Danger of a pandemic

The following post is a part of a series that discusses 'managing risk for development,' the theme of the World Bank’s upcoming World Development Report 2014.

On February 15, 2013, an asteroid 45 meters across sailed past the Earth at 4.9 miles a second.  This was the closest encounter on record with an asteroid this big. Such rare events trigger fear because people overestimate the risk of unusual events – at least for a while. The odds of other rare events are often underestimated. People have a hard time understanding frequencies that are longer than a human lifetime; politicians discount probabilities of disasters that are unlikely to hit while they are in office and so they underinvest in prevention. In sum, we have trouble assessing low-probability, high-impact risks – the kind of events dubbed as Black Swan by Nassim Taleb. 

Responding to concerns about the asteroid, The Economist (Danger of death! Feb. 14, 2013) created a graphic to illustrate how we are unlikely to die from asteroid impact (odds of one in 75,000,000). The chart showed that more prosaic, but still rare, dangers were worse.  For instance, 27 people died in 2008 in America from contact with dogs (a one in 11,000,000 chance of death).  The ranking also showed the odds of death in any given year from a range of causes, such as heart disease, choking, falling down stairs, cycling, and bee stings.

We have two recent reminders of the ever-present pandemic threat. A novel coronavirus probably originated in bats in the Middle East and then infected goats and camels, which in turn passed the virus to humans; the virus is now known to spread from person to person. Since the end of March, the H7N9 avian flu infected more than 120 people in China, with a 20 percent fatality rate. This is how contagion can start and, if unchecked, become a pandemic. Because these are respiratory diseases, their spread could be rapid, and no country would be safe. These pathogens will not necessarily cause pandemics. But with a non-zero probability, either of them could do so and at short notice. Other pathogens have a similar potential.

This brings up the question: do we know the odds of dying in a pandemic? As we decide how much to fear different causes of premature death, we ought to take preventative measures accordingly. It is therefore important to know whether a pandemic is more like an asteroid or like severe storms (odds of death in any given year of 1:2,300,000), or like alcohol (odds of 1:151,000), or like fire (odds of 1:105,000), or like gun violence (odds of 1:25,000), or like all accidents and injuries combined (odds of 1:1,700). Where do flu pandemics rank in The Economist’s graphic?  And are they as lethal as road traffic accidents globally (odds of 1:5,400)?

The worst of the four flu pandemics in the last 100 years killed up to 100 million people in 1918-19, when world population was less than 2 billion. The toll was 3-6 percent of populations in many countries. Although growing antibiotic resistance hampers treatment of flu complications like pneumonia, suppose optimistically that the next bad flu pandemic kills "only" 1 percent of the population. Next, suppose that the probability of occurrence of such a pandemic is 1 percent in any year. Since flu pandemics have recurred every 30-40 years on average, assuming that a severe pandemic is a “once-in-100-years” event is not unduly alarmist.

Thus, the expected odds of death in any year from a severe flu pandemic are 1:10,000—about ten times more than the danger from fire or choking and about the same as global odds of death in a road traffic accident in two years.  In a poor country, with vulnerable populations and weak public health systems, the mortality rate in the same pandemic could be 3 percent. In that case, for many of the world’s citizens, the annual odds of dying in a pandemic are more like 1:3,300, or on par with the odds of dying from all accidents and injuries in hlaf a year in the United States. 

What about our children? Assuming that the mortality rate does not vary with age, an older person with 10 years left to live faces odds of 1:1,000.  But Maya, a child with 70 years of life ahead of her, can expect to experience a severe pandemic with 70 percent probability during her lifetime.  So, she might face lifetime odds of dying in a pandemic of 1:140 in a wealthy country and 1:40 in a poor country.

What’s the point? A pandemic is far more dangerous than an asteroid from outer space and resembles instead the pedestrian risks we face. This is why inadequate management of pandemic risk is such a concern, especially for the young and for future generations. Sure, medical science has made progress since 1918, so many people are lulled into slumber by a vague belief that the probabilities cited above are changing in our favor. Sadly, this is not the case. Vaccines and antivirals may help, but only much too late, after a pandemic has emerged—they will not prevent a pandemic. They will, at best, mitigate the impact by protecting the privileged few who will be first in line to get them.

Death will be only a part of the impact.  Billions would suffer from illness and survive. Nearly all, ill or healthy, could face chaos and the effects of a global economic recession caused by people’s reaction to an invisible threat, high labor absenteeism, plummeting consumer and business confidence, and cascading service failures.  Imagine megacities lacking food deliveries, transport, and police. Pandemics are not mainly a health problem because they can disrupt the functioning of economies and societies. Reducing pandemic risk is a low priority in the health sector, which, moreover, cannot address the externalities involved. Non-medical measures and whole-of-society plans for business continuity are required to reduce the potentially catastrophic impacts on health, economies, and society, but such preparedness for complex emergencies is rare.

The policy implications are straightforward if progress is to be sustained toward poverty reduction and shared prosperity goals.  Pandemics should be prevented early at the source, by the veterinary and human public health authorities. Since contagion could start anywhere, all countries have to be able to detect and control disease in livestock and in humans before it spreads.  Adequate funding and inter-governmental cooperation should strengthen the weakest veterinary and human public health systems, within a framework of internationally agreed standards and governance.  Spending on pandemic prevention is now only about one-eighth of the $3.4 billion required annually in developing countries. The expected benefits of this investment are at least $37 billion annually. When early control fails and contagion spreads, then it will pay to have response plans for business continuity in hand.

Present and future generations would benefit if the international community set a goal to reduce pandemic risk. The goal would empower international organizations to raise risk awareness, and motivate prevention and preparedness; provide relevant knowledge, capacity building, and technical assistance to developing countries; assess the performance of national veterinary and human public health systems and their collaboration; and mobilize resources for bringing the systems to a minimum standard required for global public service delivery.