The most severe acute effect of hurricane landfall is death from drowning, electrocutions, or physical trauma [6–9]. Older age increases the risk of death—nearly 85% of people killed during and in the immediate aftermath of Hurricane Katrina were aged 51 and older, and almost half were older than 75 years of age [9]. Residents of nursing homes that sheltered in place were among those killed by drowning [10]. As with other natural disasters, low-income populations may be particularly vulnerable [11].

The causes and age pattern of deaths during the impact phase of Superstorm Sandy were generally consistent with prior storms. In the acute phase, Sandy caused 43 deaths in NYC. Death was caused most frequently by drowning associated with the storm surge (n = 34, 79%). Other deaths were caused by falling trees, falls, electrocution, and other trauma. Nearly half of fatalities occurred among adults aged 65 or older (n = 20, 47%), and more than half of deaths occurred on Staten Island (n = 23, 53%). Epidemiologic studies will be needed in the coming months and years, however, to assess the storm's full impact on excess mortality from accidental and natural causes, as well as other health impacts, in impacted communities.

Heavy precipitation weather events can also cause flash flooding, which occurs when water from heavy rains collects in a relatively short time and runoff is accelerated in mountainous or narrow valley terrain [12]. Nationally, flash floods are the most common cause of flood deaths via drowning [12, 13]. Flash floods occur in NYC but historically have rarely been life threatening because of local topography. However, much of NYC's infrastructure, especially in low-lying or poor drainage areas, cannot cope with more than one inch per hour of rainfall [14].

Hurricane landfall can also result in a range of non-fatal injuries, including blunt trauma, puncture wounds, lacerations, sprains/strains, motor vehicle crashes, animal bites, and electrocution [15–18]. Falls, traffic accidents, and other injuries can also occur in period immediately before a major storm hits, particularly among the elderly, as people try to evacuate and prepare their residences [18].

Those who do not evacuate prior to a storm and shelter in place, by choice or necessity, may risk injury or death during a coastal storm [9]. Poor and minority populations, and elderly nursing home residents, are more likely to lack transportation during disasters [19]. These populations often have a high prevalence of chronic health problems, which increases their vulnerability to other storm-related hazards.

Exposure to natural disasters may trigger acute coronary syndromes [20]. Researchers have also observed a 3-fold increase in the incidence of acute myocardial infarction among Tulane Health Sciences Center hospital patients two years after Hurricane Katrina related to emotional stress [21].

While evacuations of health care facilities are undertaken to avoid health risks that result from major destruction such as flooding, power outages, and interruptions in medical care if health facilities are rendered inoperable, safe evacuations require advance planning and technical expertise. In the evacuation phase, frail or medically incapacitated people may need assistance getting to trains or buses or may require other modes of transport. For municipal health and transit staff, as well as private agencies serving this population, this represents a large undertaking, requiring organization, training, and adequate staffing, which could be difficult to ensure during an emergency. Furthermore, receiving facilities located outside the storm surge zone will require additional supplies and patient care capacity. Delayed (post-storm) evacuation of health care facilities can compound logistical problems and risks because of loss of power and damage to communications and transportation infrastructure, making it more difficult to transport and track patients who may already be compromised by failure of medical equipment or exposure to heat or cold [22].

A survey of twenty Gulf state nursing homes identified significant logistical and response problems during Hurricane Katrina related to transportation, staffing, maintaining complex medication regimens, insufficient emergency provisions, and failure to follow emergency plans [23]. All facilities encountered problems, but there were slightly more negative effects (including dehydration, depression, and skin tears) reported by administrators from facilities that evacuated rather than sheltered in place. Nursing home administrators reported little support from state and local emergency responders in evacuation decisions and implementation [23].

In NYC, five hospitals shut down due to Sandy. Three of them—NYU Langone Medical Center, Bellevue Hospital, and Coney Island Hospital—had to evacuate patients after the storm hit because of flood damage to critical equipment located in low-lying parts of the hospitals. Many nursing and long-term care facilities also lost power and had to evacuate patients.

Local or widespread power outages could result from flood damage to underground and low-lying electrical infrastructure, as well as damage to utility poles and aboveground wires from high winds and downed trees.

Following widespread power outages, carbon monoxide (CO) poisoning is a major health hazard. Deaths and illness can occur when portable generators, cooking appliances, and other fuel burning equipments are used indoors or improperly [24–34].

