What do people need to know to make solid decisions about health after a disaster? What would help people recover fast from a disaster? What information would help healthcare systems recover? A few weeks after Hurricane Sandy made landfall in New York these burning questions and others became the focus of intense discussion among a group of scientists and subject matter experts who gathered at the New York Academy of Medicine.
They developed a set of research priorities. Many of the priorities identified by the group - including questions about the health outcomes of evacuated patients, the morbidity and mortality among at-risk populations, and the health system response to the storm - are most effectively answered by examining large sets of data - big data - about health services.
Past studies that use administrative claims data make assumptions about a patient’s exposure to the storm, often using county-level damage assessments. While such studies can yield valuable public health information, their findings are imprecise because not everyone in a given county or other geographic region is impacted equally by a disaster.
For Hurricane Sandy recovery researchers, ASPR found a way to improve the research process: linking the data available through multiple federal agencies so that the outcomes for individual (but unidentified) patients can be followed using various non-medical criteria. These datasets are compiled from patient data and medical and housing claims data in a way that protects patient privacy. As such, they do not include fields for such information as patient name, but can include de-identified data on the amount of damage sustained by a patient’s home, or the length of time a patient was without electricity.
This ASPR Sandy dataset promotes research efficiencies by coordinating researcher access to the data and reducing duplicative requests for data. The Hurricane Sandy researchers will have access to a single, consolidated data source that can be used by multiple investigators, now and in the future.
This Sandy dataset represents the first time ASPR has compiled and linked this type of data for researchers. The lessons learned from this process will pave the way for similar research projects to be initiated more swiftly after future disasters. It’s an important step toward ensuring that health services researchers can gain prompt access to important sources of data to answer pressing – potentially lifesaving – questions about response, recovery, and preparedness.
ASPR recently funded four new projects, collaborative grants that utilize the ASPR Sandy Dataset to build upon, augment, or enhance the original ASPR, CDC, and NIEHS Sandy recovery studies.
In the first one, a New York State Department of Health researcher is using big data to assess the health impacts of Hurricane Sandy on elderly residents in affected areas of New York State, evaluate the infrastructure and surge capacity of nursing homes and assisted living facilities, and develop a predictive model and risk assessment tools to predict and mitigate the impact of future storms on the elderly.
The second is being conducted by a researcher from Rowan University in collaboration with another researcher from Virginia Polytechnic and State University. These scientists are using big data to determine how environmental changes impact older adults’ health care utilization in the months following a disaster, and inform healthcare providers and policy makers of the unintended downstream health consequences of disaster-related disruptions in housing and community-based healthcare facilities.
A researcher from the Feinstein Institute of Medical Research is using administrative claims data to investigate the differential health effects of Hurricane Sandy between Medicare/Medicaid and non-Medicare/Medicaid populations. This study is called “Post Hurricane Sandy Implementation of a Regional Public Health Surveillance System in Long Island.”
In the last study, a New York University School of Medicine researcher is using the dataset to determine whether early, real-time data from prehospital, emergency department, and overall healthcare utilization can predict the geographic regions and vulnerable populations most affected by natural disasters.
Even if you’re outside the Hurricane Sandy impact zone, the results from these studies may have important implications so watch for their findings. The studies could help your community plan and respond to disasters in ways that help everyone recover faster. And when that happens, your community become more resilient and our nation’s health becomes more secure.
