NEW YORK (NYTIMES) - Can blood from coronavirus survivors help other people fight the illness? Doctors in New York will soon be testing the idea in hospitalised patients who are seriously ill.

Blood from people who have recovered can be a rich source of antibodies, proteins made by the immune system to attack the virus. The part of the blood that contains antibodies, so-called convalescent plasma, has been used for decades to treat infectious diseases, including Ebola and influenza.

"It's kind of difficult scientifically to know how valuable it is in any disease until you try," said Dr David Reich, president and chief operating officer of Mount Sinai Hospital, which will be using the treatment. "It's not exactly a shot in the dark, but it's not tried and true."

Dr Reich said it would be tried as a treatment for hospitalised patients who had a moderate form of the disease and had trouble breathing but not for those who are in advanced stages of the disease.

"The idea is to get to the right patients at the right time," he said. "But it's experimental."

Researchers at Mount Sinai were among the first in the United States to develop a test that can detect antibodies in recovering patients, an essential part of this treatment strategy.

On Tuesday (March 24), the Food and Drug Administration (FDA) gave permission for the plasma to be used experimentally on an emergency basis to treat coronavirus patients, and hospitals in New York quickly began asking to participate, said Dr Bruce Sachais, chief medical officer of the New York Blood Centre, which will collect, test and distribute the plasma.

"Our main focus is, how do we implement this quickly to help the hospitals get product to their patients?" Dr Sachais said. "We have blood centres in New England, Delaware and the Midwest, so we can do the same thing in other regions. We're working with other blood centres and hospitals that may collect their own blood and want to do this. We may not be able to collect enough plasma in New York to help the entire country, so we want to share with other centres to help them."

Dr Reich said that an email asking Mount Sinai staff members who had recovered to consider donating plasma went "a little viral" and quickly drew 2,000 responses.

But volunteers will have to be carefully screened to meet strict criteria. The donors will include people who tested positive for the virus when they were ill, recovered, have had no symptoms for 14 days, now test negative and have high levels, also called titers, of antibodies that fight the virus.

Dr Reich said that because there were delays and shortages in testing, the number of people who qualify may be low at first.

"Our expectation, based on reports from the Chinese experience, is that most people who get better have high-titer antibodies," Dr Sachais said. "Most patients who recover will have good antibodies in a month."

People who qualify will then be sent to blood centres to donate plasma. The procedure, called apheresis, is similar to giving blood, except that the blood drawn from the patient is run through a machine to extract the plasma, and the red and white cells are then returned to the donor. Needles go into both arms: Blood flows out of one arm, passes through the machine and goes back into the other arm. The process usually takes 60 to 90 minutes and can yield enough plasma to treat three patients, Dr Sachais said.

People who have recovered have antibodies to spare, and removing some will not endanger the donors or diminish their own resistance to the virus, Dr Sachais said.

"We may get rid of 20 per cent of their antibodies, and a couple days later they'll be back."

The plasma will be tested to make sure it is not carrying infections like hepatitis or HIV or certain proteins that could set off immune reactions in the recipient. If it passes the tests, it can then be frozen or used right away.

Each patient to be treated will receive one unit, about a cup, which will be dripped in like a blood transfusion. As with blood transfusions, plasma donors and recipients must have matching types, but the rules are not the same as those for transfusions.

He said the decision to try this approach was based in part on reports from China that it seemed to help patients. But the reports are not based on controlled studies or definitive data.

Dr Sachais said an article in a journal that was not peer-reviewed described treating 10 patients in China with one unit each of convalescent plasma and said it appeared safe and seemed to quickly lower their virus levels.

"It's anecdotal," he said.

A researcher not associated with the new treatment plans said there was evidence to support using plasma from survivors.

"Four to six or eight weeks after infection, their blood should be full of antibodies that will neutralize the virus, and that will theoretically limit the infection," said Dr Vineet Menachery, a virologist at the University of Texas Medical Branch.

In studies in mice, he said, "if you can drive the virus replication down tenfold to hundredfold, that can be the difference between life and death."

He described the use of convalescent plasma as "a classic approach that is a really effective way to treat" - if there are enough donors with enough of the right antibodies. A potential risk, he said, is that the patient's immune system could react against something in the plasma and cause additional illness.

Although hospitals will gather information about the patients being treated, the procedure is not being done as part of a clinical trial. There will not be a placebo group or the other measures needed to determine whether a treatment works.

"People are so desperately ill now, it isn't the right time," Dr Reich said. "They're in the hospital, they're sick, in intensive care, on ventilators. Some get sick so quickly, and it's such a severe illness in some people, we feel it's not the right moment."