Mortality from other accidental and natural causes may also increase during power outages. A study of the August 14-15, 2003, Northeast blackout that affected NYC found increased mortality from both accidental and natural causes that resulted in approximately 90 excess deaths (an increase of 28%) [35]. Seniors aged 65–74 years were most vulnerable. Researchers theorized that increased mortality could be related to more physical demands on vulnerable people due to non-functioning elevators and subways, increased call times for ambulances, closed stores and pharmacies, and potentially the effects of increased exposure to air pollution and ambient air temperatures [35].

Power losses may lead to increased emergency medical services (EMS) calls and emergency department (ED) visits from patients who rely on electrically powered medical equipments like ventilators and oxygen [24, 27, 32, 34, 36–39]. Frail residents of nursing homes and other health care facilities that shelter in place can be especially vulnerable to power outages in facilities without backup generators or with backups located in flood-vulnerable places such as basements. Risks include the failure of medical equipment and exposure to hot or cold ambient temperatures, possibly compounded by a lack of supplies or adequate staff in facilities that are not well prepared [22, 40]. Hospital admissions for respiratory conditions may also increase [41].

Incarcerated populations are also at potentially high risk. In New Orleans during Hurricane Katrina, a lack of emergency preparedness and planning by corrections officials led to chaos during the storm at the over-crowded Orleans Parish Prison. Prisoners locked in cells were left alone without power, food, water, or even sufficient ventilation during the storm. Prisoners described lack of access to medical care and interruptions in care for serious chronic illness during and after the storm [42]. Prison populations often have higher rates of mental and physical illness and substance use disorders than the general population and are at risk for exacerbations [43].

After Superstorm Sandy made landfall, hundreds of thousands of NYC residents initially lost power. However, even after the electric grid had been largely restored, many residential buildings in storm-inundated areas still lacked electric power, heat, or running water, often because of salt water flood damage to buildings electrical and heating systems. Many people who did not evacuate in advance of the storm sheltered in place in housing conditions that lacked one or more of these essential services. Exposure to hot or cold ambient temperatures from lack of climate control could result in heat- or cold-related illness, including heat stroke or hypothermia, as well as exacerbation of respiratory, cardiovascular, and other chronic diseases [41, 44, 45].

In the days following Sandy, health department surveillance data reflected the impact of people living without power or heat and, in some cases, trying to provide power or heat in unsafe ways. From the storm impact until November 9 (10 days), CO-related emergency department visits and Poison Center (PCC) calls for CO exposure were elevated for the time of year; PCC data frequently identified storm-related sources of exposure including charcoal grills and household cooking appliances used for heating, as well as portable generators. Counts of cold illness syndrome emergency department visits (including hypothermia) were also elevated through November 9 [46].

While the potential health impacts resulting from sheltering in inadequate housing after a storm have not yet been formally evaluated, exacerbation of chronic health conditions, including physical, mental, and substance use disorders, could occur from a disruption of care due to lack of light, telecommunications, and elevator service that makes it difficult for people to access outside care, obtain medications, maintain self-care, and receive home-based care services. There may also be stress-related exacerbation of chronic physical and mental health problems related to isolation.

Injuries could include fire risks among those using stoves for heat or candles for light [47] and risk of falls from inadequate lighting in dwellings, hallways, and stairwells [25]. Inability to access food and fresh water, particularly for people living in high-rise apartments where water delivery is dependent on electric pumps, could lead to infectious disease risk because of inability to properly wash hands or food, bathe, or flush toilets.

People living in residential buildings without electricity may be more vulnerable to foodborne disease because they cannot refrigerate food. After a 2003 summer power outage in NYC, increases in diarrheal illness caused by the consumption of spoiled foods, especially meat and seafood, were observed [48]. Furthermore, unless active steps are taken to remove spoiled food from residences and restaurants (which typically pay private contractors to collect waste) during prolonged outages, pest populations could increase [24]. It is possible that proliferation of pests, including rodents and roaches, because of difficulty with cleaning and removing trash may exacerbate allergies, asthma, and other respiratory conditions.

Older adults, young children, individuals with pre-existing health conditions, and those living on the upper floors of high-rise buildings or those who are disabled may be especially vulnerable to health effects resulting from power and utility outages. Safety concerns stemming from lighting outages and disabled safety systems in hallways and stairwells may deter this population from seeking or receiving assistance. In addition, it is possible that this population may be at risk for mental health outcomes observed in populations who have experienced long-term displacement, including post-traumatic stress and other mental health problems.

