Someone who lived with a Missouri resident infected with bird flu also became ill on the same day, the Centers for Disease Control and Prevention reported on Friday.

The disclosure raises the possibility that the virus, H5N1, spread from one person to another, experts said, in what would be the first known instance in the United States.

On Friday night, C.D.C. officials said that there was “no epidemiological evidence at this time to support person-to-person transmission of H5N1,” but that additional research was needed.

...

And neither the initial patient nor the household contact had any known exposure to the virus via animals or raw milk.

Neither patient has been identified, and details are scant. The household contact was not tested, so officials cannot be sure that the individual actually was infected with the bird flu virus.

...

Neither the C.D.C. nor Missouri health officials had mentioned the household contact’s illness before the report was posted to the C.D.C. website.

In a news briefing on Thursday, C.D.C. officials said it was unclear how the initially identified patient had become infected and called the case “a one-off.”

On Thursday evening, Missouri health officials said that “all contacts are known and remained asymptomatic during the observation period.”

On Friday, C.D.C. officials acknowledged that the household contact’s illness “should have been mentioned in the press briefing, along with the additional context.” The risk to the public remains low, officials said.

Outside experts were sharply critical of the omission.

“There are absolutely no circumstances in which it is acceptable to not have disclosed that information yesterday,” said Jennifer Nuzzo, director of the Pandemic Center at the Brown University School of Public Health.

Both illnesses might have resulted from simultaneous exposure to H5N1 in some unknown way, the C.D.C. said. But other experts said that the appearance of illness cannot be used to determine timing of exposure because people develop symptoms at different rates.

...

The patient with confirmed bird flu was hospitalized on Aug. 22. The close contact developed symptoms on the same day, but the case was missed by routine flu surveillance because the person was not hospitalized, according to Missouri health officials.

Even now, however, antibody testing could reveal exposure to H5N1. Missouri officials said such tests were “being considered.”

Dr. Nirav Shah, the C.D.C. principal deputy director, said on Thursday that it was too early to detect antibodies in the hospitalized patient. Dr. Bhattacharya disagreed.

“No way, that’s plenty of time to be able to tell,” Dr. Bhattacharya said. “It’s been weeks now, so I don’t see how you wouldn’t even try yet.”

Other experts said they were frustrated with the lack of information about the investigation.

“Aside from getting more details on the case, it would be good to get more information on who was included in case investigation and what criteria were used,” said Dr. Nahid Bhadelia, director of the Center on Emerging Infectious Diseases at Boston University.

In addition to the household contact, a health worker attending to the hospitalized patient became ill but tested negative for the flu, the C.D.C. said on Friday night.

C.D.C. scientists were unable to obtain the full genetic sequence of the virus from the hospitalized person because of the sample’s low quality, they said, so they could not be sure of how the virus might be changing.