A small study of patients who were severely ill from the coronavirus hints that treatment with antibodies from recovered patients may modestly help recovery and survival, scientists reported on Friday.
The study, although far from conclusive, is said to be the largest of subjects recovering from Covid-19, the illness caused by the coronavirus. Thirty-nine hospitalized patients were given intravenous infusions of antibodies from patients who had recovered from the condition.
The course of illness in patients who received the convalescent plasma was compared to that of similar patients identified through electronic health records who did not get the treatment.
This is a weak form of comparison, prone to error. And researchers are wary of studies that take place at a single institution, because the results often are not applicable to patients elsewhere.
Researchers at the Icahn School of Medicine at Mount Sinai in New York reported that 18 percent of those who got the plasma of convalescent serum became sicker, compared with 24.3 percent of the patients identified through medical records.
The death rates were 12.8 percent among those who got the antibodies, compared with 24.4 percent among the patients who did not get this treatment. The researchers provided percentages but not actual numbers of patients.
Analyses like these are fraught with difficulties. The only way to know for sure if the treatment works is to randomly assign patients to receive antibodies or a placebo.
And it can be impossible to find many patients who agree to have their treatment randomized to an unknown treatment, noted Dr. Arturo Casadevall of the Johns Hopkins School of Medicine in Baltimore.
The Food and Drug Administration has already authorized the use of convalescent serum in very sick patients. Few hospitalized patients may want to forgo a treatment that appears to be safe and might be effective.
“That train has left the station,” Dr. Casadevall said.
Another way to assess convalescent plasma, favored by researchers elsewhere, is a randomized trial to see if antibodies from recovered patients can prevent infections in health care workers who are at high risk of being infected.
The idea is to have a multicenter trial, to avoid the distortions that can arise when all the data come from one medical center. Researchers not associated with the study then would assess the efficacy of the treatment.