COVID-19 research briefs: prediction models for COVID-19 have important limitations

Daily POEMs

Published: 2020-04-26 © 2020 John Wiley & Sons, Inc.

Printer Friendly

Clinical question
What are the key predictors of a poor prognosis for patients with COVID-19?

Bottom line
Very limited evidence suggests that predictors of a severe prognosis in patients with COVID-19 include age, sex, features derived from computed tomography scans, C-reactive protein level, lactic dehydrogenase level, and lymphocyte count. Further details regarding specific demographics were not reported. (LOE = 2c)

Reference
Wynants L, Van Calster B, Bonten MM, et al. Prediction models for diagnosis and prognosis of COVID-19 infection: systematic review and critical appraisal. BMJ 2020 Apr 7;369:m1328. doi: 10.1136/bmj.m1328.

Study design: Systematic review

Setting: Uncertain

Synopsis
Research Brief #7 (Wynants, 2020) In this systematic review of studies that reported diagnostic and prognostic prediction models, the authors identified 27 studies that reported 31 prediction models. Only one study took place outside of China and most were at high risk of bias. Nonetheless, the most frequently reported predictors of the presence of COVID-19 included age, body temperature, and signs and symptoms. The most reported predictors of severe prognosis (deterioration to severe or critical disease, prolonged hospitalization, or death) in patients with COVID-19 included age, sex, computed tomography findings, C-reactive protein level, lactic dehydrogenase level, and lymphocyte count. Given the high risk of bias, estimates of model accuracy are unreliable.

Henry C. Barry, MD, MS
Professor
Michigan State University
East Lansing, MI

Copyright © 2020 John Wiley & Sons, Inc.