Narrative Abstract
Coordinated health research and targeted investments are needed to respond to the increasing frequency and severity of natural disasters. This study aims to describe health research following high-impact disasters from 2010 to 2020 in the United States (U.S.) and to identify gaps and opportunities for improving post-disaster research to reduce adverse health impacts across communities.
Even among the most costly high-impact disasters in the U.S. over the past decade, health research responses were fragmented, largely descriptive, and lacked exposure data and standardization for cross-comparisons. Additionally, longitudinal follow-up of impacted populations, including research focused on the recovery phase or effective interventions, was lacking.
Improvements in post-disaster data collection and generalizable research studies are needed to better characterize health impacts, evaluate interventions, and foster evidence-based strategies to reduce morbidity and mortality of those impacted. Enhanced efforts to improve health research investments and processes are critical to advance disaster preparedness, response, and recovery efforts.
Introduction
Increasingly, severe disasters are resulting in a growing array of short- and long-term health impacts among affected communities. Coordinated research and targeted investments are critical for identifying data-driven approaches to improve public health actions across all disaster phases.Reference Lurie, Manolio and Patterson 1 , Reference Malilay, Heumann and Perrotta 2 Despite an overall increase in research investments over the past several decades, disaster-health research lacks objective evaluations based on validated measures necessary to drive evidence-based practice.Reference Downey, Brown and Calonge 3 , Reference Smith, Burkle and Aitken 4 A closer examination of disaster research and federal programs reveals how the interdisciplinary field often studies individual disasters as stand-alone events, and research results are rarely translated into practice.Reference Picchione and Finegan 5 The authors underscore the stated need for the development of a national disaster research strategy to address the lack of funding, coordination, and prioritization both across and within the federal agencies that support these efforts.Reference Picchione and Finegan 5 Understanding gaps and opportunities to improve upon the existing body of research is essential to the development of this strategy. Between 2010 and 2020, natural disasters cost the United States (U.S.) over $1.1 trillion. 6 This study evaluated the research response for a subset of high-impact disasters, with costs exceeding $1 billion, 6 for this period to assess study characteristics and data collection efforts to inform recommendations for strengthening health research, ultimately helping to create more resilient communities. This initial assessment of the literature aims to identify specific areas of improvement to better target resources.
Methods
High-impact disasters were identified from the U.S. National Oceanic and Atmospheric Administration’s (NOAA) National Centers for Environmental Information (NCEI) U.S. Billion-Dollar Weather and Climate Disasters database.Reference Picchione and Finegan 5 Between 1980 and 2024, NOAA tracked climate/weather disasters in the U.S. with losses exceeding $1 billion. First, disasters occurring between 2010 and 2020 were identified (n = 153), and then, for each disaster type, one disaster with 50 or more deaths was selected. Floods and winter storms had fewer than 50 deaths; therefore, a second disaster was selected from the same category by number of deaths or to diversify regional coverage. A single freeze event occurred during the sample period and was included. A total of nine high-impact disasters were selected based on this approach. These events were chosen as they have significant economic and societal impacts and thus were regarded as having a higher likelihood of generating health research.
To evaluate the research response for the nine selected disasters, articles were identified from PubMed using combinations of the disaster names, types, dates, impacted locations, and relevant health outcome terms. Articles were screened using the following inclusion criteria: (1) peer-reviewed research and public health reports, (2) published between January 2010 and November 2023, (3) studies investigating the disaster of interest and human health outcomes. Article information and study characteristics (e.g., study populations, health outcomes, types of exposures, timing of outcome assessments, follow-up periods, research impact) were extracted and recorded.
Results
A total of 129 articles associated with the 9 selected disasters were included in this analysis (Table 1). Figure 1 compares the number of articles evaluated to the NOAA NCEI CPI-adjusted cost for each disaster. This comparison illustrates the “research response,” evaluated using the total number of articles included for each disaster. Hurricane Maria had the highest cost ($111.6 billion) and largest research response. The research response varied by cost for the other eight disasters. The drought, wildfire, and severe storm events had 12 to 13 articles each, but costs varied from $14.3 to $41.7 billion, and little to no research response was observed for the freeze (n = 0), floods (n = 0 to 3), and winter storms (n = 0).
