Older adults have increased vulnerability to injuries in disasters because of their reduced physical flexibility, decreased sensory responsiveness, medical conditions, and potential socioeconomic challenges that may impede preparation and response capabilities during disasters. 1 - Reference Aldrich and Benson 4 The elderly population aged and older is estimated to surpass 70 million by 2030.Reference Aldrich and Benson 4 , 5 Disaster emergency planning and response for the elderly population deserve careful observations from evidenced-based recommendations and prudent consideration. External attributable risks to prevent injuries in older adults during disasters and public health emergencies may be the absence of proper disaster planning for elderly care services and lack of coordination among health care providers and community emergency personnel. Additional external factors include lack of knowledge and training in care services surge capacity and disaster response preparation by health care staff can render older adults without treatments and urgent care for their injuries.Reference Banks 2 In this study, we are using data collected from New Jersey following Hurricane Sandy to understand the clinical caseload of health care services utilized by those aged 65 and older during and after the storm at emergency departments (ED) and hospitals across New Jersey. This study’s purpose is to augment knowledge for geriatric patient care during and after a disaster and to provide insights on clinical caseload for pre-disaster planning.
Hurricane Sandy was the most destructive and deadly storm of the 2012 hurricane season and the second costliest hurricane in US history.Reference Blake, Kimberlain and Berg 6 - 8 The storm’s impacts in New Jersey led to a virtual statewide shutdown of transportation. All public transportation was suspended, pending limited restoration of scheduled service as assessments were taken.Reference Blake, Kimberlain and Berg 6 New Jersey sustained the brunt of Sandy’s force as devastating damage occurred in almost every section of the state.Reference Manuel 7 As the storm traveled up the northeast coast of the United States, it became the largest Atlantic hurricane on record (as measured by diameter, with severe winds spanning 1100 miles).Reference Blake, Kimberlain and Berg 6 - 8 Current estimates of the US damage wrought by the storm total about $72 billion.Reference Manuel 7 The New Jersey state government estimated construction costs of $29.5 billion to repair and replace the damage caused by the storm.Reference Blake, Kimberlain and Berg 6 - 8 Mildenhall et al reported the storm created damage >346,000 housing units, of which 22,000 were entirely destroyed. 8
The vulnerability of older adults during disasters was revealed in the aftereffects of Hurricane Katrina in 2005. People aged 60 and older accounted for 15% of the population before Hurricane Katrina in New Orleans. 1 However, >70% deaths attributable to the hurricane were 60 and older, according to Grantmakers in Aging, which has been active in the hurricane relief effort. 1 Approximately 47% of all deaths in Louisiana attributed to Hurricane Katrina were people 75 years old or older, with 103 deaths occurring in nursing facilities.Reference Banks 2 It is also likely that behavioral health conditions such as post-traumatic stress disorder, major depressive disorder,Reference Henry, Cooke-Hull, Savukinas, Yu, Elo and Van Arnum 9 and substance use disorder were affected.Reference Goldmann and Galea 10 Stress and chaos in disasters-related conditions are particularly challenging for seniors with mental health conditions such as dementia and those who experienced traumatic life events such as a war and the Holocaust.Reference Haynes 11 A recent study suggests that the need for mental health services was unevenly distributed geographically among Medicaid beneficiaries in the days and weeks following the hurricane.Reference Davis, Wilson and Brock-Martin 12
Hurricane Sandy’s impact on the local health care delivery infrastructure was significant, causing changes in both the supply and demand for health care services.Reference Blake, Kimberlain and Berg 6 , Reference Manuel 7 The immediate surge in demand in emergency and inpatient health care services due to trauma casualties, power outage, and suspended public transportation to outpatient facilities caused a spike in cases that challenged EDs across the state. To gain better insights into medical resources planning and staffing needed to adequately address volume and case mix of health care services presented in the emergency and inpatient settings to the elderly population during the primary surge period of the hurricane, we analyzed comprehensive health care utilization data to show patterns and duration of clinical services provided to patients aged 65 years and older immediately following the hurricane in the State of New Jersey. Complex interactions among the pre-, intra-, and post-disaster health of a given population may also affect the caseload and volume of health services delivered. Davis et alReference Davis, Wilson and Brock-Martin 12 demonstrated a distinct difference between the primary and secondary surge in health service needs. Primary surge is defined as the need for health care during and immediately after an incident generated by trauma, injury, and illness caused by the incident, or by pre-existing chronic conditions not properly treated during the incident.Reference Davis, Wilson and Brock-Martin 12 , Reference Sharp, Sun and Ledneva 13 Followers of the news during Hurricane Sandy may recall that many health authorities also feared the emergence of other threats such as water-borne illnesses enabled by large bodies of standing water and disrupted public infrastructure. After all, the storm’s passage over Haiti produced more than 9000 cholera cases, as reported by the Center for Economic and Policy Research. 14