Exposure to mold in flood-damaged buildings could worsen allergic and asthmatic symptoms—among those with pre-existing allergic sensitization—and respiratory infections [49]. Although increased concentrations of outdoor and indoor mold were detected after Hurricane Katrina in areas that had experienced flooding [49, 50], the impact of the mold on health following Katrina has not been well characterized, possibly because of under-reporting or under-detection of health problems, or by population displacement that reduced exposure [49].

Coastal storms can cause release of untreated sewage through direct damage and flooding of treatment facilities, power outage, or combined sewage overflows (CSO). CSOs are caused when heavy rains overwhelm combined systems of collecting storm and precipitation water runoff causing the discharge of untreated sewage into rivers. Secondary exposure to sewage-contaminated floodwaters and wastewater, along with impaired access to potable water and flushable toilets, may lead to gastrointestinal infections, acute respiratory infections, skin infections, and insect bites [15, 16, 51, 52]. There may also be a risk of increases in vectorborne diseases like West Nile virus because sites where water collects after heavy rains could become breeding grounds for mosquitoes.

Rainfall, wind, and runoff in the watershed area can contribute to high turbidity levels, which can interfere with the disinfection process of drinking water. In NYC, a coastal storm will not necessarily impact the drinking water supply because the watershed and reservoirs are located up to 125 miles from the city. However, an increasing number of heavy precipitation events in the watershed may lead to more instances of high turbidity levels. The NYC Department of Environmental Protection monitors turbidity at over 1,400 locations in the watershed and distribution system and activates the Ashokan Waste Channel to reduce turbidity levels [53].

Storm damage may compromise sites storing toxic waste, and flood waters may move hazardous substances to new areas. Following Hurricane Katrina, hazardous substances such as volatile organic compounds (VOCs), lead, and arsenic were detected in the air, soil, and sediment samples. No health effects have been directly observed in a storm-specific context [54–59]. However, the potential for a toxic release of hazardous substances after a storm exists [55–57]. Following Sandy, initial testing of two Superfund sites indicated that contact with contaminated water from these areas was not a major health threat [60].

Displacement led to a host of adverse effects following Hurricane Katrina and major storms in other areas. Health risks were often related to living in congregate shelters, disruption of access to health care, or some combination of both.

The spread of infectious diseases, such as norovirus, was documented among residents of temporary shelters or evacuation centers in the wake of Hurricane Katrina [61]. Waterborne illnesses and infectious disease spread in shelters have been more frequently observed among young children and infants with naïve immune systems [61, 62].

Displacement can lead to interruption in medical care and exacerbation of chronic health conditions. Contributing factors to disruption of care after Katrina, Hurricane Ike, and other natural disasters included evacuees without medical history information, medications or knowledge of medication names and doses, and access to medical records [63, 64]. Following Hurricane Katrina, health care providers and focus groups also reported that chronic disease treatment interruptions—especially among patients with cancer, hypertension, end stage renal disease, cardiovascular disease, and respiratory illnesses—were problems [63]. Dialysis sessions may be missed [65]. In a sample survey of Louisiana Red Cross shelter residents in the weeks immediately following Katrina, slightly more than half of the residents had chronic medical conditions including hypertension, hypercholesterolemia, diabetes, and lung diseases [64]. About a third needed acute medical attention. Shelter residents may have pre-existing mental health conditions that require ongoing care [64].

After Hurricane Katrina, there was also a need for urgent medical supplies, such as oxygen tanks [63]. Emergency responders may also need access to sufficient quantities of vaccines (and facilities with proper storage capabilities), medications, and other preventive medical supplies for displaced populations.

Women who are displaced may have trouble accessing contraceptives and other reproductive health services. Following Hurricane Ike in Texas, racial disparities in access to contraceptives were also observed [66]. Disruptions in prenatal care for pregnant women, including folate supplementation, could potentially lead to adverse birth outcomes [67, 68].

Displacement can also result in other physical health effects. Displacement at 12 months following Katrina was associated with increased risk of hip fracture among seniors, particularly among women, those with co-morbidities and a history of hip fracture [69]. Among nursing home residents in the Gulf Coast evacuated before a recent hurricane, 30- and 90-day mortality and hospitalization rates were higher compared with rates during non-hurricane control years [10].

