The idea of "excess mortality" first came to my attention during the Great Panic. Undoubtedly, a great many people died during that time. There have been disagreements voiced in academia regarding whether some people died who had contracted SARS-Cov-2 or whether there was proof that SARS-CoV-2 killed those people. The distinction is perhaps only of academic interest. For the person who dies and their loved ones, the cause is not as important as the outcome itself. And yet, causation was a topic.
Some of the discussion revolved around "excess mortality." Over the years, we have learned that death rates can be predicted with relative accuracy. That study in the field of actuary science has led to a viable and profitable prediction business called life insurance. I tell my students that when you purchase this produce you are literally betting you will die. The seller is conversely betting you will live. The same is true for any indemnification purchase. The underwriter bets you won't flood, get sick, wreck your car, etc. Much like the "house" in Vegas, the actuaries have better and more complete information in that gamble.
Despite their sound actuarial predictions, it is possible for the expectations in any particular sense to defy accurate forecasts. When that happens regarding life expectancy, the market experiences a mortality rate that exceeds (excess) or is less (reduced). The predictability is important in a variety of perspectives, both economic and health care. When it is wrong in either direction, the mathematicians become curious and there is a search for causes. That search is intended to build knowledge and therefore make better bets in the next attempts.
This certainly or predictability is of interest in the realm of public health. People passing away is a reality of life. But, volumes occurring that are unexpected or unpredicted casts challenges that are worthy of consideration and study. Understanding causation might be of benefit in terms of resource deployment, preventative medicine, education, and more.
Nature published an article in early October 2024 regarding a recent study of mortality. It addressed deaths from natural disasters with a particular emphasis on tropical cyclones, without distinction between the general storms and those that reach hurricane designation. It concluded that there are clearly related deaths from these, "directly observed," which it labels "immediate deaths." These are seemingly simple to connect to a storm and to accept in terms of causality.
However, the study reported there concludes that there are "delayed downstream outcomes" that are "indirectly caused by the disaster." These, it concludes, are attributable to cyclones even though they "are difficult to trace back to the initial event." The study scientists conclude that there are "lasting economic impacts" tied to storms and those may contribute to the long-term "full health impact(s)" of such events.
The reader will likely recall I have encountered many cyclones. My life has been directly impacted by Ivan, Dennis, and Sally. I have engaged in recovery following Katrina and Zeta. I have sat here watching many others form, threaten, and yet divert to unleash horrors elsewhere. Some discount that latter experience, but there is an emotional toll from each storm, even when it hits elsewhere. For me, Irma, Michael, and Katrina are prime examples.
The study reported focused on mortality following all cyclones "between 1930 and 2015." That is what actuaries call a large data set. Following the rate of demise in the areas in which storms struck, these scientists are comfortable concluding "that the average TC (cyclone) generates 7,000–11,000 excess deaths, exceeding the average of 24 immediate deaths reported in government statistics."
Some would argue with this. Their inclination would be to focus on the “immediate deaths” more specifically. However, there may be some that would take issue in some instances even with the causation of some portion of those. In Hurricane Helene, there were motor vehicle collision deaths that were attributed to slick roads. Such conditions might occur to some degree in any thunderstorm, not necessarily a cyclone. Causation may be debatable in various contexts.
The study authors say they addressed "the effects of 501 historical storms" "between 1930 and 2015." That is a significant volume, much higher than the 165 tropical cyclone landfalls listed by the National Oceanic and Atmospheric Administration in that period. That said, the 501 storm figure equates to 3.5 to 5.5 million deaths using the averages the authors provide (7,000 to 11,000). That is a great many deaths.
Just as the cause and effect may be very difficult to connect linearly in the “immediate” category, it may be more of a challenge for deaths that are further removed temporally from a storm’s immediate impact. The fact is that cyclonic impacts have broad impacts on people, structures, employment, social systems, and more. Therefore, the broad connection between the geography of cyclonic impact and the geography of death leaves many unanswered questions.
Of particular concern to the study authors is that there appears to be more significant “excess” impact "among infants (less than 1 year of age), people 1-44 years of age, and the Black population." Why is there a disparate impact perceived in the mortality rates of these particular demographics? Some might have expected a disproportionate impact among the aged and the ill as their medical usage is usually higher than the norm and service interruption is a noted challenge the authors mention. Nonetheless, the disparity is worthy of note.
The authors suggest that precipitate causes of "excess" may include health "behaviour (sic), healthcare systems, and pollution." Thus, there are interests here in long-term community impacts and ancillary potential causes. Some might therefore see any connection between deferred mortality and a particular storm as somewhat tenuous, The study authors, however, are quick to jump to allegations that similarly the "tracking (of) only these 'direct deaths'" may be a "misrepresent(nation)," signaling perhaps an allegation or suggestion of intent or ill will on the part of the official reporters (government at some level).
The authors are careful not to make suggestions of causation per se regarding the “excess mortality.” Instead, they provide insinuation of outcomes allegedly "traceable" to cyclones. There is mention of potentiality and suggestions of probability that are perhaps supported only by similarities in the data as regards various cyclone impacts, and the later measure of “excess mortality” in those areas.
