When you see the sometimes vague and swiftly changing death counts after wildfires out west, tornadoes down south, hurricanes across the coast, or extreme heat literally anywhere, you might ask yourself: “Why is it so hard to figure out how many people died after a natural disaster?”

The short and unsatisfying answer is, “It’s complicated.” Here’s a quick primer.

There are many people who respond to a disaster, but when it comes to deaths, the local medical examiner or coroner (ME/C) office takes the lead in determining whether a death is disaster-related and reporting that fact on the death certificate.

Most disaster-related deaths are certified by the ME/C office, since they typically take jurisdiction over deaths that are sudden, unexpected, violent, or suspicious. However, some disaster-related deaths happen in hospitals, nursing facilities, and homes without obvious evidence that the death was related to a disaster. In those instances, an attending physician might certify the death, but ideally, these cases would also be turned over to the ME/C office.

Death certificates are typically certified through state-specific vital statistics systems, which are used to determine how many deaths were related to the disaster, based on the information on the death certificate. The National Vital Statistics System (NVSS) compiles data from vital registration systems across the United States, enabling officials to determine a death toll for a disaster that impacted multiple jurisdictions.

Well, lots of things, but here are three big ones:

1. Not enough evidence. If there isn’t enough information collected during the death investigation or if the circumstances of the death aren’t described very well in the death investigation report, an ME/C may not be able to tell if the death was a result of the direct physical forces of a disaster or unsafe conditions caused by the disaster. If this happens, an ME/C cannot “count” the death as disaster-related. 

2. Different definitions for “disaster-related death.” Historically, there has been some debate about what is considered a “disaster-related death.” If different ME/Cs do not use the same definition, some disaster-related deaths may not be reported, which could lead to an inaccurate death count. For example, if an ME/C only defines disaster-related deaths as those caused by the physical forces of a storm (e.g., blunt trauma from flying debris or drowning) and not unsafe conditions (e.g., impassible roads/facility closures preventing someone from accessing dialysis or other health care services), then that could lead to an undercount of disaster-related deaths.

3. Incomplete information about the disaster on the death certificate. If the death certificate does not use the disaster (e.g., Hurricane Helene) or describe it in some way (e.g., Kentucky tornado, 2025), it makes it hard for vital statistics agencies to count the number of deaths and provide an accurate death toll after the disaster.

    

So, like I said, the reason it’s hard to figure out how many people died after a natural disaster is because it’s complicated.

For the past decade, NORC has been working with the Centers for Disease Control and Prevention (CDC) to develop tools, resources, and trainings to improve this process and address these three reasons. In 2017, we developed a toolkit to help death investigators collect specific disaster-related information that is often only available immediately after the incident. More recently, we developed a quick guide that supplements the toolkit, includes key data elements, and summarizes common causes of death and risk factors for five common disasters. Both resources help to address reason #1 above.

We have also developed several materials to help ME/Cs determine whether a death is disaster-related and how and where to record disaster-related information on the death certificate (which addresses reasons #2 and #3 above). These resources include a poster-sized flowchart that can be hung on the ME/C office wall and a series of fact sheets that provide background information about common causes and risk factors for death, summarize the action steps to determine if the death is disaster-related, and provide examples of completed death certificates for six big disaster types: heat, winter weather, tornado, hurricane, thunderstorm, and wildfires.

If that wasn’t enough, we even designed and built a comprehensive online training that covers everything you could ever want to learn about disaster mortality. It explains the process in greater technical detail than I have, walks through how to use each of the available tools I mentioned above, and, best of all, lets you practice using the tools based on real-life scenarios for hurricane, heat, winter weather, tornado, and thunderstorms.

Now that you have a better sense of why it’s so hard to figure out how many people die after a disaster and what NORC has been working on to improve the process and support the ME/C offices that do this critical work, I want to briefly tell you why getting it right matters. It’s not just an accounting problem. Getting it right means closure and support for families. If a disaster-related death is correctly attributed to the disaster on the death certificate, families can use that documentation to apply for funeral benefits from the Federal Emergency Management Agency.

Getting it right means researchers can better monitor the causes and circumstances of death to measure the burden and severity of a disaster and identify risk factors, which ultimately helps public health officials and emergency managers plan for the next disaster and prevent future deaths. So, the next time you read an article about how many people died after the latest natural disaster, I want you to remember that the process is complicated and so important to get right!