88% on oxygen support.
    • 24 invasively ventilated, 19 noninvasively ventilated/on high-flow oxygen, 27 on low-flow cannula.
  • Each 0.1-unit increase in the IL-6:IL-10 ratio across days 0-4 was linked to more severe clinical outcome:
    • OR, 1.28 (P=9.3×10−8).
  • More severe clinical outcomes also seen for each 10-unit IL-6 increase across days 0-4:
    • OR, 1.14 (P=6.2×10 5 ).
  • Using the Dublin-Boston score, severe outcomes risk increased for every 1-point increase:
    • OR, 5.62 (P=1.2×10−9 ).
  • At day 7, the association with clinical outcome was consistent across the patient spectrum, and the ratio slope and 4-day change in ratio were similar when stratified by clinical location.

Study design

  • Prospective cohort analysis.
  • Dublin-Boston score was calculated as follows: days 0-4 change in IL-6:IL-10 ratio multiplied by 2, rounded to a whole number, and restricted to a 5-point scale (−2 to 2); higher score=worse prognosis. 
  • Funding: American Thoracic Society; others.

Limitations

  • Small sample size, no replication cohort.
  • Limited generalizability.
  • Asymptomatic patients not included.