Skip Navigation
Skip Navigation

Review of Clinical Avian Influenza A (H5N1) Infection in Humans

By Eric Toner, M.D. and Lucina Borio, M.D., October 3, 2005

On September 29, 2005 the New England Journal of Medicine published a review article on avian influenza A (H5N1) infection in humans by The Writing Committee of the World Health Organization (WHO) Consultation on Human Influenza A/H5. Below are the article’s key points with respect to clinical features, management, and hospital infection control of human infection with influenza A (H5N1). It is important to note that many questions remain and, as more information about this disease is gathered, modifications to current recommendations will likely follow.

Clinical Features

  • The clinical spectrum of disease described is based on hospitalized patients. The overall frequency of mild, atypical, and perhaps asymptomatic cases is not known. Most patients have been young and were previously healthy.

  • Generally, the incubation period has been between 2-5 days, but has ranged up to 8 days after exposure (i.e., longer than other human influenza viruses). 

  • The initial symptoms of H5N1 are not like typical human influenza. High fever, cough, and pulmonary infiltrates have been present in nearly all patients, but upper respiratory symptoms are variable, seen in fewer than 50% of patients. Gastrointestinal symptoms (especially watery diarrhea) have been common and may precede respiratory symptoms. Encephalopathy with diarrhea but no respiratory illness has been described in a limited number of patients.

  • Lower respiratory tract disease develops early in the course of illness, with clinical and radiographic manifestations of pneumonia. Most patients have required assisted ventilation within 48 hours of hospital admission; many patients have progressed to acute respiratory distress syndrome. Multi-organ system failure has been common. 

  • Laboratory findings include leucopenia, thrombocytopenia, and elevated aminotransferase levels. 

  • Mortality among hospitalized patients, mostly from progressive respiratory failure, has been as high as 89% for children younger than 15 years of age.


  • Diagnosis may be confirmed by viral isolation and/or H5-specific RNA. In contrast to human influenza A infection, pharyngeal swabs have been more sensitive than nasal samples. Commercial rapid antigen tests for influenza A have been relatively insensitive (36%) and do not discern between the different viral subtypes.


  • Early treatment with neuraminidase inhibitors should not be delayed pending definitive diagnosis, given the potential benefit of this class of drugs if given early in the disease course. The optimal dose and duration of therapy is not known, but recent animal studies employing oseltamivir suggest that a higher dose for a longer period than approved for human influenza may be necessary. For adults with severe infections, the authors suggest doubling the approved dose of oseltamivir and duration of treatment. 

  • Oseltamivir-resistant viruses were detected in several patients who had been treated with oseltamivir, but these viruses remained sensitive to zanamivir and partially sensitive to the investigational agent peramivir in vitro.

  • Corticosteroids have been widely used without any clear benefit. This may be in part because it has been instituted late in the disease course.

Hospital Infection Control

  • Transmission of human influenza A occurs by inhalation of infectious droplets and droplet nuclei, by direct contact, and by indirect (fomite) contact followed by self-inoculation of the respiratory tract and/or conjunctival mucosa. For human influenza H5N1 infection, transmission has occurred by way of bird-to-human transmission, possibly environment-to-human, and limited human-to-human transmission.

  • In healthcare facilities, combined standard, droplet, contact, and airborne isolation precautions (N-95 masks and negative pressure isolation) should be employed. Because viral shedding may be protracted, especially in children, patients should be isolated for at least 7 days after the resolution of fever or possibly up to 21 days. 

  • The number of healthcare workers (HCW) having contact with H5N1-infected patients should be limited and these HCW should not care for noninfected patients. They should also be monitored for the development of fever. 

  • Osletamivir prophylaxis should be considered for HCW who develop unexplained fever; who may have had unprotected exposure to aerosols, droplets, or fluids; or who participate in high-risk aerosol generating procedures.

Reference: N Engl J Med 2005;353:1374-85.


BioWatch Sensors in D.C. Detect Bacteria on National Mall

By Lucina Borio, M.D. and Eric Toner, M.D.

A number of BioWatch air sensors detected the presence of bacteria related to Francisella tularensis in Washington, D.C. on 9/24/05. Thousands of people were on the Mall for antiwar demonstrations and the National Book Festival, and public health officials were alerted. Because people from all over the country had been potentially exposed on the Mall, the CDC released a nationwide HAN alert on Friday evening.

Tularemia has an incubation period that normally ranges between 1-6 days. More than a week has elapsed since the detection and to date there are no reports of related illness. (Key Questions to Help Identify the First High-Risk Patients with Tularemia have been posted at

F. tularensis has been detected by BioWatch air sensors several times in the past in a number of locations. The first detection occurred in Houston in 2003, where assays tested positive for low levels of the organism in multiple locations. There were no tularemia cases reported after the incident. The bacteria were ultimately believed to be present in the environment, but the levels observed were of no known consequence to human health.

Like the Houston incident, the sensors in D.C. detected very low concentrations of bacteria. The most likely cause of the alarm was an unusual airborne presence of naturally occurring environmental bacteria related to F. tularensis; however, an intentional release cannot be completely ruled out at this time. Public health and law enforcement officials continue to investigate. It appears to have presented no threat to public health.

For further CDC information: