A Mayo Clinic study indicates that many hospitalized influenza patients shed flu virus for at least 7 days after they fall ill, suggesting that the current recommendation to use infection control precautions for 5 days may be inadequate.

In the study, one type of test showed that 22 of 41 patients were still shedding the virus 7 days or longer after the onset of their illness, according to the report in Infection Control and Hospital Epidemiology. Another testing method showed that 12 of the 41 patients were still infective at 7 days or later.

"This finding raises the concern that hospitalized patients who are older and/or have chronic illnesses could shed influenza A virus beyond the 5- to 7-day period that has traditionally been considered to be the time of infectivity," says the report by Surbhi Leekha and colleagues at the Mayo Clinic College of Medicine in Rochester, Minn.

The authors recruited 50 patients who were hospitalized for at least several days for influenza from December 2004 through mid-March 2005. The patients´ median age was 72, and 48 of the 50 had an underlying chronic disease, such as heart or lung disease or diabetes. About two thirds of them had received a flu shot.

Polymerase chain reaction (PCR) was used to test respiratory tract specimens for influenza A. If the initial test was positive, additional throat swabs were taken and tested at 2, 3, 5, and 7 days after the first specimen, if the patient remained in the hospital. The 41 patients used in the final sample included 35 from whom specimens were obtained at least 7 days after symptom onset, plus another six whose samples tested negative earlier than 7 days after onset. (For the other nine patients, researchers could not determine whether shedding persisted at 7 days, so they were excluded.)

PCR testing was positive for influenza in 22 (54%) of the 41 patients at or after 7 days, whereas two culture methods (tube and shell vial) yielded flu virus in 12 (29%) of the 41 patients, the report says. All the patients who tested positive by culture also had positive PCR results. The longest period of flu virus shedding shown by any method was 14 days.

The researchers found no link between prolonged virus shedding and either antiviral treatment or previous vaccination. Those who received antiviral treatment shed the virus an average of about 2 fewer days (5.7 versus 7.5 days) than those who were not treated, but the difference was not statistically significant.

The Centers for Disease Control and Prevention (CDC) recommends "standard" and "droplet" isolation precautions for hospitalized flu patients for 5 days after symptom onset, the report says. This recommendation, the authors say, "could be insufficient for such patients, and prolonging isolation measures for the entire duration of the hospital stay might be more prudent to prevent outbreaks in hospitals during the influenza season."

The researchers cite several limitations to their study that might have swayed the results toward either underestimation or overestimation of how long patients were infective. For example, because other patients and hospital staff were not tested, it is uncertain whether the detection of flu virus by PCR meant that the patients were actually capable of spreading the virus.

Nonetheless, the results "suggest that hospitalized patients who are older and/or have chronic illnesses could shed influenza A virus for several days beyond the period those individuals are traditionally thought to be infective," the article states. "A recommendation for prolonged isolation precautions might be necessary to prevent influenza outbreaks in acute care facilities."

Michael L. Tapper, MD, hospital epidemiologist and director of the Division of Infectious Diseases at Lenox Hill Hospital in New York City, praised the study but said it would be premature to recommend changing the infection control guidelines for hospitalized flu patients without further research.

Citing an author-acknowledged limitation of the study, he told CIDRAP News, "The majority of the findings [of continued shedding] were based on PCR rather than culture, and PCR might have detected nonviable virus. The fact that a PCR assay remained positive at 5 or 7 days, I´m not sure how to interpret that in terms of clinical communicability."

Even for patients who tested positive by culture, "being culture positive doesn´t necessarily imply that you´re infectious," Tapper added. "I think the study nicely shows that virus may be detected for a longer period of time, but whether that translates into a longer period of transmission is not clear."

Noting that the patients in the study were mostly elderly and had underlying chronic conditions, he said, "These are exactly the patients you´d expect to shed for a longer period of time. So it´s hard to generalize from this group to the general population."

"I don´t know that I would jump from anything in this article to a recommendation to change the isolation guidelines in any setting," Tapper said.

He suggested that further studies on the topic should involve larger numbers of patients and should test whether PCR-based findings of virus shedding are associated with infectivity in animal models.