Abstract Number: 242 | ID: 2017-242
Evaluating Injury Emergency Department Visit after Hurricane Sandy and Developing Community Vulnerability Index
Shao Lin(Department of Environmental Health Sciences, University at Albany, State University of New York, United States, firstname.lastname@example.org), Wangjian Zhang(Department of Medical Statistics and Epidemiology and Health Information Research Center & Guangdong Key Laboratory of Medicine, School of Public Health, Sun Yat-sen University, China), Ziqiang Lin(Department of Mathematics and Statistics, College of Arts and Sciences, University at Albany, State University of New York, United States), Wayne Lawrence(Department of Epidemiology and Biostatistics, University at Albany, State University of New York, United States), Jillian Palumbo(Department of Environmental Health Sciences, University at Albany, State University of New York, United States), Yuantao Hao(Department of Medical Statistics and Epidemiology & Health Information Research Center & Guangdong Key Laboratory of Medicine, School of Public Health, Sun Yat-sen University, China)Background/Aim: Extreme weather events such as hurricanes was found to be related to multiple health outcomes. However, significant gaps remain in our understanding of the impacts of Hurricane Sandy on injury especially the modification effect of community-level factors. Furthermore, little is known about the vulnerability of population during the events. This study was designed to assess injury emergency department (ED) visit after Sandy in New York State (NYS), and to develop community vulnerability index (CVI) to identify areas that might be more vulnerable to future storms.
Methods: Distributed Lag Nonlinear Models (DLNM) were used to examine the impact of Sandy on injury ED visit for each county. Meta regression was used to pool county-level estimations and identify the modification effect of community-level factors on Sandy-injury association. Sandy-attributable number of cases estimated with DLNM and community-level factors identified with meta regression were then incorporated into Boosted Regression Tree (BRT) models to determine the weight and the direction of the impact of each factor for developing CVI.
Results: Overall, the risk of injury significantly increased 3-4 days after Sandy. The impact of Sandy varied across the study areas, with the RR highest in directly affected areas (RR: 1.1~2.5). In terms of community vulnerability, percentage of mobile homes had the largest contribution (P=0.001, weight=-25.87%) to the spatial variation of cumulative RR of injury related to Sandy, followed by the percentage of uninsured people (P=0.001, weight=24.90%), M/F sex ratio (P<0.001, weight=-10.26%) and the percentage of Hispanic (P<0.001, weight=8.95%). Counties with high CVI (>M+1.5SD) was clustering in directly affected areas.
Conclusions: The risk of injury ED visit increased after Hurricane Sandy in directly affected areas, compared to indirectly affected areas. Community-level factors have significantly modified the impact of Hurricane Sandy. The residents living in Sandy areas have the highest vulnerability.