This study’s approach was to capture and present the primary surge caseload of health care services utilized by elderly adults aged 65 and older in the urgent settings such as EDs and hospitals. It provided a time series of monthly caseload for the top categories of clinical diagnosis in the EDs across New Jersey prior and post the hurricane over a 12-month period. The study also estimated the cost impact of the storm on the hospitals and EDs caring for the elderly population aged 65 and older for the top categories of clinical caseload. Because of the public transportation suspensions, power outages, roadblocks, and driving challenges, the elderly population would likely resort to temporary medical tents, EDs, and medical centers to seek care.Reference Blake, Kimberlain and Berg 6 , Reference Manuel 7 Due to the severity of infrastructure damage and public transportation suspensions, we hypothesized that the hurricane would produce significant demand increases during the primary surge for medical treatments related to acute trauma-related conditions in the elderly population aged 65 and older in urgent medical settings such as the EDs and hospitals. Common chronic conditions for the elderly population including heart disease, hypertension, diabetes, and arthritis would be among the top medical conditions that needed urgent care and treatment in the emergency and hospital settings. 1 To quantify the clinical caseload during the primary surge period and verify the prevalence of chronic conditions presented to EDs and hospitals from the elderly population post hurricane, we conducted analysis on multiple data sets from the Agency for Healthcare Research and Quality (AHRQ), a division of the U.S. Department of Health and Human Services. AHRQ publishes annual updates to the “Healthcare Cost and Utilization Project” (HCUP). The data offered a wide-ranging and comprehensive emergency and inpatient care response picture for the elderly population during the primary surge period of Hurricane Sandy. 15 This study also made available a time series graph illustrating clinical ED caseload over time. It did not delve into the quantification of additional resources needed in the ED and hospital settings to care for elderly patients for the secondary surge and recovery process.
Methods
Data Source
Data was obtained from the AHRQ HCUP, which consists of health care databases collected through a federal-state-industry partnership with sponsorship by AHRQ. 15 HCUP databases are gathered via efforts from state data organizations, hospital associations, private data organizations, and the federal government to generate a national information supply of encounter-level health care data. HCUP contains the largest collection of hospital care data in the United States, with all-payer, encounter-level information starting from 1988. 15 These databases allow researchers to conduct studies in wide range of health policy issues, including health care services utilization, cost and quality of care, accessibility to health care services, and treatment outcomes at the national, state, and local market levels. The HCUP State Inpatient Database (SID) and State Emergency Department Database (SEDD) are databases that capture discharge information at the encounter level to be used for research purposes. The SEDD includes discharge information on all ED visits for a state that do not result in an admission. The SID encompasses all New Jersey community hospital discharges. 15 We obtained both the SID and the SEDD for the state of New Jersey for each year from 2010 to 2013.
Study Population
This study examined encounters for patients aged 65 and older who presented to a New Jersey ED or were admitted to a New Jersey hospital between January 1, 2010, and December 31, 2013. All general acute-care hospitals in New Jersey were included in the data set. In 2012 there were 1,043,274 inpatient stays and 3,124,655 ED visits for all ages. There were 380,989 inpatient stays and 365,962 ED visits for patients aged 65 and above accounting for 36.5% of all inpatient stays and 11.7% of all ED visits in 2012. A similar volume of patients was seen in the other years studied. Treatment patterns were analyzed based on principal diagnosis codes to identify changes in utilization and in the types of health conditions being treated in the elderly population of New Jersey in ED and inpatient facilities in the aftermath of Hurricane Sandy.