Experiencing a hurricane or major natural disaster may exacerbate existing mental health conditions or contribute to new mental health and interpersonal problems [70–74]. Particularly in the months immediately following exposure to a natural disaster, increases in levels of post-traumatic stress disorder (PTSD), as well as other mental health problems have been observed [72]. More than a year after Hurricane Katrina, anxiety and mood disorders in the New Orleans metro area were substantially elevated, and mental health conditions were broadly distributed in the population [75]. Serious mental illness was typically accompanied by PTSD, and important predictors of mental health problems were storm-related physical illness or injury, physical adversity, and property loss. In addition, two years after Katrina, the prevalence of self-reported psychological and physical intimate partner violence increased among Mississippi residents affected by the hurricane [76]. Self-reported poor physical and mental health before and after a storm has also been correlated with self-reported poor mental health after the storm [77].

The duration of mental health problems may depend on the nature of exposure to the storm and on ongoing stressors related to the storm. One study of mental health conditions after Hurricane Ike in 2008 showed that prevalence of storm-related PTSD decreased within 18 months [72]. However, elevated levels of PTSD and psychological distress among vulnerable populations have also been observed up to five years after a hurricane [83]. Following Hurricane Katrina, researchers have suggested that slow government responses may have exacerbated mental health problems and argued that an efficient emergency response can also help to minimize the mental health impacts of natural disasters [75].

For those who have experienced displacement, short- and long-term mental health effects are the most commonly cited storms-related health outcomes in the literature. Evacuees at the Red Cross Shelter in Austin, TX, USA, following Katrina, were at increased risk of short-term acute stress disorder, while populations who were displaced or who experienced or witnessed traumatic events were at increased risk of long-term mental health effects, including PTSD, depression, anxiety, and suicidal ideation [70, 77, 78, 81, 82, 84, 85]. Women, African-Americans, and those with prior psychiatric history, poor physical health, and weak social networks have been identified as particularly vulnerable [75, 78, 79, 81, 82, 85]. Katrina evacuees living in Houston were also found to be at risk for increased substance use [80, 86].

Many studies on mental health conditions were cross-sectional, and pre-storm depression and PTSD levels could not always be ascertained, thereby limiting conclusions that could be drawn from the data about the cause-effect relationship between storms and subsequent mental health outcomes. Nevertheless, preventing the long-term mental health effects following a storm through ongoing mental health surveillance, appropriate intervention, and adaptation strategies should remain a priority.

The clean-up and recovery period after a major storm may also present significant health risks. Many occupational fatalities have occurred during post-storm clean-up and reconstruction. One study found that, at median, occupational deaths occurred 36.5 days after a storm event [7] and were most often associated with clean-up (44%), restorative construction (26%), public utilities restoration (8%), and security/policing (6%). Residents and volunteers trying to clean affected homes could suffer non-fatal injuries (including cuts, wounds, sprains, and strains) and other health risks during the course of removing debris or during minor and major home repairs (for instance, removing wet building materials or wet wall insulation). Indoor dust created during cleaning, exposure to mold, fumes from temporary heating sources, and the use of strong cleaning products can also irritate the eyes, throat, and lungs.

The delivery of and access to health care and other basic services—as well as efforts to respond in storm- or flood-damaged areas—could be impeded if workers and volunteers have difficulty traveling to and within the areas. Large-scale displacement could increase demands on the transportation infrastructure, and there may be occupational health concerns for municipal workers deployed to address flooding in transportation and communications infrastructure [7].

During recovery, air quality may be negatively affected by dust from the home clean-up, debris movement, emissions from truck traffic, and the use of outdoor temporary boilers and emergency generators. Following Sandy, routine monitoring at rooftop air monitors in New York City showed that the levels of fine particles (PM_2.5)—the pollutant most likely to be associated with combustion of fuels, dust from streets, and debris—were not significantly elevated compared to levels typical for the season. Similarly, two-week average PM_2.5 concentrations at street level monitors near storm- impacted areas in the four weeks after the storm showed that levels were typical for the time of year and similar to those elsewhere in the city. Asbestos was not detected in the dust [87]. In early December 2012, supplemental continuous air monitors were placed in flood-impacted neighborhoods with ongoing recovery operations. For the most part, 24-hour average PM concentrations in these locations through the winter tracked with levels at regional monitors elsewhere in the city, although on several days concentrations were somewhat higher near areas with concentrations of temporary generators and boilers or reconstruction activity [88].