The trauma signature of 2016 Hurricane Matthew and the psychosocial impact on Haiti

ABSTRACT

Background. Hurricane Matthew was the most powerful tropical cyclone of the 2016 Atlantic Basin season, bringing severe impacts to multiple nations including direct landfalls in Cuba, Haiti, Bahamas, and the United States. However, Haiti experienced the greatest loss of life and population disruption.

Methods. An established trauma signature (TSIG) methodology was used to examine the psychological consequences of Hurricane Matthew in relation to the distinguishing features of this event. TSIG analyses described the exposures of Haitian citizens to the unique constellation of hazards associated with this tropical cyclone. A hazard profile, a matrix of psychological stressors, and a “trauma signature” summary for the affected population of Haiti - in terms of exposures to hazard, loss, and change - were created specifically for this natural ecological disaster.

Results. Hazard characteristics of this event included: deluging rains that triggered mudslides along steep, deforested terrain; battering hurricane winds (Category 4 winds in the “eye-wall” at landfall) that dismantled the built environment and launched projectile debris; flooding “storm surge” that moved ashore and submerged villages on the Tiburon peninsula; and pummeling wave action that destroyed infrastructure along the coastline. Many coastal residents were left defenseless to face the ravages of the storm. Hurricane Matthew's slow forward progress as it remained over super-heated ocean waters added to the duration and degree of the devastation. Added to the havoc of the storm itself, the risks for infectious disease spread, particularly in relation to ongoing epidemics of cholera and Zika, were exacerbated.

Conclusions. Hurricane Matthew was a ferocious tropical cyclone whose meteorological characteristics amplified the system's destructive force during the storm's encounter with Haiti, leading to significant mortality, injury, and psychological trauma.

KEYWORDS:

• hurricane
• Hurricane Matthew
• natural disaster
• trauma signature analysis
• tropical cyclone
• TSIG

Introduction

Exposure of populations to natural disasters is known to elevate risks for psychological distress and common mental disorders (CMDs), including post-traumatic stress disorder (PTSD), major depression, generalized anxiety disorder (GAD), panic disorder, and alcohol dependence and abuse.^1-6 According to Pietrzak et al.,⁷ for studies that have examined longitudinal data, symptoms and prevalence of CMDs tend to be higher in the early aftermath of a natural disaster and to decline over time, as confirmed on follow-up assessments.^1-6 Studies have documented differential constellations of CMD symptom elevations depending upon that nature of the disaster, prompting Pietrzak and colleagues⁷ to recommend assessment of a spectrum of CMDs, not just PTSD and depression.

The present analysis will explore mental health and psychosocial outcomes for the population of Haiti following Hurricane Matthew's direct impact during the first week of October, 2016. Guidance will be sought from the expanding literature of published studies^8-14 that have examined the range of psychological sequelae for survivors of tropical cyclones. Tropical cyclones are powerful, high-velocity windstorms that are variously named hurricanes, typhoons, and cyclones, depending upon the oceanic “basins” where they form around the globe.¹⁵

Hurricane Matthew was the most powerful tropical cyclone of the 2016 Atlantic Hurricane season, affecting multiple island nations as well as coastal South America and the United States. The 13-day trajectory of destruction associated with the storm encompassed an immense geographic swath, beginning in the Lesser Antilles and extending from the tropical latitudes of Venezuela and Colombia, South America, to the Canadian Atlantic Maritimes. The central core and eyewall of Matthew made sequential landfalls, first passing over Haiti's Tiburon Peninsula and moving across the extreme eastern tip of Cuba, before traversing the entire expanse of the Bahamas (Lucayan Archipelago), hugging the Florida coastline, and coming ashore several days later in South Carolina, before sending extra-tropical remnants moving northward toward Canada.

Disaster ecology model and trauma signature (TSIG) analysis

In the present study, an established methodology of trauma signature (TSIG) analysis¹⁶ was applied to this wide-ranging natural disaster event. TSIG analyses have been conducted across a spectrum of disasters, both natural and anthropogenic.^17-23 In the depiction of Hurricane Matthew, TSIG analyses are applied to illuminate the manner in which the physical forces of harm associated with this tropical cyclone were tranmuted into psychological outcomes.

Methods

Trauma signature (TSIG) analysis^16-23 was applied to examine the psychological and mental health effects of Hurricane Matthew. TSIG analysis is defined as “an evidence-based method that examines the interrelationship between population exposure to a disaster, extreme event, or complex emergency; and the inter-related physical and psychological consequences; for the purpose of providing timely, actionable guidance for effective disaster behavioral health (DBH) support that is tailored to the defining features of the event.”¹⁶

TSIG analysis looks at disaster survivors' exposures to empirically-documented risk factors for psychological distress and CMDs.^1-6 Consistent with the Disaster Ecology Model,^24-26 which provides the theory base for TSIG analysis, the assumption is that each disaster exposes the affected population to a novel constellation of hazards, losses, and life changes. Experiencing a disaster may be described as a potentially traumatizing exposure (PTE). Disasters expose citizens to “forces of harm” at both the individual and the collective, population-wide levels.

Once the hazards, losses, and life changes are fully elucidated, a disaster's unique “signature” of exposure risks may serve as a key predictor of needs for mental health and psychosocial support (MHPSS).¹⁶ When TSIG analysis is performed in “real time,” while a disaster is unfolding, or in the short term aftermath, results can be used to guide the on-scene MHPSS response.¹⁶

The TSIG analyses reported here entailed the following steps:

• Step 1: Overview of the Haitian context of disasters. Medical anthropologists on the author team with extensive research experience in Haiti, and onsite response to the hurricane, provided the description of the Haitian context.

