In 2001-02, reduced air travel delayed flu season

Sep 12, 2006 (CIDRAP News) – Researchers reported today that reduced air travel in the wake of the Sep 11, 2001, terrorist attacks delayed and prolonged that year's influenza season.

The study, published this week in the online journal Public Library of Science–Medicine, provides what the authors call the first empirical evidence for an effect of air travel on the spread of flu. As such, the findings may elevate the role of travel restrictions in pandemic planning.

Using 9 years of weekly influenza and pneumonia mortality data (1996 to 2005) from the Centers for Disease Control and Prevention (CDC), researchers from Harvard University and the Massachusetts Institute of Technology in Boston measured the rate of influenza spread across nine regions. Then they used government estimates of air travel volume to assess whether it was related to the timing of yearly flu outbreaks.

The researchers found that the national peak date for flu mortality was within 2 days of Feb 17 in the first 5 seasons they examined (1996-97 to 2000-01). But in 2001-02, the peak date was Mar 2, 13 days later than the previous average. In the ensuing 3 seasons, the peak date progressively returned to the Feb 17 baseline.

For comparison, the authors analyzed flu and travel data for France, where air travel was not restricted after the terrorist attacks. The investigators found no evidence of a delay in France's flu season that year.

The researchers also reported an overall finding that the seasonal flu spread more slowly when the number of domestic air travelers was low, particularly in November.

"Our results suggest that for a nonpandemic year, travel during the Thanksgiving holiday may be central to the yearly national spread of influenza in the United States," they write.

Also, the team found that the flu season peaked later when the number of international air travelers was lower, particularly in September.

The authors say their findings suggest that airline passenger volume explains about 60% of the year-to-year variation in the spread of flu across the country. They say other factors remain unknown; they were unable to link the timing of seasonal flu deaths to the weather.

The spread of highly pathogenic H5N1 influenza has intensified debate about whether border controls and travel restrictions can slow a pandemic, the article notes. "Our results suggest that limiting domestic airline volume may have a measurable impact on the rate of spread of an influenza pandemic, and particularly on spread across regions," the investigators write, adding that the benefits would hinge on early detection and immediate intervention.

Brownstein JS, Wolfe CJ, Mandl KD. Empirical evidence for the effect of airline travel on inter-regional influenza spread in the United States. PLoS Med 2006 Oct;3(10) [Abstract]