Data Analysis
We compared all International Classification of Diseases Version 9 primary diagnosis codes listed for ED and inpatient encounters for elderly patients during November 2012, the month following Hurricane Sandy, to those during November for the 2 years before the hurricane (2011 and 2010) as well as 1 year after the storm (2013). 16 This comparison was done to look for changes in volume and case mix of the elderly population age 65 and above seeking ED and hospital care immediately following Hurricane Sandy, relative to what is typical in November, based on prior and subsequent years. The comparison was based on the proportion of each diagnosis volume that occurred in November to the diagnosis volume of a 12-month period. The volume increase was controlled by calculating the proportion of a diagnosis for the primary surge month of November over the total number of the diagnosis seen in a 12-month period. This method adjusted for confounding factors such as capacity increase, seasonal-related outbreaks, and potential local epidemics that may account the volume increase in the EDs and inpatient settings outside of the hurricane effect.
ED and inpatient data were analyzed separately to explore differences in diagnoses presented at each type of health care facility independently. Diagnosis codes were examined at the 3-digit level because we observed changes in coding practices beyond the third digit across years. In presenting the top categories of health services contributing to the highest volume and caseload increase post hurricane, we set a threshold of ~10% increase in the November share for each diagnosis code in 2012 when compared with the November share for 2010, 2011, and 2013 for this study. The diagnosis codes that were above this threshold were identified as the minimum benchmark standard in this study for conditions with a significant increase in utilization in the month following the hurricane. By setting the benchmark to a minimum of a 10% increase from baseline, it allows this study to locate and focus on the health services that significantly contribute to the primary surge in health services demand of the elderly population post hurricane. The Wilcoxon rank-sum test was used to determine if there is a significant difference in volume for November 2012 compared with other months in the study period. The rank-sum test is a non-parametric test of the null hypothesis that it is equally likely that a randomly selected value from November 2012 will be less than or greater than a randomly selected from all other months in the study period. The Tukey test was performed to identify clinical caseload outliers in November 2012 in comparison with other months from September 2012 to December 2013. Outliers are values below Q1−1.5(Q3−Q1) or above Q3+1.5(Q3−Q1). The first quartile, denoted Q1, is the value in the data set that holds 25% of the values below it. The third quartile, denoted Q3, is the value in the data set that holds 25% of the values above it. The interquartile range is defined as follows: Q3−Q1. When there are outliers in a sample, the median and interquartile range are used to summarize a typical value and the variability in the sample, respectively. 17
Results
In this investigation, we produced a timeline of elevated clinical caseload for the elderly population aged 65 and older post Hurricane Sandy in the State of New Jersey. We presented clinical caseloads with the highest increase from the baseline during the month of November post Hurricane Sandy to establish the primary surge effect. The Wilcoxon rank-sum test demonstrated a significant difference in the volume of cases by diagnosis code in November 2012 compared with other months for both inpatient and ED patients (P-values: inpatient<0.0001; ED<0.0001). The baseline was defined as the average of the month of November in 2010, 2011, and 2013 (2 years before and the year after the hurricane) for elderly patients aged 65 and older. Figure 1 displays the monthly ED clinical caseload for the 3 most frequent physical injuries and mental disorder diagnosed and treated in the ED over a 12-month period. Physical injuries such as fractures, open wounds and fractures, and mental illness were the most frequently treated conditions for the month of November 2012 immediately after the hurricane. Figure 1 also illustrates the immediate surge of different types of clinical caseload and captures any notable variations in clinical caseload following the hurricane. We conducted an analysis to quantify clinical caseload variations over a period of 12 months following the hurricane. The data showed the variations in clinical caseload diminished beyond 9 months post Sandy. Therefore, Figure 1 displays a 9-month post hurricane caseload data to demonstrates any notable variations. The caseload decreased rapidly from the peak in the next 3 months after November to reach a temporary trough. The trough then took an upward turn and increased after February 2013. Six clinical category high outliers for the month of November 2012 were identified by the Tukey outlier test. The 6 clinical high outliers were fractures, injury other and unspecified, encounter for other and unspecific procedures and aftercare, observation and evaluation for suspected conditions not found, visual disturbances, and repeat prescription refills.