• Step 2: Creation of a synopsis of Hurricane Matthew's impact on Haiti. A narrative description of the disaster event was developed as a segue to the creation of the hazard profile of the event.

• Step 3: Development of a hazard profile. A TSIG hazard profile uses an epidemiologic approach to disaster description that incorporates hazard, person, place, and time dimensions. Classification of the type of disaster was adapted from several available taxonomies.^27,28

• Step 4: Review of the literature on psychological risk factors for tropical cyclones: hazards, losses, and life changes. The disaster mental health literature on human population exposure to tropical cyclones was searched to identify evidence-based risk factors. A primary focus was placed on recent storms that generated literature on this topic, particularly Atlantic Basin hurricanes Katrina (2005), Ike (2008), and Sandy (2012), and Northwest Pacific Supertyphoons such as Haiyan (2013).

• Step 5: Description of the Haiti situation in relation to hurricane hazards, losses, and life changes.

• Step 6: Creation of an exposure and stressor matrix. Based on review of the scientific literature and news accounts, a table was constructed of disaster stressors experienced by residents of Haiti during this nation's encounter with Hurricane Matthew. A “stressor matrix” was developed that presents exposures to 3 types of psychological risk factors (hazards, losses, life changes) cross-classified with 3 disaster phases (pre-impact, impact, post-impact).

• Step 7: Creation of trauma signature summary. A composite trauma signature summary table was constructed, displaying the most significant evidence-based psychological risk factors, grouped under the headings of hazard, loss, and change. The trauma signature summary table presents exposure severity ratings for the risk factors. The ratings use order-of-magnitude (10-fold) differences between adjacent categories. Based partially on the international disaster database maintained by the Centre for Research on the Epidemiology of Disasters (CRED),²⁷ dating from 1900 to the present, ratings of “Extreme” for a specific risk factor are reserved for those rare disasters that produce consequences at that order of magnitude only several times every 50 to 100 y. “Very Severe” ratings reflect the order of magnitude threshold reached with a frequency of 1 or several times within a 10-to-20 y period, while “Severe” ratings occur 1 or several times within a 3-to-5 y period.

• Step 8: Enumeration of the distinguishing features of Hurricane Matthew. As an opportunity to summarize key findings, co-authors developed a listing of the unique and defining features that set this disaster apart.

Results

TSIG Step 1: Overview of the Haitian context of disasters

Haiti has experienced natural disasters almost every year over the past 4 decades. Compared to other nations in the Caribbean region, Haiti “has a higher number of disasters per km².”²⁹ Natural disasters coupled with slow economic growth only aggravate the condition of poverty faced by many in the island nation. By 2012, approximately 8.8 million Haitians were living in poverty and most were unable to meet their basic food needs.²⁹ This phenomenon of extreme poverty disproportionately affects Haitians living in rural areas such as those directly affected by Hurricane Matthew in the Sud and Grand'Anse regions of southern Haiti.^29,30

Limited access to safety nets render the average Haitian vulnerable to the damage, and disruptions of daily life, caused by natural disasters. Annual shocks such as displacement, educational disruption, and changes in nutritional profiles caused by hurricanes, flooding, and the 2010 earthquake, affect three-quarters of all households.²⁹ The impact of Hurricane Matthew on the southern region of Haiti is set to continue this trend with total estimated damages of USD $1.89 billion.³¹

TSIG Step 2: Creation of a synopsis of Hurricane Matthew's impact on Haiti

On September 22, 2016, a mass of thunderstorms moved rapidly westward off the coast of Africa on an uncharacteristic southerly arc, attracting the immediate attention of atmospheric scientists. By September 25, the National Hurricane Center designated this fast-paced wave as an area of interest for investigation - INVEST 97L. At 11 AM, September 28, just before passing over the Lesser Antilles in the vicinity of St. Lucia, this system gained tropical cyclone status and was christened “Matthew” with winds of 60 mph, an unusually high initial wind speed. Moving into the Caribbean, the system encountered super-heated waters, a vestige of the record-setting 2015/2016 El Niño that was just waning, coupled with 2016's trajectory toward becoming the hottest year on record. Intensification was brisk; Matthew became a major hurricane on September 29 and reached Category 5 wind speeds one day later, on September 30. Matthew's forward motion stalled for several days as the system meandered in the Caribbean south of Hispaniola (Fig. 1). Then northwesterly movement advanced the storm with accelerating speed as it made landfalls in Haiti on October 4 and eastern Cuba on October 5. Matthew traversed the Bahamas throughout the days of October 5 and 6, and snaked along the entire length of Florida's Atlantic coastline on October 6 and 7. The storm began to weaken as it moved close to northern Florida and finally made landfall in South Carolina on October 8 as a Category 1 hurricane. Matthew reentered the Atlantic as a post-tropical system, moving along the North Carolina coastline on October 9, with remnants moving northward toward Canadian latitudes.

Figure 1. NASA satellite image of Hurricane Matthew at Category 4 strength near the north coast of South America and south of Haiti on Oct 2, 2016. Photo by NASA. Source: https://worldview.earthdata.nasa.gov/?p=geographic&l=VIIRS_SNPP_CorrectedReflectance_TrueColor(hidden),MODIS_Aqua_CorrectedReflectance_TrueColor(hidden),MODIS_Terra_CorrectedReflectance_TrueColor,Reference_Labels(hidden),Reference_Features(hidden),Coastlines&t=2016-10-02&z=3&v=-97.34765625,1.44140625,-52.34765625,23.58984375

Haiti situation

Hurricane Matthew was the most impactful tropical cyclone during the 2016 Atlantic Hurricane season. Effects were felt in multiple small island developing states along the Windward Islands, the northern coast of South America (primarily Colombia and the ABC islands), Jamaica, Haiti, Dominican Republic, Lucayan Archipelago, and the southeastern United States.

