Risks of Ebola

Like everyone else, we at Chatham have been gripped by the news of this year’s Ebola epidemic. We offer our sincere condolences for those who are bereaved of loved ones, compassion for those who are currently afflicted, and concern for how to stop this terrible virus from spreading further.

A crisis always displays heartfelt motives in stark relief, and this one is no different. We’ve seen health workers and missionaries go directly into the heart of this dread disease’s path, in some cases giving their very lives to heal the sick and halt Ebola’s spread. How admirable has been their courage in the face of peril, and how noble their sacrifice!

Crises like this also require grappling with very tough questions. What can be done to prevent the disease’s further spread, and what options are on the table? Also, how bad could it become? In other words, what are the realistic worst-case scenarios for which to be contemplated and prepared for? With human lives on the line, the risks could not be higher. So what’s the right way to think about them?

To be clear, we don’t equate a facility for managing financial risks appropriately with an ability to plan for and manage epidemiological risks. However, there are areas of overlap – related to statistical methods, current versus future risks, or worst-case risk managements – that touch every risk category. Here are a few:

Risk assessments depend upon a number of assumptions, so it’s deceptive to rely on a single number rather than a range. The CDC estimated that in a worst-case scenario, the total number of Ebola cases in Liberia could be as bad as 1.4 million by January 2015. Fixation upon this single (admittedly frightening!) number without understanding the assumptions that underpin it can paint a significantly bleaker picture than may really exist. For instance, this estimate depends upon the continued spread of the disease without successful intervention – i.e. without effective contact tracing, case isolation, and safe burial. As the international community is able to help fund appropriate intervention, like hospital beds and treatment centers, the estimates shrink in commensurate fashion. If 70% of patients can be treated in confined settings and the dead buried safely, fewer than 30 thousand cases are projected in the same study.

Not all risks scale in the same fashion, so it can be misleading to lump them together. There are plenty of articles being forwarded around about the far greater risk of dying from plane crashes, lightning strikes, or shark attacks then getting Ebola in the U.S. at the moment. To be sure, these are accurate statistics about the present chances of being harmed, but they say nothing about how these risks could change over time. This is because shark attacks and plane crashes can’t increase one thousand-fold in one year – the numbers of great white sharks or airplane flights don’t grow exponentially from year to year – but infected populations can grow virally. This means the current risk of contracting Ebola in the U.S. is orders of magnitude lower than the risk of dying from a shark attack, but in six months the opposite could be true.

Not all upper or lower limits are ironclad. The New York patient with Ebola passed through the enhanced screening at JFK completely asymptomatic. Given the incubation period of the Ebola virus cited by the WHO of 2-21 days, it’s clear that even the enhanced screening will not be completely effective in keeping out Ebola. In fact, even the WHO’s upper boundary of 21 days is based on a 95% confidence interval, not a hard absolute limit. Their New England Journal of Medicine-published piece shows cases from Liberia and Sierra Leone where the time between disease exposure and onset was greater than 30 days. It’s important to know which numbers represent statistical two- standard-deviation boundaries, and which represent absolute upper limits that could never be crossed.

Once risks materialize, the best time to intervene decisively is now. Experts are still debating on whether or not we could see a harrowing increase in the number of Ebola infections. Several professors published an analysis in Lancet Infectious Diseases on Friday estimating that every infected person in Montserrado County, Liberia is infecting 2.49 others on average, and that without intervention there could be as many as 170,000 cases by December 15th. Others disagreed, stating that the best available recent data shows the number of new cases in the area has leveled off or even begun to decrease. But all of them agree that intervening quickly – through adding thousands of beds to treatment centers, speeding up detection, and getting protective kits (bleach, gloves, and masks) to homes in the area – is what the region, and the world, desperately needs.