公衆衛生：米国におけるサイクロンによる超過死亡数の定量化

The average tropical cyclone in the USA generates 7,000 to 11,000 excess deaths, and these cyclones accounted for around 3.2–5.1% of deaths on the Atlantic coast from 1930 to 2015, according to a new study published in Nature. Increased excess deaths may persist for around 15 years after the event. These findings illustrate that the death toll of a tropical cyclone may be a broader public health issue than previously thought.

Tropical cyclones are a regular threat on the coastlines of the contiguous USA and cause damage to infrastructure, population relocation, societal disruptions and death. Previous analyses of the effect of natural disasters on populations have primarily focused on direct deaths, for which the disaster is the most immediate danger or cause (such as drowning in flood waters). However, researchers have hypothesized that solely tabulating direct deaths may discount the true footprint of a natural disaster such as a tropical cyclone, as disasters frequently trigger a domino effect of other threats to affected populations (such as reduced access to healthcare or increased living expenses).

Solomon Hsiang and Rachel Young modelled the effect of 501 tropical cyclones that directly affected US coastlines between 1930 and 2015. Federal government statistics report that there was an average of 24 direct deaths following these storms — but when accounting for cascading effects, the researchers found that around 7,000 to 11,000 indirect deaths were triggered by tropical cyclones.

In an analysis of data from the Centers for Disease Control and Prevention, Hsiang and Young found that these excess deaths were from a generalized category that includes diabetes, suicide, sudden infant death syndrome or another cause that was not recorded. Cardiovascular disease was the next most common cause, followed by cancer. While the analysis identifies that tropical cyclones are associated with excess mortality, it does not identify the underlying mechanisms. However, the authors propose five factors that might explain the association. For example, economic disruption from job losses or payments needed for post-cyclone repairs might reduce future spending on healthcare. Government responses to support recovery might reduce spending on healthcare infrastructure. Changes to social networks, the natural environment and stress are also put forward as contributing factors.

The authors note that a key limitation of their study is that they only modelled wind speeds in their analysis, and not other characteristics of tropical cyclones such as storm surges, rainfall or flooding. They conclude that these findings offer insight into the true magnitude of the healthcare needs of populations affected by tropical cyclones, as members of those populations may not be aware of just how much their health was affected by the storms.