However, the most extensive physical damage and human harm occurred in Haiti, particularly in the Tiburon Peninsula where Matthew made direct landfall. Matthew drifted in the Caribbean for the first days of October prior to making its northward-moving assault on Haiti (Fig. 2). During that time when the near-stationary storm was positioned south and slightly west of Haiti, towering thunderstorm bands pounded the nation with drenching rainfalls.

Figure 2. NASA satellite image of Hurricane Matthew as Category 4 strength directly over Haiti on Oct 4, 2016. Photo by NASA. Source: https://www.nasa.gov/feature/goddard/2016/matthew-atlantic-ocean

An unusual feature of the storm was the “blob” of thunderstorms that remained to the east of the center of circulation for several days, producing heavy downpours over much of Haiti, well in advance of the storm winds that finally arrived on October 4 (Fig. 3).³² Haiti was subjected to days of deluging precipitation with rainfall totals exceeding 30 inches in some locales. With rugged, steep, deforested terrain, Haiti was vulnerable to flooding and poised for mudslides, creating the hazards that accounted for much of the mortality. Portions of western Haiti were subjected to Category 4 hurricane winds near the eyewall when the storm shifted from static motion south of Hispaniola to northward progression, making landfall on October 4.

Figure 3. Water vapor image of Hurricane Matthew, Monday evening, Oct. 3, 2016. Source: NOAA Geostationary Satellite Server. http://www.goes.noaa.gov/

In Haiti, a total of 546 persons were confirmed dead and 438 were injured based on official government statistics compiled through October 31, 2016.³³ More than 2.1 million persons, equivalent to one-fifth of the national population, were affected and 894,000 (42.6%) were children. An estimated 1.4 million persons, equivalent to 66.7% of those affected, were identified as needing assistance during the post-impact period. Furthermore, approximately 141,000 persons were internally displaced due to the hurricane. Storm-associated morbidity, mortality, and population displacement were especially acute in the departments (states) of Grand'Anse and Sud (Fig. 4).

Figure 4. Aerial Damage Survey: Jeremie-Haiti. © Pacific Disaster Center. Reproduced by permission of Pacific Disaster Center. Permission to reuse must be obtained from the rightsholder. Source: http://www.pdc.org/event-products/preview/?id=1321

Food insecurity was a dominant feature of the post-impact landscape during the first month following Matthew's impact; the United Nations Office for the Coordination of Humanitarian Assistance (OCHA) estimated that 800,000 individuals had an urgent need for life-sustaining food, including 69% of households in Grand'Anse and Sud.³³ OCHA also indicated that “negative coping strategies” were prevalent (85% of households).³³

Adding to the complexity of the response and worsening the hardship, Matthew struck during a time when cholera was spreading in epidemic fashion and 3,500 new suspected cholera cases were diagnosed during the first 4 weeks following hurricane landfall.

TSIG Step 3: Development of a hazard profile (Table 1)

Table 1. Hazard profile: 2016 Hurricane Matthew in Haiti.

Disaster Type	Initial Event Classification	Natural Disaster:
		• Meteorological: Category 4 Hurricane
		• Hydrological: Surge, floods, mudslides
	Ongoing Event Classification	Complex Emergency:
		Natural disaster followed by large-scale displacement, infrastructure damage, and epidemic cholera
	Forces of Harm	Wind hazards:
		• Hurricane force winds
		• Tropical storm force winds
		• Tornadoes
		Water hazards:
		• Deluging rains
		• Coastal surge
		• Storm surge
		• High tides
		• Mudslides
		• Inland flooding
Magnitude and Severity	Hurricane winds	Category 4
	Saffir-Simpson Hurricane Wind Scale (SSHWS)
	Precipitation (rainfall)	Up to 30+ inches
		Duration: 3 d
		Area: Entire nation
Place Dimension	Hurricane landfall	Tiburon Peninsula close to Jeremie
	Rainfall	Precipitation throughout entire country (27,750 km^2)
	Geography/Geology	Steep, deforested mountainous terrain
		Low-lying peninsula
	Built Environment	Poorest nation in Western Hemisphere
		• Human Development Index: 163/188
		• 55 percent below extreme poverty index
		• Widespread structural deficiencies
Time Dimension	Landfall	October 4, 2016
	Duration	Landfall and eyewall encounter: 8 hours
		Storm system interaction effects: ∼4 days
	Health risk: duration	Months
	Post-storm disruption: duration	Years

Tropical cyclones are closed, rotating circulations of thunderstorms around a well-defined center that originate over warm tropical or subtropical waters. The system is sustained by the extraction of heat energy from warm ocean waters.¹⁵ The principal hazards associated with a tropical cyclone are population exposures to wind and water.¹⁵

Wind-related hazards include the powerful circulating winds of the rotating thunderstorms that comprise the cyclone itself. On the right side of the forward-moving storm, precisely where Haiti was situated as the storm approached, the speed of forward motion of the system is added to the speed of the spiraling winds. The strongest winds are found in the cyclone's eyewall, an organized band or ring of cumulonimbus clouds that surround the calm and windless eye. Matthew was a Category 4 storm when it struck Haiti on October 4 (based on the Saffir-Simpson Hurricane Wind Scale with ratings ranging from 1 to 5). Adding to the ferocity of the hurricane winds, tornadoes may also be spawned, particularly in the forward right quadrant of the hurricane.