Articles included for review (N = 129)
Table 1. Long description
The table is organized into six columns: Specific disaster, Author and year, Title, Study population, Health outcomes, and Exposures.
* Hurricane Maria (September 2017): This section contains the largest volume of entries, including studies by Abadie (2022, 2023) on mental health and drug markets, Aranda (2023) on resilience, and Buckley (2022, 2023) on social impacts among older adults. Other studies cover mortality (Cruz-Cano, 2019; Kishore, 2018), perinatal outcomes (Kello, 2023; Lee, 2022), and public health systems (Jiménez Chávez, 2020; Luna-Pinto, 2021). Populations studied are primarily adults, with some focus on children and pregnant women.
* South Carolina and East Coast Flooding (October 2015): Includes Botchway (2021) on social impacts for older adults and Gibson (2018) on mental health for caregivers.
* Southeast/Ohio Valley/Midwest Tornadoes (April 2011): Features studies on hospital admissions (Bell, 2018), child injuries (Campbell, 2012), and mortality (C D C, 2012; Chiu, 2013).
* U.S. Drought/Heat Wave (2012): Lists research on occupational stress (Berman, 2021), respiratory issues (Figgs, 2019, 2020, 2021), and infectious disease (Head, 2022).
* Western Wildfires, California Firestorm (Summer-Fall 2018): Includes studies on air pollution and respiratory health (Casey, 2020), dermatological impacts (Fadadu, 2021), and patient access to opioids (Tseregounis, 2022).
ED, Emergency department.
Cost and article frequency of select high-impact disasters (N = 9) occurring between 2010 and 2020.
CPI, Consumer Price Index.
Figure 1. Long description
The x-axis represents C P I-adjusted cost in billions of dollars, ranging from 0 to 150. A legend at the bottom identifies disaster types by color and article frequency by bubble size.
From top to bottom, the data points are:
* Hurricane Maria: Grey bar, $115.2 billion, large blue bubble, 89 articles.
* U.S. Drought or Heat Wave: Tan bar, $41.7 billion, medium blue bubble, 12 articles.
* Western Wildfires, California Firestorm: Orange bar, $30 billion, medium blue bubble, 12 articles.
* Southeast, Ohio Valley, Midwest Tornadoes: Dark green bar, $14.3 billion, medium blue bubble, 13 articles.
* Texas and Oklahoma Floods: Light green bar, $3.4 billion, tiny dot, 0 articles.
* Midwest, Southeast, Northeast Winter Storm: Teal bar, $2.9 billion, tiny dot, 0 articles.
* South Carolina and East Coast Floods: Light green bar, $2.7 billion, small blue bubble, 3 articles.
* Groundhog Day Blizzard: Teal bar, $2.6 billion, tiny dot, 0 articles.
* Southeast Freeze: Light blue bar, $1.3 billion, tiny dot, 0 articles.
Color coding: Grey for Tropical Cyclone, Tan for Drought, Orange for Wildfire, Dark Green for Severe Storm, Light Green for Flooding, Teal for Winter Storm, and Light Blue for Freeze.
Sixteen different health outcomes were studied, and they varied by disaster type (see Supplemental Table 1). Mental/behavioral health outcomes were studied across all disaster types, and studies of Hurricane Maria included the largest range of outcomes (n = 12). The second most studied outcomes were social impacts and hospitalizations/emergency department visits. Disaster-specific outcomes were also examined, including heat-related illnesses, infectious disease and respiratory outcomes for the drought.
Most studies (n = 92, 71%) used only surrogate measurements of exposure, like proximity to the disaster, to assign exposure groups rather than assessing specific exposures (Table 1). Of those that attempted to assess exposures more directly, 9 studies (7%) of drought assessed meteorological factors (e.g., standardized precipitation index, daily mean ambient temperature), 5 studies (4%) assessed chemical pollutants associated with wildfire smoke, and 22 studies (17%) assessed trauma or stress.
Regarding the time between the disaster and the beginning of health outcome data collection, we observed that most studies (n = 79, 61%) collected data one year or more after the disaster. Few studies (n = 7, 5%) collected data within days or weeks following the disaster.
Study populations were mainly comprised of adults (n = 101, 86%), with only 28 (22%) including higher health risk populations such as children, pregnant women, or older adults (Table 1).
