iScience

Using understandable machine learning to predict how well low-dose methylprednisolone works in long COVID

Updated

Abstract

Essence

An interpretable logistic-regression model may help predict which long COVID patients respond to low-dose methylprednisolone.

Evidence

This retrospective three-hospital machine-learning study used 330 treated long COVID patients split into training, test, and two external validation sets, with logistic regression AUCs of 0.8715, 0.7198, 0.8419, and 0.8676.

Caveat

The model predicts observed treatment response in retrospective treated cohorts and does not prove methylprednisolone effectiveness or prospective clinical utility.

Simplified

Full Text

We can’t show the full text here under this license.

what lands in your inbox each week:

  • 📚7 fresh studies
  • 📝plain-language summaries
  • direct links to original studies
  • 🏅top journal indicators
  • 📅weekly delivery
  • 🧘‍♂️always free