Monthly Variation of Emergency Departments Clinical Caseload Over a 12-Month Timeline. The top 3 physical injuries (fractures, contusions and open wounds) and mental illness diagnosis with their monthly percentage of caseload over a 12-month timeline for population aged 65 and older in the State of New Jersey. It illustrates the primary surge peak in November 2012 post Hurricane Sandy, a decrease after 3 months and an upward trend that follows.
The primary surge of physical injuries and mental disorder ED caseload is illustrated in Figure 2. Trauma-related injuries such as open wounds, fractures, contusions, sprains and strains, and other unspecified injuries had the highest increase by caseload of 880 cases seen and treated in the ED for the month of November 2012 relative to the baseline. The surge of ED clinical treatment demand caused an extra $7.3 million of health care costs for the top 15 clinical caseload categories in the month of November 2012. On a daily basis, there was a volume increase of an average of 57 cases for the most frequently encountered medical conditions in EDs for the month of November 2012 across New Jersey for the elderly population aged 65 and older. Mental illness caseload was the second most frequent after the injuries related diagnostic category of services needed care and treatment in the ED. Table 1 reports health care charges and presents the caseload volume of the top ED surge clinical diagnostic categories.
Hurricane Sandy Primary Surge in Emergency Departments (ED) Cases From Baseline in New Jersey. The primary surge of physical injuries and mental illness cases in New Jersey ED clinical cases for November 2012 following Hurricane Sandy relative to baseline (averages of November 2010, 2011, and 2013) for the elderly population aged 65 and older.
Number of Emergency Department (ED) Primary Surge Cases in November 2012 and Associated Total Charges
The highest increase of ED clinical caseload categories and their charges during primary surge period of Hurricane Sandy in the State of New Jersey for population aged 65 and older.
Chronic conditions were frequently observed in the elderly population in the ED and inpatient settings. Essential hypertension was the most frequently encountered chronic condition in both ED and inpatient during the aftermath of Hurricane Sandy. An average of 16,000 cases of essential hypertension were seen in these 2 settings. Essential hypertension represented over 10 times as many as the average number of 1400 cases in injury-related diagnoses seen in the ED. Disorder of lipoid metabolism was the second highest chronic condition treated in both ED and inpatient settings. However, chronic condition caseload did not change significantly compared with the baseline. Table 2 illustrates the relative comparisons of different chronic conditions with acute conditions in the ED and inpatient settings.
Top Chronic Conditions in Emergency Department (ED) and Inpatient Settings (Patients Age 65 and above), November 2012 New Jersey
The most frequent chronic conditions treated in ED and inpatient settings during November 2012 in New Jersey for elderly patients aged 65 and older.
a Chronic conditions counted by looking at secondary diagnosis codes for ED and inpatient settings.
b Among ED diagnoses with a large increase over baseline in November 2012.
Most ED trauma presentations did not result in inpatient hospitalization. The biggest increase in caseload for November 2012 relative to baseline was for respiratory system and chest symptoms. These symptoms accounted for an increase of 543 admissions for respiratory-related problems. The second highest increase in admissions was for disorders of fluid electrolyte and acid-base imbalance balance resulting from dehydration. The highest increase by percentage of admissions relative to the baseline was intracerebral hemorrhage with an increase of 3% as illustrated in Figure 3. The caseload was 146 associated with intracerebral hemorrhage. It was presented as the highest by percentage because the denominator of admissions for the diagnosis was smaller.