Longitudinal assessments were exceedingly limited, with only 3 studies (2.3%) following participants for over a year.
Finally, efforts to test interventions or translate research into helpful information for future events were classified for each study. Few studies informed public health policies (n = 20, 16%) or development of new disaster resources (n = 7, 5%) (e.g., toolkits, frameworks, or templates for emergency or risk communication). No studies evaluated public health interventions. The majority were descriptive, reporting health outcomes among impacted populations (n = 102, 79%).
Discussion
This study identified a patchwork of research responses following nine high-impact disasters in the U.S. over a 10-year period. The cost of the 9 disasters included in this assessment of the literature totaled $214.1 billion, nearly 20% of the cost of all disasters in the decade studied. These results indicate that critical health research needed to inform health understanding and protective measures is lacking, even following large-scale, heavily publicized events like Hurricane Maria. Data-informed decision-making is critical for disaster response and public health actionsReference Lurie, Manolio and Patterson 1; however, despite the billions of dollars in damages, there is not a proportionate research response to gather information systematically and comprehensively to inform clinical and public health actions. Overall, exposures were not well characterized, perishable data were not often collected, there was a lack of data standardization across disasters, and impacted populations were rarely followed longitudinally. Additionally, research was focused on the recovery phase with limited interventional research or studies of other phases of the disaster management cycle.
These deficiencies limit our nation’s ability to learn about complex disaster-related exposure-response relationships needed to inform and improve measures to further response, recovery, and future preparedness. The present study investigated specific study characteristics of the disaster-health research response to uncover commonalities, differences, and trends across nine different disasters, which allowed us to identify priority areas. For example, utilization of standardized, validated data collection tools and protocols would enhance both health and exposure data collection and further our ability to understand complex disaster-related exposure-response relationships. Additionally, it is essential to follow those impacted, as well as assess any deployed interventions, to understand longitudinal health impacts and the efficacy of any protective measures in reducing adverse health consequences. Strengthening research translation and implementing iterative evaluations to apply evidence to practice has the potential to break the cycle of siloed research, where each disaster is studied as a stand-alone event, and instead share and apply research findings across the disaster cycle.Reference Malilay, Heumann and Perrotta 2 –Reference Picchione and Finegan 5
Consistent with past studies,Reference Downey, Brown and Calonge 3 , Reference Picchione and Finegan 5 the present study highlights the need for a national disaster research strategy. Expanding existing federal programs, such as the National Institutes of Health’s (NIH) Disaster Research Response ProgramReference Miller, Yeskey and Garantziotis 7, the Centers for Disease Control and Prevention’s (CDC) Public Health Emergency Preparedness and Response Capabilities, 8 the National Science Foundation’s (NSF) Disaster Risk and Resilience Program, 9 and the National Aeronautics and Space Administration (NASA) Disasters Program, 10 provides an opportunity to advance needed national disaster research capabilities and connect research to practice.
Limitations
This study has limitations. Articles for named disasters may have been easier to identify, potentially causing under-detection of articles for unnamed disasters. As this assessment of the literature aimed to describe study characteristics, scientific quality was not evaluated or discussed.
Conclusions
Enhanced efforts to improve investments in, and processes to support the conduct of, health research across all phases of the disaster cycle are critical for producing the evidence base and actionable insights necessary to improve disaster preparedness, response, and recovery efforts for our communities.
Supplementary material
To view supplementary material for this article, please visit http://doi.org/10.1017/dmp.2026.10465.
Funding
This work was supported by the National Institute of Environmental Health Sciences (NIEHS) Clinical Research Branch, Contract # HHSN273201600002I, and in part by the Intramural Research Program of the NIH.
Acknowledgement
The contributions of the NIH authors are considered Works of the United States Government. The findings and conclusions presented in this paper are those of the authors and do not necessarily reflect the views of the NIH or the U.S. Department of Health and Human Services.
Author contributions
J. Barkoski—Conceptualization and design, literature review, synthesis of results, led writing and review process. E. Van Fleet significantly contributed to the literature review, synthesis of results, and writing process. A. Liu supported the literature review and contributed to the synthesis of results and writing process. A. Miller—Conceptualization and leadership, contributed significantly to the review and writing process. All authors approved the final draft of the article.
Competing interests
None.
Open access
