Ticket to ride: spreading germs a mile high

Many of us believe air travel is also a lottery ticket for an upper respiratory infection. Breathing recirculated air while strapped into a seat for hours next to a stranger seems the ideal recipe for the transmission of an infectious disease. But to what extent does scientific evidence support these fears? In this issue of The Lancet, Mark Gendreau and Alexandra Mangili describe what we know about a topic of substantial interest to doctors, their patients, family and friends, and to public-health officials everywhere.

The picture that emerges is both reassuring and troubling. An aeroplane cabin provides the smallest volume of available air per person of any public space,¹ but air movement is predominantly transverse, not front to back. Large commercial aircraft typically recirculate about 50% of cabin air, passing it first through high efficiency particulate air (HEPA) filters. It is often said that these filters are effective for viruses,² although supporting data seem insufficient and have been questioned.³ Moreover there is no regulation requiring carriers to use HEPA filters, and only 85% of commercial airliners that carry more than 100 passengers in the current US fleet and recirculate cabin air are equipped with them.⁴ Many fewer of the smaller regional jets that recirculate cabin air use HEPA filters.

Although contagious diseases such as tuberculosis, severe acute respiratory syndrome, measles, and influenza have been transmitted during commercial air travel, published reports of this happening are uncommon. The conventional wisdom is that any risk of contagion is related to the proximity of the index case (plus or minus two rows),⁵ which might be little comfort if you are sitting next to, in front of, or behind someone with a hacking cough or explosive sneeze. But at least in this regard aeroplanes are not much different from other public places or forms of mass transit.

Nevertheless, as severe acute respiratory syndrome aboard Air China Flight 112 showed, we still have much to learn, because cases occurred in passengers at least seven rows in front of and five rows behind the index case (almost the entire length of the coach cabin).⁶ There was spread to as many as 25 passengers.⁷ How often such spread occurs with other infectious diseases is hard to say, because outbreak investigations are inherently difficult. The exposed population disperses widely and becomes sick days or weeks later with non-specific and non-reportable symptoms. Gendreau and Mangili, and others,⁸ appropriately call for intensified study to estimate the risks of disease transmission aboard commercial aircraft, and the effects of mitigating measures such as ventilation.

But more is needed. If a contagious disease were only transmitted locally, it would take a considerable time to reach global dimensions. In 14th century Europe, it took 3 years for bubonic plague to diffuse from southern Italy to Britian on the backs of rats. But if influenza H5N1 makes the transition to person-to-person contagion, it will have a ticket to ride. Our world is now interconnected to an unprecedented, perhaps qualitatively different, extent. From a public-health standpoint, air travel is one of the most important kinds of interconnections. Network theory shows that shortcuts between local enclaves can dramatically shrink the average path-length between individuals, giving rise to the so-called small-world phenomenon in which each of us is related to anyone else on average by, at most, six degrees of separation.9, 10 Consider that airline networks are arranged in a hub and spoke pattern, and that some people such as cabin attendants or frequent travellers come in contact with a vastly larger number of geographically separated people than most of us. We are just now beginning to realise the influence of specific kinds of network topologies on the ease and speed with which disease is spread. Sophisticated analysis might show which nodes to apply our preventive measures to first or most vigorously.⁷ Like many difficult public-health problems, air travel as a disease vector will only yield to an intensive interdisciplinary research effort, bringing together ventilation and infectious disease experts, mathematicians, and others. Unlike many other problems, however, there is an urgency here that begs for immediate attention.

Regulations requiring HEPA filters for any aeroplane that recirculates air should be seriously considered. Advice on what an individual can do remains generic: good personal hygiene to protect yourself (wash hands frequently, particularly before eating), cover nose and mouth when sneezing or coughing, and wash hands afterwards to protect others. Beyond that, might we suggest one of the many magic little rituals the public uses to allay their general anxiety when flying. When Niels Bohr was asked by a reporter if he was superstitious because he had a horseshoe over his laboratory bench, he said of course not. He was a scientist. But he understood it worked even if you did not believe in it.