In the case of Hurricane Matthew, the water-related hazards were geographically more widespread and associated with more deaths and injuries than the wind hazards. Water hazards included deluging rains that triggered mudslides, coastal surge due to wave action as the storm approached from the south, storm surge directly beneath the eye of the storm that pulled a dome of water onshore, and overland flooding. Although hurricanes are typically presented as windstorms – meteorological events – the water-related hazards frequently account for the majority of deaths and a significant proportion of damage and destruction. In the case of Matthew impacting Haiti, this disaster involved both prominent meteorological and hydrological features. Overland flooding that occured onshore and inland, based on a combination of rain, surge, and river overflow, added to the human harm and property destruction.

TSIG Step 4: Review of the literature on psychological risk factors for tropical cyclones: Hazards, losses, and life changes

The research literature that examines the linkages between exposure to tropical cyclones and mental health consequences is steadily expanding. Psychosocial research is limited to studies conducted on the small subset of major hurricanes that have caused significant harm to human populations, infrastructure, and the built environment. Much of the psychological literature derives from Atlantic Basin hurricanes that have made landfall in the United States, including 1992 Hurricane Andrew, 2004 Florida hurricanes (Charlie, Frances, Ivan, Jeanne), 2005 Hurricane Katrina, 2008 Hurricane Ike, and 2012 Hurricane (“Superstorm”) Sandy.

The studies from Hurricane Ike provide a particularly sophisticated analysis of the psychological risks for populations exposed to tropical cyclones.^7,34-48 Multiple research teams conducted a coordinated set of studies that together elucidate the interconnections between hurricane stressors and psychological outcomes. Several of these studies are reviewed here to provide an overview.

Pietrzak and colleagues⁷ evaluated symptom levels for multiple CMDs, including PTSD, major depression, GAD, panic disorder (PD), and suicidality for survivors of 2008 Hurricane Ike. Their study was distinguished by early and repeated assessment of a carefully-selected sample that accurately represented the pre-disaster resident population of the storm-impacted counties. Assessments were conducted 2–5 months, 5–9 months, and 14–18 months post-hurricane. Across the 3 measurement points, past-month prevalence of any CMD, and hurricane-related PTSD, declined sharply. In contrast, non-disaster-related PTSD, depression, suicidality, PD, and GAD prevalence rates remained relatively unchanged over time.

To be classified as “hurricane Ike-related PTSD,” study participants needed to meet all 6 DSM-IV-TR criteria for PTSD,⁴⁹ specifically reporting: terror or helplessness during the event; being bothered “moderately” or more by at least one re-experiencing symptom (e.g., repeated, disturbing dreams of Hurricane Ike), at least 3 avoidance symptoms (e.g., avoiding activities or situations that remind the respondent of Hurricane Ike), and at least 2 increased arousal symptoms (e.g., feeling jumpy or easily startled); symptoms persisting for at least one month duration; and significant impairment or distress resulting from symptoms.⁷

Tracy and colleagues (2011)⁴⁶ carefully dissected the types of exposures to hazards and hardships that contributed to PTSD and depression in survivors of 2008 Hurricane Ike. These authors differentiated PTSD as “a disorder of event exposure.” Persons most likely to develop hurricane-related PTSD were those who had been “physically present” and directly experienced the hurricane winds and flooding. Persons who presented with symptoms of PTSD frequently reported potentially traumatic events such as real or perceived life threat for self or others, sustaining physical injury, or losing a family member or close friend who was killed in the hurricane. Authors described “hurricane-related stressors” as displacement from home for one week or longer, serious damage to home or property, loss of sentimental possessions, or significant financial loss. Their findings corroborated those of earlier investigations with survivors of Hurricane Katrina.^50,51

Tracy et al.⁴⁶ found that both direct exposure to hurricane hazards and hurricane-specific health problems were prominent predictors of PTSD. Depression was significantly related to lower educational attainment, lower household income, more hurricane-specific stressors, more lifetime stressors, and direct hurricane exposure. Loss of sentimental possessions was related to both PTSD and depression. Severe financial losses predicted depression but not PTSD. Authors explained that while PTSD is driven by event exposure, contributors to depression are more nuanced and include non-traumatic life stressors and “personal vulnerabilities.” The central role of stressors in fueling psychopathology is noted by the authors as a key point for study and a focus for intervention.

Hurricanes frequently disrupt public access to vital services and essential supplies such as electrical power, food, water, transportation, and communications. These services may be unavailable for prolonged periods of time, leading to physical discomfort and psychological distress. In developing nations, food insecurity may be life-threatening. Gros et al. (2012)³⁶ conducted surveys of 2008 Hurricane Ike survivors and ascertained a strong linkage between extended loss of services and symptoms of PTSD, depression, and “worry,” with higher rates of symptom elevations in non-urban settings. These findings underscore the importance of distributing survival supplies and reactivating essential services as rapidly as possible post-disaster.

Ruggiero and colleagues (2012)⁴⁵ conducted structured telephone interviews with 255 survivors of Hurricane Ike; 3-in-4 had evacuated their home neighborhoods and 40% remained displaced from their homes for one week or longer. The interview screening instruments assessed symptoms for PTSD, major depression, and GAD. Hurricane exposure severity was measured as a tally of usual resources (e.g., electrical power, food, water, cash, transportation, clean clothing, etc.) that became unavailable for periods of one week or more due to Hurricane Ike. CMD symptoms were associated with resource disruptions and especially with more pronounced and longer-duration losses such as loss of home and severe financial impact. Authors highlighted the importance of replenishing basic needs, restoring resources, and reestablishing essential services during the disaster recovery period.