The similarity of net effect at future points following such landfalls may provide the support the study proposes, even in the absence of causation proof. Or, it is possible that there are other more tenuous, and as yet not identified, parallel causes of "excess," that are worthy of examination and perhaps blame.
Because of this complexity, many approaches to measuring the mortality impact of disasters focus narrowly on enumerating cases where a disaster is the most immediate plausibility. These are described with words like "might" or could. There is characterization in the article of the logic path of "complex cascades of events," which may suggest that at least some of the perceived connections are tangential or inferential. Some may be downright tenuous.
Nonetheless, there remains the evidence of increased mortality, pure numerical inconsistency, that begs for some explanation. And, there may be value in this study in that other potential causes might be explored, debated, and either validated or excluded through the analysis. Elimination of any cause is helpful in narrowing the population that might be actual cause.
The authors see indirect cyclone effects that they deem worth of discussion:
- damage to infrastructure, homes, and businesses
- population relocation
- social and economic disruptions
- ecological changes
- reduced access to basic services
- increased pollution
- crop damage
- insurance payouts, and
- political actions
Some of these are intriguing. There is some patent logic, for example, that there is population increase and thus societal and service burdens might constrain overall service access; but why would this not desperately impact the old (see above)? On the contrary, there is seemingly less explanation as to how insurance payouts or "political actions" might impact mortality. In this regard, there are both intuitive and perhaps counterintuitive examples cited.
The methodology the study employed involved both geography of cyclonic impact sites and geography without. This is not a side-by-side comparison of two communities, one hit and one not, but instead a chronologically different comparison. In other words, the "control" group for analysis on some particular locality after a cyclone is that same community or locality in a prior time period, before cyclonic impact ("each state before a (cyclone) serves as the control for the same state after").
There would be those who would criticize either method (“its not the same kind of city” or “it is a different period, era” etc.). Both are valid potential criticisms. Any study will have its methodology challenges and there will be a likelihood of criticism in any such attempt.
This Nature study dovetails into a recent report from CNN regarding the “harsh new world” of disasters. The author describes “back-to-back jolts of anxiety, uncertainty, and fear.” This is seen as “compounding the suffering inflicted” by the two storms (Helene and Milton) in a matter of days. The article concludes that there is a resulting “vast trail of despair and frayed nerves,” and the potential for “susceptib(ility) to depression and post-traumatic stress syndrome.”
There is a citation by CNN to a UCLA psychologist’s 2022 study that connects increased “mental health issues” tied to prior Florida storms (“Irma and Michael”). There is also suggestion of “exacerbate(ion) (of) existing mental health problems” as part of the overall psychological impacts alleged. Much of that article is devoted to conclusory opinions regarding “climate change” and the opinions of experts regarding whether storms will be more robust or frequent in years to come.
One may validly wonder how much the media's hype of storms contributes to mental health issues, anxiety, and worse. Every storm brings the dire predictors, the adventurous reporters, and the stress on the bad outcomes. I have many times seen the media focus on a single decimated structure in a particular neighborhood. No one tunes in at 6:00 and 11:00 for video of the boats that were not wrecked, the houses that were not destroyed, or the people who were not displaced. There is sometimes a tendency to sensationalism.
The news media learned long ago that the public wants the shocking and the disturbing. The news that delivers that is the one that garnered yesterday's Neilson ratings and today's clicks. How much of society's mental health impact is driven by media, repetition of gloomy stories, and our seemingly innate desire for such shocking scenes?
In much of our current conversation, there is a forgiveness and accommodation for people having their own “truth.” See Infantilized (September 2024). Perusing the media, there is perhaps less accommodation for the "truth" of others suggested in some contexts than others. To some, it appears the media takes sides with its own "truth" in some instances or is forgiving of some "own truth" at times. However, at other times, it is quick to label someone's "own truth" as false. Is there a rational distinction or reason why some "own truth" is more acceptable or excusable than others?
Nonetheless, there will perhaps be more opinionating and prognosticating about the results of storms, the probability of future storms, and the questions of contribution or causation that challenge this or that "truth." Or, there is the chance that only one truth will be deemed worthy of adherence and acceptance. That is, the media may endorse one "truth" over others and heretical labels will be attached to those who dare hold contrary views.
Regardless, the Nature article is an interesting read. The cited CNN story brings some interesting perspectives also. Each drives contemplation and consideration of potentials for consideration and debate to the extent our commonality accommodates expression of views and ideas. That is, to the extent disagreement will be welcome or at least tolerated as healthy to the debate and not heretical or unhinged beliefs simply dismissed by the media or others (if a report labels someone's perceptions "a lie" that perspective is being dismissed by the media, rightly or wrongly).
The bottom line supported by either story is that there is at least the potential for natural events to impact human emotions, naturally or through media amplification. Those who operate businesses in affected communities, send employees into such areas, or have employees with affected family or friends in such areas would do well to consider how to best monitor mental health, address emotional fatigue risks, and accommodate remediation.
These are employer and employee concerns. Both are likely to benefit from open discourse and resource that address both the "immediate" and the "delayed downstream outcomes." It is a societal issue, a work issue, and a personal issue all intertwined and therefore complex.
