The ghost of pandemics past

Prominent North American and European politicians convened recently for a table-top exercise designed to illustrate the international scope of bioterrorism.¹ Participants were asked to respond to a hypothetical apocalyptic incident, in which the simultaneous release of lyophilised variola in six countries caused 74 000 cases in a single generation. Not surprisingly, such an attack, irrespective of the agent, defied easy control in the exercise. Issues about the manufacture and distribution of massive amounts of smallpox vaccine dominated the discussion. Three false lessons were conveyed to readers of the ensuing press release, and doubtless to the lay participants: a smallpox pandemic is likely; ordinary methods of public health cannot control person-to-person infections; and vaccine stores are the key to control. Poppycock, poppycock, poppycock!

3 years ago, the Bush policy of vaccination against smallpox on a large scale2, 3 was in part motivated by Dark Winter, a similar table-top exercise, also with questionable assumptions.⁴ Sceptics questioned the benefits of mass vaccination, and raised concerns about vaccine complications.5, 6 Doubts about the net benefits of mass vaccination grew after reports of subsequent cardiovascular events,7, 8 and only a few people have been vaccinated.

In the meantime, a different global pandemic attracted attention. A culinary choice in south China led to a fatal infection in Hong Kong, and subsequently to 8000 cases of severe acute respiratory syndrome (SARS), and nearly 1000 deaths in 30 countries on six continents.9, 10 Despite rapid geographical dissemination and much economic angst, clinicians and epidemiologists proved up to the task, and SARS was controlled in all locations after about 3 months. Control was accomplished by well-informed managers¹¹ using the same basic strategy of surveillance for cases and isolation of contacts that had eradicated smallpox 25 years before,¹⁰ augmented only by high-tech record-keeping and communication.¹²

In fact, SARS presented a challenge far more difficult than smallpox. Like smallpox, SARS spread efficiently at the bedside to family members and caregivers5, 6 by means of large droplets. Like smallpox, the coronavirus was capable of wide dissemination from only a small number of individual cases,¹³ with the help of favourable air currents and contaminated clothing or linens.

However, neither vaccine prophylaxis nor sophisticated diagnostic methods were available to combat SARS. Whereas unvaccinated people with infectious smallpox have a characteristic vesicular rash, which is likely to keep potential contacts at a distance, SARS was disseminated in material or faeces from people with early non-specific symptoms.⁹ Control of smallpox, but not SARS, was facilitated by a 2-week period between exposure and infectiousness in which to identify and isolate contacts.

Had the outbreak been of smallpox instead of SARS, control in Toronto, for example, would have been quickly established by the old-fashioned techniques of public health: surveillance and isolation, identifying the exposed for vaccination, and avoiding mass vaccination to the extent possible politically. On the basis of the early sequence of events in this largely nosocomial outbreak,¹⁴ containment of smallpox probably would have occurred by the appearance of the third disease generation, much sooner than SARS was contained.

SARS will probably reappear, and there is a remote chance that smallpox will be introduced by a terrorist act (but not with 74 000 first-generation cases). An outbreak of either disease would be costly, but experience suggests that control would again be quickly established. Of greater concern are the more virulent strains of influenza. Reconstruction of the DNA of the 1918 Spanish virus is not only possible, but also a reality,¹⁵ and we can only hope it does not escape into the population. Moreover, the number of human cases of avian influenza in Asia has been increasing.¹⁶ The first clear person-to-person transmission of influenza A (H5N1) reported from Thailand¹⁷ was even more limited to family members than smallpox or SARS, despite the H5N1 exposure of other family members and hospital staff. As yet there is no change in the host specificity of the viral receptor's binding preference, which would probably enable more efficient human-to-human transmission, but that might be only a matter of time.¹⁶

Whatever pandemic agent next appears (and smallpox is certainly not the most likely), each community must be prepared. Vaccines will doubtlessly evolve, although avian influenza will require a constant struggle,¹⁸ prototype SARS vaccines now appear strain-specific,¹⁹ and the qualities of any new smallpox vaccine can never be evaluated.²⁰ Even so, it is easy and politically safe to budget for vaccine production, with or without plans for mass vaccination as a pre-emptive strategy or as outbreak control. However, no weapon is effective without an appropriate delivery system. It is harder to budget for field personnel in public health, and justify them to the accountants, than it is to budget for vaccines, even if these people are absolutely essential.

Identification of cases and of contacts eligible for prophylaxis requires active surveillance, and any isolation policy requires a detailed plan. Both actions will be needed to break a chain of person-to-person infection, and both require management skills that are grounded in biology, local sociology, and the medical-care system. History tells us that protection against a viral pandemic will always depend on the availability of a local cadre of skilled public-health workers. Although the return on an investment for the support of such personnel is hard to quantify, we assign it low priority at our peril.