TB on a plane. Checking the evidence.

What do we know about transmission of tuberculosis on an airplane? Not much, apparently. There is very little literature on it and not a single case of active TB has ever been traced to an airplane contact. On the other hand, it isn't very easy to estimate the risks. The only way you can do it is in cases such as the current one where you know someone with TB got on an airplane and potentially exposed others. Then you would do intensive follow-up of fellow passengers and crew to see if you could find others who might have gotten infected. Now the problems really start.

First, there's the task of finding passengers, who disperse immediately upon disembarkation and where airlines don't have contact information post flight, getting them to agree to testing on two occasions two weeks apart and then trying to figure out what the results mean. Skin testing is the usual means to see if the person has ever been infected with the TB organism at any time prior to the test (even decades before), and there are both false negatives and false positives for the test. More than 90% of those infected will not have active TB, that is the signs and symptoms of disease. Their immune defenses will have walled off and neutralized the bug. So there are a lot of people with positive skin tests who aren't sick, aren't contagious and may have been positive for long before the airplane flight. That's why two tests are done, one as soon as possible and before the part of the immune system responsible for the skin reaction has gotten underway, the second 8 to 10 weeks later to see if the person has "converted" in the interval, evidence the infection was recent and not years old. But in many parts of the world (not the US), people are routinely vaccinated with an attenuated TB organism (BCG vaccine), so they show up as positive skin reactors for a different reason.

Thus if you want to estimate the risk of TB from air travel you need an index case and as many passenger contacts as you can find, tested twice several weeks apart with medical histories to help you interpret the results of the test for your purpose. Surprisingly, this has been done several times. In 2006, WHO published the second edition of their Guidelines for the prevention and control of tuberculosis and air travel. I don't usually quote very long passages here, but in this case I'll give you most of Chapter 2 from the Guiedlines so you can see context. I've omitted the cites, but you can find them in the link, above. I've looked at them and don't think their rendition is out of context. Here are the relevant passages:

Research has shown that there is a very small risk of any infectious disease being transmitted on board aircraft [the cite here is to the Mangili and Gendreau review we discussed in our initial post]. However, transmission probably occurs more frequently than reported because most diseases have an incubation period longer than the duration of air travel. Of the airborne and droplet-borne diseases that are potentially transmissible on board aircraft, the most important are TB, infl uenza, meningococcal disease and measles.

TB is an infectious disease, caused in most cases by M. tuberculosis and transmitted by exposure to tubercle bacilli in airborne droplet nuclei produced by a person with infectious TB during expiratory efforts, such as coughing, sneezing or singing. TB develops in the human body in two stages: fi rst, the individual exposed to M. tuberculosis becomes infected; second, the infected individual develops the disease. However, only a small minority of infected individuals will develop the disease. While the subsequent risk of progressive TB is greatest within the fi rst year or two after infection, latent infection may persist for life. The latent period between acquiring TB infection and developing active TB (i.e. in the minority of infected people who go on to develop the disease) varies between 2 weeks and several decades. Primary TB occurs between 4 and 12 weeks after infection, and this can be confi rmed by a demonstrable primary lesion or conversion of the tuberculin skin test. The majority of patients who ultimately develop TB will do so within the fi rst fi ve years after infection. To date, no case of active TB has been identifi ed as a result of exposure on a commercial aircraft. Furthermore, no evidence of TB disease has been reported among those known to have been infected with M. tuberculosis during air travel. From 1992 to 1994, the United States Centers for Disease Control and Prevention (CDC), together with state and local health departments, conducted seven contact investigations, one centred on a cabin crew member and six on passengers with infectious TB who had fl own during this period. Concern was raised that the closed aircraft cabin environment may have enhanced transmission of M. tuberculosis. The number of potentially exposed passengers and cabin crew exceeded 2600 on a total of 191 flights involving nine different types of aircraft.

All index patients were highly infectious, i.e. smears from spontaneous sputum specimens from all index cases were heavily positive for acid-fast bacilli (AFB) and all patients were culture-positive and had evidence of extensive pulmonary disease on chest radiography. One patient also had biopsyand culture-confi rmed laryngeal TB, the most infectious form of TB. In two instances, strains of M. tuberculosis resistant to at least isoniazid and rifampicin were isolated, i.e. multidrug-resistant tuberculosis (MDR-TB). Organisms isolated from the other patients were sensitive to all anti-TB drugs. Two passengers, who were flying to the United States for medical care, knew that they had active TB at the time of their flights but did not inform the airline of their disease. In the other fi ve cases, TB was diagnosed after the flights.

In only two of the investigations was there evidence to suggest transmission of M. tuberculosis infection: one from a cabin crew member to other crew members, and another from a passenger to other passengers. In the fi rst report, evidence of transmission was limited to cabin crew with at least 12 hours' exposure to the infectious source. In the other, transmission of infection occurred to only a few passengers seated in the same section as and in close proximity to the passenger with infectious TB, and only on one fl ight lasting more than eight hours.

These results suggest that the risk of infection with M. tuberculosis during air travel is similar to that associated with other activities in which contact with potentially infectious individuals may occur (e.g. train travel, bus travel, any gathering in enclosed spaces). No case of TB disease has been reported among the infected people in the seven studies carried out by CDC. No other instances of possible TB transmission on aircraft have been published since then.

Other potentially serious airborne or droplet-transmitted infections that could merit public health interventions if encountered during air travel are infl uenza (both seasonal infl uenza and any unusual strain such as avian infl uenza affecting humans), measles, meningococcal disease and SARS. Box 1 presents a short description of these diseases. For further information and advice for travellers, see International travel and health (http://www.who.int/ith). (pp. 5 - 6 from Tuberculosis and air travel: guidelines for prevention and control, second edition, WHO, 2006; cites omitted)

The current case of TB and air travel differs in two important respects. First, the index case was not highly infectious, had no cough or fever, minimal chest x-ray changes in comparison to the reported cases, and has had several negative sputum smears. On the debit side, the infecting strain is much more worrisome because it is resistant to the first and second line drugs. This makes any cases especially dangerous to the infected person. Note, however, that XDR TB is not thought to be either more virulent or more infectious than other strains. Its danger comes from the fact that we can't treat it easily or sometimes at all. This raises the stakes considerably and undoubtedly was the reason for the CDC actions in trying to prevent his travel home on a commercial flight.

Am I reassured by the available evidence? Yes and no. The problem has been looked at and despite considerable contact tracing (2600 contacts on a total of 191 flights involving nine different types of aircraft) there was little evidence of transmission and no evidence of active TB. But the evidence is still scant, follow-up still short, many difficulties in interpreting the data. There was evidence of transmission and the lack of active TB is to be expected because most infections do not produce active disease ("active TB" is infection with clinical signs and symptoms of disease). So I think it is certainly possible that XDR TB can be transmitted this way with a resulting outbreak (a cluster of related cases). So figuring out a procedure for handling this kind of problem is definitely indicated. On the other hand, you won't prevent it. There are too many undiagnosed TB cases in the world. This kind of thing has probably happened multiple times without anyone knowing it. People get TB (14,000 new cases a year in the US) and much of the time we don't know where they got it from. With international travel into areas with high prevalence of TB infection, this kind of event is going to happen. Probably often.

So I'm not reassured. But I am resigned.

Update: Other posts here.