DisastersThe Complications of Counting Casualties after Natural Disasters

Published 19 September 2019

There are many gray areas when collecting data on how and why people died in a disaster. A new study now underway aims to identify best practices for collecting, recording, and reporting death and illness data during and immediately after large-scale weather disasters.

There are many gray areas when collecting data on how and why people died in a disaster.

A National Academies of Sciences, Engineering, and Medicine study that is now underway aims to identify best practices for collecting, recording, and reporting death and illness data during and immediately after large-scale weather disasters.

Sometimes, it can take months or even years for a disaster death toll to become fully known. Death counts can include drowning, or factors like disruptions to medical care, infections from contaminated water, or an injury from hurricane-proofing the roof in the days before the storm.

NAS notes that collecting this information is critical for informing recovery efforts and for preparing for the next disaster. However, the challenges have become ever apparent in the aftermath of Hurricane Dorian in the Bahamas. First responders have to balance addressing people’s fundamental needs (food, water, shelter, and clothing) with counting fatalities - often in vast areas of devastation.

“There’s a difference in how you’ll respond to and prepare for disasters if more deaths were attributed to electrocution, versus people not having power to refrigerate their insulin,” said José Cordero, head of the Department of Epidemiology and Biostatistics at the University of Georgia, at a recent public workshop that kicked off the study’s information-gathering process. “Accurate data collection can not only lead to policy change, but change in how we design our interventions,” he said.

Standardizing and Streamlining Data Collection
Currently, the types of data that are collected are usually specific to one disaster or region, said Amy Davidow, associate professor at the Rutgers University School of Public Health in New Jersey. Workshop attendees discussed the merits of a standardized set of data items that are collected on the national, state, and local levels — so responders don’t have to “wing it” after every disaster.

“What if, after every event, we agreed, ‘we should look at these X number of things’?” she said.

Edward Kilbane, a physician-forensic pathologist who works in the federal National Disaster Medical System, added that sometimes in disaster data collection, less is more. Although the committee’s report will focus on weather-related disasters, epidemic response efforts offer important lessons learned. During the 2014 Ebola outbreak in West Africa, Kilbane was part of a team of responders that was trying to collect data on 200 variables across several small villages. “While any researcher can make the case that those variables were critical, focusing on 200 variables didn’t help us respond in a timely way,” he said.  

Umair Shah, executive director of the public health department of Harris County, Texas, said data collection has come a long way since Hurricane Katrina in 2005, when people had little choice other than to resort to paper. Nearly 15 years later, there’s a “sea of information” coming from electronic health records, social media, and health wearables, he said, calling for better workforce training to ensure responders can make sense of the data.

Disparities and Disasters Often Go Hand in Hand
Natural disasters tend to have a disproportionate impact on vulnerable populations, including people with disabilities, the elderly, and communities that are already hard-hit by other factors like poverty and unemployment.

“If you take care of the most vulnerable in the wake of disaster, you take care of everyone,” said Maureen Lichtveld, a member of the study committee and professor at Tulane University’s School of Public Health and Tropical Medicine.

Data collection efforts should not only look at whether people had their basic needs met, but whether they are getting social and emotional support as well.

In the two years since Hurricane Maria, the mental health toll is still palpable, said Francisco Murphy-Rivera, a family physician in Adjuntas, Puerto Rico, who was on the frontlines of the Hurricane Maria response. However, mental health surveillance often falls through the cracks in data collection, and in Puerto Rico, a persistent stigma connected to mental health deterred many residents from seeking help, he said. Recognizing these challenges, his practice began offering primary care, behavioral health care, and social services, including food and water assistance, under one roof.

Survival versus Resilience
Lichtveld emphasized that there’s a difference between surviving a disaster and being able to recover from the stress and hardship. A National Academies report from earlier this year recommends that communities evaluate such measures of resilience after a disaster, including social, economic, and physical/environmental.  

José Cordero, the University of Georgia public health professor, added that every disaster is an opportunity to improve data reporting, but responders must not lose sight of the human aspect of the work.

“When we’re talking about morbidity and mortality — the dots on a map — remember, those are people,” he said.

Visit the workshop page for the archived webcast and presentations.

For more information on this study, including future workshops, visit this page. The committee expects to release its final report in 2020.