Cerdá and her research team (2013)³⁴ examined the role of ongoing exposure of hurricane survivors to potentially traumatic events and stressors during the aftermath in relation to “post-disaster mental health.” Specifically, investigators compared the effects of acute hurricane exposures, experienced during the impact phase, to stressors continuing into the recovery period, on posttraumatic stress symptoms (PTSS) and functional impairment (FI) for 658 Hurricane Ike survivors during a series of 3 sequential interviews. Because of the direct applicability to the experience of Hurricane Matthew survivors in Haiti, the next sections further explicate the major findings from Cerda's study of Hurricane Ike victims.

Hurricane-related exposures and stressors measured on the initial survey

As a point of reference, acute exposures included traumatic events or severe losses experienced during Hurricane Ike (e.g., storm-related physical injury, death of a family member or close friend, witnessing dead bodies, assaultive violence around the time of the hurricane).

Hurricane-related stressors – at the time of the hurricane – included disruption of essential resources (shelter, food, water, electrical power) for more than one week, loss or severe damage to personal property (home, possessions, appliances, vehicles), loss or damage to “sentimental possessions,” serious hurricane-related health problem for self or household member, and/or “experiencing significantly increased demands after the hurricane.” Also assessed were the respondent's emotional reaction and “fear severity” during and immediately after the hurricane. This included self-report of sensations and behaviors such as shortness of breath, trembling, shaking, or buckling knees.

Post-hurricane exposures and stressors

Post-hurricane exposures also included traumatic events such as robbery, mugging, or sexual assault. Hurricane-related ongoing stressors that were assessed at each of the 3 interviews focused on displacement from home for one week or longer, job loss, or decreased income due to the hurricane. Post-hurricane life stressors included relationship break-up, divorce, legal problems, or serious financial crises.

Analyses examined the exposures and stressors individually and in the composite. Researchers were able to confirm their hypotheses that exposure to hurricane impact stressors predicted PTSS and FI during the initial assessment and also more significant FI increases over time. Post-hurricane daily stressors were not associated with PTSS or FI at the time of first interview but predicted elevations on both measures at interviews 2 and 3.

Cerdá's team of investigators conclude that interventions in the immediate aftermath may be beneficial for mental health in the short run but “prevention of ongoing stressors” is an essential component of intervention to optimize mental health for the longer term.³⁴

Members of the same team conducted additional subanalyses of the cohort over 3 measurement points (Waves 1,2,3). Lowe and Tracy (2013)⁴³ examined longitudinal patterns of hurricane stressors in relation to ongoing stress and psychopathology. They found that hurricane survivors who reported more stressors during and immediately following hurricane Ike (immediate stressors) on the Wave 1 assessment were more likely to experience stressors, PTSD, and depression symptoms at Waves 2 and 3. Investigators also showed that longer-term stressors, persisting over multiple measurement points, predicted continuing stress, distress, and symptoms of disorders.

Ongoing analyses of Hurricane Ike survivors provide insights into special populations and diverse contributors to both psychological consequences on the one hand, and also to psychological recovery, wellness, and resilience, on the other. For example, Hirth et al. (2013)³⁹ conducted a cross-sectional survey of young women, ages 16–24 years, who were affected by Hurricane Ike. Participants were asked about a series of hurricane exposures and consequences including personal injury, injury or death of family members or friends, hurricane-related damage, and job loss, in relation to symptoms of PTSD and depression. All of these exposures predicted increased hurricane-associated PTSD symptoms, while job loss and hurricane damage were related to increased depression symptoms.

Wu and colleagues (2015)⁴⁸ were able to examine perceived stress and sleep quality data, collected on an ongoing basis, for women clients who were receiving services from Texas family planning clinics and who later experienced Hurricane Ike. This allowed a pre-hurricane to post-hurricane comparison of risk factors for higher levels of perceived stress. Researchers found that perceived stress was higher among poor sleepers than good sleepers throughout the period from 6 months pre-Ike to 12 months post-Ike. Stress increased for poor sleepers with increasing levels of Hurricane Ike storm damage, but not for good sleepers. Only good sleepers experienced decreased stress levels with increasing time since the hurricane. Higher stress levels in poor sleepers were associated with Ike-related stressors compounded with major life events and emotional dysregulation. Among good sleepers, higher stress levels were related to lower levels of self-control and history of psychiatric diagnosis. Authors conclude that sleep quality influences perceived stress in survivors of a major natural disaster and therefore, disaster survivors should be provided with interventions to improve sleep quality and diminish stress.

West and team (2013)⁴⁷ found that community support moderated the effects of disaster exposure on PTSD and depression symptoms in nonurban settings when assessed in Hurricane Ike survivors 12–17 months after the storm.

While acknowledging that natural disasters produce adverse psychological outcomes in significant minority of survivors, “resilience” is the modal response. Lowe et al. ⁴⁰ examined resilience (defined as low levels of psychological symptoms and time-limited functional impairment in the near-term aftermath of trauma exposure, with restoration of function thereafter), mental health wellness (resilience across mental health conditions), and general wellness (resilience across mental health, physical health, and role functioning) in Hurricane Ike survivors. Predictors of mental health and general wellness included lower peri-event emotional reactions and higher collective efficacy in the respondent's community. Three types of losses – losses of financial resources, sentimental possessions, and pets – negatively affected mental health wellness while loss of personal property was associated with diminished general wellness. Authors suggest that understanding this complex interplay of factors is important for meeting the psychosocial needs of disaster-affected individuals and their communities.

Two related publications by Gruebner et al.^37,38 demonstrated that it is possible to detect geospatial patterning of mental health and wellness, including “resilience trajectories” for areas that have been impacted by a hurricane or other natural or anthropogenic disaster. In fact, it is possible to examine geographic distributions of persons with specific symptom elevations for depression, PTSD, and other CMDs. These advanced analytics map the disaster-affected community in relation to the severity of physical damage, resource losses, and socio-demographic indicators, in order to predict the MHPSS services needed for each locale.