Percentage of November 2012 Inpatient Clinical Caseload Comparison With Baseline. The comparison of the November 2012 primary surge for inpatient caseload to baseline percentage of clinical caseload in November 2010, 2011, and 2013 in New Jersey for patient population age 65 and above seen following Hurricane Sandy.
Similarly, Table 3 presents the proportional increases from the baseline for the top ED cases in November 2012. For the month of November following the hurricane, there was a total increase of 35% in ED encounters for the top 10 diagnoses for adults aged 65 and older who visited EDs throughout the state of New Jersey. The highest percentage increase in ED presentations was for repeat prescription refills, which increased by 10.7% in comparison with the baseline. It is important to note that this percentage increase represents a very modest number of actual prescriptions. The baseline prescriptions and refills cases amounted to just 134 counts, as compared to the much larger numbers of trauma presentations. This large increase is again due to the relatively small denominator used for calculation.
Increase in Emergency Department (ED) Cases in November 2012 From Baseline by Diagnosis
The percent share of November 2012 encounters per diagnosis in EDs and the increase from baseline for the state of New Jersey in the years 2010–2013 for population aged 65 and older.
a Repeat prescription refills was observed to have the highest increase from the baseline.
Discussions
Elderly adults with multiple chronic conditions typically require regular medication refills and office visits for management of their conditions. Chronic conditions are prevalent among elderly adults aged 65 and older. Close to 80% of older adults have at minimum, 1 chronic condition such as heart disease, cancer, diabetes, or stroke, and 50% have 2 or more. 1 The CDC’s National Center for Health Statistics reported that hypertension occur in nearly 50% of older adults aged 65 and over. Other frequent chronic conditions include arthritis, heart disease, cancer, and diabetes. 1 Medical and public health challenges are to be expected after disasters such as Hurricane Sandy. It is important to first identify and treat storm-related casualties requiring immediate attention.Reference Redlener and Reilly 18 This study confirmed that trauma-related cases were the leading primary cause of the surge for ED presentations involving elderly citizens. The primary surge of ED clinical treatment demand caused an extra $7.3 million of health care costs for the top 15 clinical caseload categories in the month of November 2012. The presentations for chronic conditions such as essential hypertension cases were encountered 10 times as frequently as the trauma cases. Essential hypertension-related health services and medication demand presented much more frequently than trauma-related health services and supplies. Furthermore, repeat prescription refills had the biggest increase in percentage relative to baseline. Elderly adults aged 65 and older who lost their medications during the hurricane or who lacked outpatient access due to transportation challenges presented at EDs for their repeat prescription refills. Lack of access to clean drinking water sources could be particularly challenging to frail elderly adults or adults with multiple chronic conditions, thus resulting in the large increase in dehydration and fluid imbalance cases. Dehydration resulting in fluids and electrolytes disturbances caused the second highest admission for elderly adults aged 65 and older after the hurricane. The same finding was reported by Swerdel et al; they found that hospitalizations with dehydration increased significantly immediately after Hurricane Sandy in New Jersey. They also found that those over 65 years of age accounted for the vast majority of the increase.Reference Swerdel, Rhoads, Cosgrove and Kostis 19
A natural disaster is a stressful and traumatic event and often can trigger an acute exacerbation of existing chronic conditions, including any elderly adults’ underlying mental health conditions. The Centers for Disease Control and Prevention reported that about 40% of shelter evacuees might have suffered from post-traumatic stress disorder after Hurricane Katrina. 1 In our study, episodic mood disorders were among the top inpatient admission diagnoses observed. Mental illness can be observed as the second highest in clinical caseload treated in the EDs after trauma-related injuries. These findings may represent an important area for the public health community to continue to work on to reduce morbidity and the associated burden on our health care systems.
