At least one hospital is doing it. But those are by no means is easy to do.

Some Hope: How blood from coronavirus survivors might save lives

https://www.nature.com/articles/d41586-020-00895-8

How blood from coronavirus survivors might save lives

New York City researchers hope antibody-rich plasma can keep people out of intensive care.

Hospitals in New York City are gearing up to use the blood of people who have recovered from COVID-19 as a possible antidote for the disease. Researchers hope that the century-old approach of infusing patients with the antibody-laden blood of those who have survived an infection will help the metropolis — now the US epicentre of the outbreak — to avoid the fate of Italy, where intensive-care units (ICUs) are so crowded that doctors have turned away patients who need ventilators to breathe.

The efforts follow studies in China that attempted the measure with plasma — the fraction of blood that contains antibodies, but not red blood cells — from people who had recovered from COVID-19. But these studies have reported only preliminary results so far. The convalescent-plasma approach has also seen modest success during past severe acute respiratory syndrome (SARS) and Ebola outbreaks — but US researchers are hoping to increase the value of the treatment by selecting donor blood that is packed with antibodies and giving it to the patients who are most likely to benefit.

A key advantage to convalescent plasma is that it’s available immediately, whereas drugs and vaccines take months or years to develop. Infusing blood in this way seems to be relatively safe, provided that it is screened for viruses and other infectious agents. Scientists who have led the charge to use plasma want to deploy it now as a stopgap measure, to keep serious infections at bay and hospitals afloat as a tsunami of cases comes crashing their way.

snip

Arturo Casadevall, an immunologist at Johns Hopkins University in Baltimore, Maryland, has been fighting to use blood as a COVID-19 treatment since late January, as the disease spread to other countries and no surefire therapy was in sight. Scientists refer to this measure as ‘passive antibody therapy’ because a person receives external antibodies, rather than generating an immune response themselves, as they would following a vaccination.

the extent of immune-response.

This is a very labor and resource-intensive (gotta have recovered patient pool after all) procedure which seems like a "no-brainer" and can certainly help in some situations with some infections. However, it is dependant on levels of protective antibody produced. It provides passive antibody that can be very short lived, but might give the patient a short time assist in producing their own antibodies.

and no antibodies? I don't understand. If they recovered from Covid 19 then their plasma should be safe.

response and others only a short term response--e.g.,the difference between post-infection response to hepatitis A (as much as 30-year immunity) vs. influenza (a year or less typically). Not to mention that early on it was not clear that significant mutation with the virus was not going on, making long term natural immunity unlikely.

That is why it is necessary to have some of the info we have only recently started to acquire on immune response both from humans and macaque monkeys infected in the lab.

And no, even if a recovered patient no longer tests positive for the presence of the viral nucleic acid on PCR (antigen tests) does not necessarily mean they have a robust and long-lasting antibody response which would be necessary in order for their plasma to benefit others.

I can see that it may not help in all cases, but what is the harm other than wasting resources.

even if a person has tested antigen negative. Very low risk, but that is why the timing of the "recovery" testing and at least two negative tests is so important.

But, while hospitals ARE trying plasma please do not underestimate the issue of a possible "wasted resource" and more, the difficulties in producing such COVID-specific protective plasma on any kind of widespread level. That is why it is unlikely to be a practical solution for therapy, except in limited circumstances.

If you read a little bit on passive antibody versus active antibody response, I think the limitations of this as "the answer" will be more obvious.

... something that should only be used when the alternatives are worse.

Years ago I had a job where one or two of our hemophilia patients were dying every week from AIDS. They'd "acquired" it from bad Factor VIII preparations made from donated blood.

Blood or blood product transplants are as dangerous as any other sort of transplant. There's always some small chance that something awful, worse than Covid 19, will make it through the process, or that some sort of negative immunological consequence will occur.