Last week was likely the deadliest in modern Maui history—all causes combined.

Was climate change the “real” cause?

Data: CDC and Maui officials. Image: Inside Medicine.

Wildfires devastated the island of Maui last week. The Lahaina fire now stands as the deadliest natural disaster in the state’s history. So far, at least 93 people have died.

That appears to be just the start. Many bodies have been recovered and are in morgues—not yet included in official death counts because they remain unidentified. It’s also likely that other deaths indirectly caused by the fire will eventually add to the toll. Some people needing dialysis may have trouble reaching a facility, for example. Or, in other emergencies in which time matters, ambulances may be delayed—if available at all. Hospital resources for a wide variety of serious conditions, both human and material, may be impossible to access temporarily.

Meanwhile, it is reported that approximately 1,000 people were missing over the weekend, a horrifying number. If true, this means that 0.6% of Maui’s entire population was (or remains) missing. Unlike the 2018 California Camp Fire, where around 1,300 people were at some point missing over 153,000 acres (85 were eventually declared dead), the Maui fire covered a much smaller area, under 2,200 acres. That suggests the death toll among those missing on Maui could be much higher.

The context.

In a typical midsummer week, Maui County records around 25 deaths from all causes combined. Assuming those other expected deaths occurred as usual last week, the mortality counts for Week 32 of 2023 will be at least quadruple the typical rates, if not far, far worse. It is therefore likely* to have been the deadliest week in Maui county’s modern history, far worse, for example, than the peaks of the Covid-19 Delta and Omicron waves. (Hawaii did comparatively well overall during the pandemic). The fire is certainly now the state’s worst-ever natural disaster, surpassing a tsunami that killed 61 on the Island of Hawaii (the “Big Island”) in 1960 and the deadliest natural disaster in the US in decades, if not longer

CDC statistics tell us that from 1968 to 1978, around 7 people per week died in Maui. From 1979 to 1998, that number rose to 11 deaths per week, as the population grew. From 1999-2020, the county averaged 19 deaths per week, again growing with the census. In 2021 and 2022, weekly deaths typically ranged from 20-30, depending on the time of year and—of course—Covid-19. (Hawaii largely kept Covid out in 2020).

Did climate change just become Maui’s leading cause of death?

While we can’t say to what degree the fires were caused by climate change, experts seem to believe it was a major factor, if not the driving one. Other factors, however, may have contributed. Still, wildfires don’t tend to destroy beachfront property in Hawaii. That means it is reasonable to hypothesize that climate change was the root cause of death for some, most, or all of these deaths. We’ll never know for sure.

And that’s a problem.

What the records won’t reflect, but probably should.

Regardless, “climate change” is not a label that the Hawaii Department of Health or the CDC will apply to any of these deaths. The death certificates of those who perished are likely to officially cite the International Diseases Classification Version 10 (ICD-10) codes “X00-X09” for fire and flame as the underlying causes. But if the true cause of these deaths was in fact climate change, a cause for which no official or agreed-upon code exists, they won’t accurately appear in our vital statistics. While a few doctors have written climate change as a cause of death on certificates in the past, there is currently no unity of practice. That makes deaths like these harder to track over time.

The epidemiology of a tragedy.

Whenever one-off mass casualty events occur in smaller jurisdictions, whatever the event was can easily temporarily become the leading cause of death in that place. In September 2001, terrorism—ICD-10 code U01.1, Terrorism involving destruction of aircraft—became the leading cause of death in New York City among residents under age 90. In Uvalde, Texas, homicide—ICD-10 code X95, Assault by other and unspecified firearm discharge—was the leading cause of death in May of 2022 for all ages (18 of the 21 were young children).

What determines the epidemiologic (as opposed to the obvious sociological) importance of these events, then, is how likely such instances are to be bellwethers for future events. In turn, awareness of those threats can turn into political will, motivating actions which can effectively prevent them, if not make them far less likely. For whatever missteps the United States government took in the “war on terrorism,” there has not been another successful similar-scale attack on US soil since 9/11, and that’s not due to luck. On the other hand, our failure to act after the Sandy Hook school massacre in Newtown, CT seems to have had the opposite effect.

Climate change deaths must be counted.

Deaths caused entirely or in part by climate change are not going away. But we should not expect them to look alike. They are as diverse as one can imagine. Ninety-three deaths from a wildfire in Maui in August. Twenty deaths, say, from heat waves in July. Ten deaths, perhaps, from a flood somewhere in June.

The numbers add up. But we are not adding them up.

Our want for a unified way to count these deaths as being connected to a common source—climate change that we have caused—stands to hurt us. We simply can’t put it all together to get a realistic account of where things stand. It’s death by a thousand cuts—or by a couple of ticks on our thermometers, year-over-year.

To use hospital lingo for a cardiac arrest in progress, climate change is causing humanity to slowly code. Unfortunately, when it comes to death certificates, we don’t have a code for that.

---

*Note: I can’t absolutely confirm this statement. Weekly data back to 2018, monthly data back to 1999, and yearly data back to 1968 suggest that a week with over 100 all-cause deaths on Maui is exceedingly unlikely to have ever occurred, but I do not have raw weekly pre-2018 data.