Of particular importance from the Hurricane Ike studies was the ability of the research groups to isolate the effects of direct hurricane exposure for residents who remained on Galveston Island and did not evacuate, thereby experiencing both extreme trauma and loss, compared to their counterparts who did evacuate to the mainland. Evacuees from Galveston Island also experienced near-total property destruction and loss but without the direct trauma exposure. Consistent with the description of PTSD as a disorder of “event exposure,” those who stayed on the island and experienced the full force of Hurricane Ike had much higher rates of PTSD. Depression was prominent for both groups due to losses combined with post-event hardships.

These findings from Hurricane Ike provide the context for the likely experiences and responses of the survivors of Hurricane Matthew in Haiti. The literature is further enriched by studies conducted with survivors of other tropical cyclones occurring in the Atlantic Basin and worldwide. We have previously written about mental health issues related to Hurricane (“Superstorm”) Sandy^52,53 and psychological research on that storm is particularly informative.^54-69

TSIG Step 5: Description of the Haiti situation in relation to hurricane hazards, losses, and life changes

Exposure to hazards

During the approach and landfall of Hurricane Matthew in Haiti, citizens throughout the nation were at risk for heavy downpours, flooding, and mudslides. In western Haiti, these risks were compounded by the strongest storm winds and the most powerful storm surges to impact the nation in recent history.

Hazards: Haitian context

Haiti's vulnerability to hazards is not just a product of geography and geology but can also be understood as a result of weak governance, poor and limited infrastructure, economic stagnation resulting in limited investments, and poor educational access and quality. Haiti is routinely exposed to hazards, as evidenced by the 2010 earthquake disaster that struck the Port-au-Prince metropolitan area, claiming the lives of over 220,000 persons and displacing 1.3 million Haitians.^70,71 While the extent of the harm caused by Hurricane Matthew has yet to be fully ascertained, prior experience and pre-existing conditions suggest that the storm's consequences will be complex and prolonged.

Prior to Matthew, slightly less than one-third of residents in the affected regions lived in extreme poverty, largely the result of food insecurity.²⁹ Sanitation services were extremely limited while rates of access to potable water were extremely low (1.2% and 10% in Grand'Anse and Sud, respectively).⁷² Even 3 weeks after the disaster, the Water, Sanitation and Hygiene (WASH) sector had only provided one-quarter of the targeted population with access to potable water for hygiene, drinking, and cooking.⁷³ These precarious post-impact living conditions are periodically exacerbated by heavy rains that may trigger life-threatening mudslides and flash floods.⁷⁴ The risks for human harm are greatly magnified by Haiti's deforestation crisis: heavy precipitation transforms barren hillsides into avalanches of mud and debris.

Prior to 2010, only 3,572 km of roads were paved, of which 10% were judged to be in good condition.⁷⁵ The road system provides access from major cities and towns to the capital city of Port-au-Prince. However, limited investment in road maintenance leads to physical isolation and separation for those regions of the nation that are farthest away from the capital such as Grand'Anse and, to a lesser extent, the Sud. Decaying infrastructure led to the collapse of several bridges during the impact phase of Hurricane Matthew, including the critical bridge connecting the capital to the southern region of Haiti.⁷⁴ This hampered the ability of the state to conduct initial damage assessments and impeded the humanitarian intervention efforts.

Exposure to losses

Losses include deaths of loved ones and close friends, leading to traumatic bereavement and complicated grief, compounded by material losses of homes and property (land, animals, crops), important and sentimental personal possesions, and livelihood-sustaining resources (including jobs, worksites, schools for children).

Losses: Haitian context

It is estimated that 2.1 million people have been affected by the disaster, including a death toll of 546 people, 438 reported injured, 132 missing, and an estimated 1.4 million in need of humanitarian assistance, including 500,000 children.³³ With 774 schools damaged or destroyed, 116,000 school children will experience educational disruption for the unforeseeable future.³³ Total damage and losses to the housing sector are estimated at just over USD $600 million, with 175,000 citizens having lost their homes.³¹ In the immediate post-impact phase, more than 140,000 storm-displaced persons were housed in about 200 temporary shelters scattered throughout the affected regions.³³ The significant loss of secure housing may trigger internal migration patterns immediately toward the towns of Jérémie and Les Cayes, and in the long-term, toward the capital.⁷⁶

In Haiti, the agricultural sector accounts for 39% of the labor market, with just under 80% of rural households engaging in farming activities.²⁹ On par with losses to the housing sector, the agriculture sector suffered significant losses to cattle, crops and fruit.^31,77 The storm's effects on the region's major crops such as coffee and cocoa will not be recovered for at least 2 y.⁷⁶ Prior to the disaster, food insecurity in rural Haiti was estimated at 34%,²⁹ and Hurricane Matthew left an estimated 800,000 people in urgent need of food assistance.³³ With estimated losses of over USD $700 million in the agriculture and environmental sectors, food insecurity is certain to worsen in the affected regions.⁷⁷

Exposure to life changes

Life changes following disasters require survivors to modify lifestyles, engage in survival behaviors, or shift from independence to dependency on services provided to maintain health and life. Population displacement following natural disasters is coupled with dislocation of social networks, loss of employment, interruption of schooling, and reliance on support from government or agencies. The psychological consequences of distressing life changes, particularly if prolonged, can be more profound than the initial exposure to trauma during the hurricane.