Limitations and Gap Analysis
This study focused on the month of November post Hurricane Sandy as the immediate impact health services utilization assessment. It did not address any ripple effect or sustainment effort required to address the effort beyond the initial peak of cases. The data source did not identify the specific day of month. Thus, this study focused on the month of November as a whole. Immediately following the storm, there were other agencies that responded to treat and care for all casualties. The U.S. Department of Health and Human Services sent over 500 personnel, including 9 Disaster Medical Assistance Teams from 8 states, to deliver care at medical shelters. The American Red Cross opened 171 shelters across 13 states, with thousands of volunteers working alongside paid personnel throughout the area.Reference Manuel 7 FEMA set up 68 Disaster Recovery Centers in Connecticut, New York, and New Jersey where people could apply for assistance and seek information on alternative housing. Within 24 hours, FEMA supplied more than a million liters of water and more than a million plate-ready meals to the New York National Guard and to volunteers to distribute to those in need.Reference Manuel 7 This study’s focus was in emergency and hospital care and thus it did not capture additional urgent clinical caseload provided by disaster relief agencies such as the ones mentioned above, or any outpatient services that were provided by free-standing physician clinics and practices.
Conclusion
In the aftermath of Hurricane Sandy, EDs and hospitals experienced an increase in case volume and health care expenditure related to trauma, other acute medical conditions, and underlying chronic conditions for the elderly population aged 65 and older in the State of New Jersey. Many elderly adults sought care in the EDs and hospitals during the aftermath of Hurricane Sandy due to transportation disruptions. The highest workload increase can be observed in injuries related to trauma such as open wounds, fractures, contusions, sprains and strains, and other unspecified injuries, which collectively accounted for the highest increase in volume seen in the ED for the month of November 2012. New Jersey EDs handled on average 1700 more cases for the month of November 2012 (post Hurricane Sandy) relative to the baseline for the top 15 most frequently diagnosed and treated ED medical conditions for the elderly population aged 65 and older. The primary surge in these ED care services required extra medical resources that generated an ~$7.3 million increase in health care costs for the month of November 2012. The elevated clinical caseload of trauma-related injuries sustained for 3 months. This study confirmed that trauma-related cases were the leading cause of the surge in elderly population aged 65 and older visiting the EDs for care and treatments during the primary surge of the hurricane. The surge in trauma-related care services generated ~$4.0 million of the $7.3 million total health care costs for November 2012. Most ED trauma presentations did not result in inpatient hospitalization.
Chronic conditions such as essential hypertension cases were encountered 10 times as frequently as the trauma cases. However, the utilization and demand for chronic conditions-related health services and medication did not increase much more significantly than the baseline during and after the hurricane. Study findings showed inpatient trauma-related workload did not increase as much as outpatient workload, as most presentations were treated and discharged without admission. Fracture injuries, specifically fractures of the humerus, radius, and ulna were reported as the fourth-highest admission diagnosis for the month of November 2012. Tomio et al summarized that a considerable number of recommendations for the disaster preparedness of chronically ill patients were developed by the lessons learned from recent disasters. Some of the recommendations were based on prior disasters’ epidemiological findings; however, an abundance of recommendations came from anecdotal evidence or expert opinions. In addition, most of these recommendations had not been scientifically studied and validated.Reference Tomio and Sato 20 Findings of this study provided a quantitative effort to understand the immediate surge workload as ED and hospital staffs responded to injured and medically ill adults aged 65 and older during and after Hurricane Sandy in New Jersey. We hope these findings contribute to literature that offers evidence-based results and insights for the elder medical care services and disaster preparedness.
Future Direction
In future studies, we plan to identify hospitals within geographical areas most impacted by the storm, study their surge levels, and compare them to hospitals within less-impacted geographical areas to further understand medical services utilization in the aftermath of Hurricane Sandy in New Jersey. We also plan to further quantify and focus on the number of mental health cases observed post Hurricane Sandy and monitor the associated medical services utilization and length of related medical treatments as a result of the impact of a natural disaster to understand any secondary surge effect of the hurricane. In the number of articles published on elderly disaster victims, not very many contain the focus on frail elders. In addition, the majority of researchers do not differentiate between frail and strong elderly adults in their studies.Reference Redlener and Reilly 18 Our future work will aim to use both age and level of physical impairment as indicators of which populations aid workers should target first at disaster relief sites.
