The U.K. trial that tested dexamethasone found that only certain hospitalized patients benefited. In this case, it was those who were sick enough to need oxygen or a mechanical ventilator. The drug cut their risk of dying by one-fifth to one-third.
But when hospital patients were not on respiratory support, the drug was no help.
The current study turned up a different line of demarkation: Blood levels of a substance called C-reactive protein (CRP), a marker of inflammation.
If patients' CRP was high (20 mg/dL and up), treatment with steroids cut the risk of death or ventilation by 77%.
But if CRP was low (less than 10 mg/dL), steroid therapy more than doubled those risks of death, the study authors reported.