Life changes: Haitian context

Hurricane Matthew created challenges for the already debilitated healthcare system in Haiti. Prior to the disaster, access to healthcare was unequally distributed, with only 5.9 doctors per 10,000 inhabitants,⁷⁸ resulting in over-dependence on external sources of aid. Infant and maternal mortality rates were the highest in the Western Hemisphere, life expectancy was 62 years, and tuberculosis rates were the highest in the Americas.^78-80 As a result of the 2010 earthquake, 80 percent of Haiti's healthcare infrastructure, mostly concentrated in the greater Port-au-Prince metropolitan region, was severely damaged.⁸¹ Even before the storm, access to quality healthcare facilities and well-trained personnel was severely lacking.

Following the hurricane, the initial assessment conducted by the Pan American Health Organization (PAHO) field teams found that 36 of 56 health facilities in the region were either damaged or destroyed.⁸² Among the health concerns triggered by this disaster was the rampant spread of cholera. Dating from 2010, Haiti began to experience the largest epidemic ever recorded in a single country and the threat of cholera reemerged in 2016. Whereas 6 cases of cholera per day were reported in Sud and Grand'Anse prior to Hurricane Matthew, by October 31, 2016, there were 97 reported cases and over 3,500 suspected cases throughout the nation that had occurred following the landfall of Hurricane Matthew.³³

TSIG Step 6: Creation of an exposure and stressor matrix (Table 2)

Table 2. Disaster stressors associated 2016 Hurricane Matthew in Haiti.

Hurricane Matthew Pre-impact and Impact Phase Stressors
Exposure to Hazards	Losses	Life Changes
Direct traumatic storm exposures	Mass mortality	Change in living situation
• Exposure nationwide to deluging rains, floods, and mudslides on deforested hillsides • Exposure along the southern coast to high wave action • Exposure in Western Haiti to tropical storm and hurricane force winds • Exposure in the Tiburone Peninsula to direct landfall with maximum winds near the eyewall • Exposure to record-setting wave heights and storm surge near center of circulation • Extensive and progressive electrical power outages Personal physical harm • Physical injury • Severe pain • Entrapment Potentially traumatic exposures • Perceptions: Fear of extreme harm Fear of threat to life • Witnessing: Injury to others Death of others Mass casualties Grotesque scenes	• Death of loved ones • Death of friends • Death of community leaders Separation from loved ones • Desperate searching • Digging through debris Compounding losses • Loss of home • Loss of possessions • Loss of lands • Damage to crops • Loss of animals Community-wide destruction • Loss of utilities (power, water) • Loss of survival necessities • Loss of community services • Loss of schools • Loss of businesses • Impact on coastal industries • Impact on agricultural production • Crop destruction • Food insecurity	• Evacuation from home • Displacemen • Stranding of survivors out of reach of help • Living without shelter • Living in shelters Community impact • Massive property damage • Extensive hurricane debris • Lack of security • Lack of power • Lack of clean water • Lack of services • Gender based violence • Transportation disruption • Schools closed • Food insecurity

Based on the literature review, coupled with the actual situation reports describing Hurricane Matthew's impacts on Haiti, and considered within the Haitian context, it is possible to create an exposure and stressor matrix that considers both the encounter with hazards and the human psychosocial response to the hurricane.

The stressor matrix illustrates the complex and compounding litany of hazards, losses, and life changes that singly and collectively caused distress for the storm-affected population of Haiti. Exposure to hazards included the experience of citizens who were subjected to the powerful forces of hurricane winds, storm surge, heavy precipitation, and flooding. Hundreds sustained physical injury and many witnessed severe harm to others and destruction of the built environment around them. Of particular psychological import, many citizens near the center of impact perceived that they were going to die – or experience extreme physical harm.

Exposure to losses was widespread, including the deaths of hundreds of loved ones and friends. Some citizens were separated from loved ones and engaged in desperate searching until they were reunited. Many citizens experienced a series of losses – including severely damaged or destroyed homes, property, and possessions. For many there was a loss of critical elements related to their livelihoods and subsistence, including loss of farm animals, damage to crops, and erosion of agricultural lands. Critical community assets, including schools and worksites, were damaged and unusable for variable periods of time.

Among exposures to life changes, the experience of disaster-induced displacement was particularly salient and distressing. Added to the widespread lack of secure shelter, the affected population experienced extreme food insecurity, coupled with extended gaps in basic services such as clean water, electricity, and sanitation.

This enumeration of stressors highlights the degree of distress and psychological impact that Hurricane Matthew survivors encountered and establishes one of the bases for constructing the Hurricane Matthew TSIG summary table.

TSIG Step 7: Creation of trauma signature summary (Table 3)

Table 3. 2016 Hurricane Matthew Trauma Signature Summary.

