The Centers for Disease Control and Prevention released its first study under the Trump administration on the current H5N1 bird flu outbreak on Thursday, finding that a spillover of the virus among dairy cattle and humans has gone undetected.

The study analyzed blood samples collected from 150 veterinarians throughout America who worked with cattle and found that three of them had H5N1 antibodies, a sign that they had recently been infected with Avian Influenza. Only one of the vets said they knowingly worked with infected live poultry.

“If the circulating H5 viruses become more transmissible between humans, we are not going to be able to control transmission as the viruses will spread rapidly and often subclinically,” an infectious disease epidemiologist with the University of Texas Medical Branch at Galveston, Gregory Gray, said to NPR.

The study was due to be released several weeks ago but was delayed by the Trump administration after it ordered federal health agencies to suspend external communications. Two other studies of bird flu have yet to be made public, according to CNN. 

Last week, a new variant was found in dairy cattle for the first time among six cow herds in Nevada.

The D1.1 variant has been associated with two human infections, including a fatal case in Louisiana, raising concern about the potential of the chance of infections among the population. 

A day later, Governor Hochul ordered all markets selling live poultry to shut down at and around New York City after a spate of bird flu infections were discovered.

A total of seven cases of avian influenza were found during routine inspections of live bird markets at Queens, the Bronx, and Brooklyn. Out of an abundance of caution, the governor said she has ordered all markets in the five boroughs and neighboring Westchester, Nassau, and Suffolk counties to be closed until Friday at the earliest.

“We should never underestimate flu,” a professor at the University of Nebraska Medical Center College of Public Health, Lauren Sauer, said to NPR.  “If cases are occurring more frequently than detected in humans, we risk missing small changes that allow the virus to begin to spread much more easily in humans.”