Risk Factors for Psychological Distress and Mental Disorders	Hurricane MatthewDisaster Characteristics	Exposure Severity
Exposure to Hazards		N/A	Low	Moderate	Severe	Very Severe	Extreme
Exposure to Tropical Storm Hazards	Entire population of Haiti exposed to strong rains associated with Matthew	N/A	<10,000	>10,000	>100,000	>1 million	>10 million
Strong Fear/Panic/Life Threat	Most intensely experienced by residents of Tiburon Peninsula close to landfall;	N/A	<10,000	>10,000	>100,000	>1 million	>10 million
	1.4 million in need of humanitarian aid
Personal Physical Injury Requiring Medical Care	438 confirmed injuries	N/A	<100	>100	>1,000	>10.000	>100,000
Epidemic Infectious Disease	Increased risk for ongoing cholera outbreak; >1,000,000 need vaccination	N/A	<10,000	>10,000	>100,000	>1 million	>10 million
Entrapment or Delayed Rescue	Entrapment in collapsed structures	N/A	<100	>100	>1,000	>10.000	>100,000
Witnessing Death or Serious Injury	Most intensely experienced by residents of Tiburon Peninsula close to landfall	N/A	<1,000	>1,000	>10,000	>100,000	>1 million
Witnessing Grotesque Scenes and/or Dead Bodies	Most intensely experienced by residents of Tiburon Peninsula close to landfall	N/A	<1,000	>1,000	>10,000	>100,000	>1 million
Human Generated Contribution	Extensive deforestation leading to mudslides and uncontrolled flooding	N/A	Low <20%	Moderate 20–40%	Severe 40–60%	Very Severe 60–80%	Extreme 80–100%
Intentional Human Causation	Not applicable	N/A	Low <20%	Moderate 20–40%	Severe 40–60%	Very Severe 60–80%	Extreme 80–100%
Multiple High Intensity Impacts	Compounding effects of damaging wind and water hazards	N/A	Low	Moderate	Severe	Very Severe	Extreme
Exposure to Loss		N/A	Low	Moderate	Severe	Very Severe	Extreme
Mortality	546 confirmed deaths	N/A	<1,000	>1,000	>10,000	>100,000	>1 million
Bereaved Survivors (Loss of Immediate Family)	Bereavement for several thousand primary family members of the victims	N/A	<4,000	>4,000	>40,000	>400,000	>4 million
Loss of Friend, Peer, Pet	Bereavement for community members, extended family of the deceased victims	N/A	<10,000	>10,000	>100,000	>1 million	>10 million
Loss of Primary Dwelling	Severe storm surge and wind damage to homes and structures with displacement	N/A	<1,000	>1,000	>10,000	>100,000	>1 million
National Financial Losses	Flash appeal for relief: $120 million Total damages, agricultural losses: >$1B	N/A	<$10 million	>$10 million	>$100 million	>$1 billion	>$10 billion
Exposure to Change		N/A	Low	Moderate	Severe	Very Severe	Extreme
Total Affected Persons	2.1 million affected; 1.4 million in need of immediate humanitarian aid	N/A	<100,000	>100,000	>1 million	>10 million	>100 million
Evacuation	142,000 displaced following storm impact	N/A	<1,000	>1,000	>10,000	>100,000	>1 million
Long-term Displacement/Relocation	142,000 displaced following storm impact	N/A	<1,000	>1,000	>10,000	>100,000	>1 million
Disruption of Essential Services	Widespread loss of electrical power, communications	N/A	<10,000	>10,000	>100,000	>1 million	>10 million
Lack of Food, Clean Water	750,000 need WASH (water, sanitation, hygiene) services for 3 months or more	N/A	<10,000	>10,000	>100,000	>1 million	>10 million
Damage to Infrastructure	Near total destruction of infrastructure in Jeremie and nearby areas on Tiburon Peninsula	N/A	Low	Moderate	Severe	Very Severe	Extreme
Ecological/Environmental Impact	Deforestation and storm damage leading to floods, mudslides, destruction of agricultural lands	N/A	Low	Moderate	Severe	Very Severe	Extreme

The trauma signature summary allows the Hurricane Matthew hazards, losses, and life changes, specific to Haiti, to be viewed on a format that displays the exposure severity for a range of documented risk factors for psychological distress and disorder. Similar presentations have been developed for other major disasters, including the 2010 Haiti earthquake, allowing rapid comparison across a variety of extreme events.

TSIG Step 8: Enumeration of the distinguishing features of Hurricane Matthew

Distinguishing features of Hurricane Matthew

From a meteorological perspective, Klotzbach and researchers at the Tropical Meteorology Project at Colorado State University indicated that Hurricane Matthew was notable on several metrics. First, Matthew attained Category 5 intensity status at the lowest latitude (closest to the equator) in the recorded history of Atlantic basin hurricanes. Second, among tropical cyclones traversing the eastern Caribbean, Matthew maintained Category 4/5 status for the longest duration on record. This was particularly significant for Haiti because this powerful storm remained south of Hispaniola for a protracted period of time as heavy rains deluged Haiti and neighboring Dominican Republic prior to the system moving northward and directly making landfall in western Haiti. Third, as a recognized indicator of the storm's power and destructive capacity, Matthew produced the most “accumulated cyclone energy” (ACE) of any eastern Caribbean storm on record. Fourth, Matthew remained at major hurricane strength (Category 3 or higher) for 7.25 days; this was the longest-duration Atlantic basin storm ever to attain major hurricane strength after September 25. Fifth, Matthew was the only tropical cyclone to make separate landfalls in Haiti, Cuba, and the Bahamas as a major hurricane.

Regarding the storm's unique physical impacts on Haiti, Matthew was notable for widespread heavy downpours coinciding with a “blob” of convection that was parked to the right of the center of circulation and created massive precipitation over Hispaniola even before Matthew's winds reached Haiti.

In terms of the storm's effects on human populations, Hurricane Matthew was particularly notable for the degree of devastation along the Tiburon Peninsula where a combination of Category 4 winds, pounding wave action, and storm surge beneath the center of the cyclone, combined with high tides and extreme rainfall to produce significant death, destruction, and psychological trauma.

Concluding comments

This trauma signature analysis has benefitted from the co-authorship of Drs. Cela and Marcelin who have been conducting intensive research and data collection in the hardest-hit areas of Haiti most affected by Hurricane Matthew's landfall.⁸³ The combination of the TSIG methodology combined with firsthand experience and deep knowledge of the culture, has facilitated the ability of our research team to tailor findings to the on-scene realities. The intended outcome of this analysis is to provide guidance for effective, targeted, mental health and psychosocial support interventions that are attuned to the exposures and experiences of the Hurricane Matthew survivors and their communities.

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.