Urban Evacuations and Rural America: Lessons Learned from Hurricane Rita

Schools of Public Health have been assisting state and local health departments with emergency preparedness education and training, including the Harvard School of Public Health (HSPH) and the Texas A&M School of Rural Public Health (SRPH). They are two of 27 Centers for Public Health Preparedness established in schools of public health across the country helping federal, state, and local officials prepare for bio-terrorism, emerging infections like avian flu, natural disasters such as hurricanes, and other emerging threats with great public health consequences. They are funded by the U.S. Centers for Disease Control and Prevention through a cooperative agreement with the Association of Schools of Public Health. These two institutions are especially concerned with rural emergencies; their staffs joined forces following Hurricane Rita to learn from the experience in Texas.

The hurricanes of 2005 revealed that public health officials responsible for management of emergencies in rural areas should be preparing for two issues: (1) human and property damage inflicted directly on rural communities, and (2) the flight of millions of temporarily homeless urban residents. This paper describes the results of interviews conducted following Hurricane Rita. Staff from both HSPH and SRPH Centers met with residents of rural southeastern Texas counties that faced serious human and property losses in their own communities as well as a concurrent influx of people due to large-scale urban evacuations. The authors focus upon the second challenge, with the hope that the experience gained during and following the storm in Texas can inform emergency officials who are planning evacuation procedures in other states. Specifically, this article will address evacuation issues related to transport, leaving problems such as serious shelter shortages for future discussion.

RESEARCH METHODS

The authors reviewed literature on urban evacuations, primarily reports and articles describing experience caused by natural disasters. They also interviewed a series of key informants in the affected areas of Southeastern Texas, speaking with more than 50 individuals from October 2005 through January 2006. They talked with hospital administrators, local emergency management officials, local public health staff, and shelter managers as well as residents. In addition, two of the authors based at Texas A&M had direct experience in the storm response, conducting rapid health assessments in a number of Texas shelters to identify potential public health threats. Finally, the authors spoke with state emergency management and public health officials and reviewed available secondary information, such as the Texas State Emergency Plan.

RURAL PUBLIC HEALTH CAPACITY

Public health and medical care capacity in rural areas of the U.S. is relatively weak, adversely affecting preparedness for emergencies of all kinds. Research compiled by the National Rural Health Association has documented that rural residents (approximately 20% of the U.S. population) are more likely to live in a medically underserved area than their urban counterparts and are more likely to die if injured.¹ In addition to the relative lack of physicians, hospitals, and other clinical resources, rural areas are also less likely to have well-staffed and equipped public health agencies. The Office of Rural Health Policy in the federal Health Resources and Services Administration has produced a document entitled “Rural Communities and Emergency Preparedness” that explains the daunting and uniquely rural challenges in coordination, communication, training, hospital capacity, and equipment that are too extensive to cite in this paper.²

In Texas, nearly one-half of the 254 counties do not have a local public health department and 65 counties do not have a hospital. Of the counties without hospitals, 38 counties have only primary care clinics, 25 counties have only EMS providers, and two counties have no available health services. To fill this gap, Texas has created eleven health service regions served by eight Regional Public Health Offices across the state. During emergencies such as the ones related to Hurricanes Katrina and Rita, a single regional office may be responsible for the public health of as many as 30 rural counties.

NATIONAL EXPERIENCE WITH EVACUATIONS

Hurricane experts at Louisiana State University published a report just prior to the 2005 hurricane season that included findings from a survey of coastal states on evacuation planning as well as an assessment of previous research.³ Research interest in evacuation has varied greatly over the past fifty or sixty years as the threats have changed along with the available technology and hazard mitigation priorities. During World War II and through the Cold War period, the focus was upon planning for urban evacuations resulting from military attacks using conventional, and later nuclear, weapons. During the 1960's and 70's, as satellites were utilized for the first time for storm tracking and forecasting, interest shifted to improving the effectiveness of hurricane evacuations. The Three Mile Island episode in 1979 created a third phase, reorienting planning efforts to nuclear power plant safety. Then the events of September 11, 2001 and subsequent anthrax attacks again altered the nation's attention and the priorities of emergency management professionals. They shifted their attention to terrorism, to mitigating damage inflicted on urban populations by nuclear, chemical, and biological weapons.

Hurricanes Katrina, Rita, and Wilma have turned our attention back to natural disasters in general, and hurricanes in particular, for two reasons: (1) the tremendous loss of life and property due to the storms, and (2) the prediction by some experts that weather patterns are changing, affected by global warming, and that the extraordinary hurricane experience of 2005 may be the beginning of an extended period of severe storms rather than an anomaly.

Hurricanes differ from the other threats mentioned above in the amount of warning time provided. Evacuation planning and execution, therefore, has relatively greater relevance. Improved hurricane tracking, predictions, and warnings have enhanced this advantage; as a result, saltwater deaths have been dramatically reduced. Flooding still claims the most lives (59% over the past 30 years), but it is now related to freshwater inland flooding. Experts have worried that enhanced weather predictions have created a false sense of security among public officials and developers, and contributed to a corresponding extensive housing boom found in coastal areas susceptible to hurricanes, including along the Gulf Coast. It remains to be seen if the damage caused by Katrina, Rita, and Wilma is sufficient to counter the economic incentives underlying this dangerous pattern.

HURRICANE RITA IN TEXAS

Plans for the evacuation of Texas residents are included in Annex E of the State Emergency Management Plan.⁴ This document describes emergency powers held by elected officials so that they can: (1) order evacuations; (2) prescribe routes, modes, and destinations; and (3) control disaster area ingress and egress.⁴ Twenty-two counties in Texas have been identified as high risk areas vulnerable to hurricane storm surge and gale-force winds.

Texas Governor Rick Perry directed the state Office of Homeland Security (OHS) to conduct a comprehensive assessment of state hurricane evacuation plans after the devastating hurricanes in Florida during 2004.⁵ During the following six months, the OHS reviewed existing evacuation routes, obtained input from local community leaders, identified potential gaps, and made recommendations for improvement. In March 2005, the OHS made public a number of recommendations, and at the third annual Texas Hurricane Conference held in Beaumont, TX in May 2005, the Governor's Division of Emergency Management provided a series of workshops to begin training and exercises for local jurisdictions on implementing these revised evacuation plans. Shelter Hubs and the Evacuation Information Centers, two new concepts in Texas evacuation planning, were introduced. A Shelter Hub is a pre-identified geographical location enabling state officials along with local partners to rapidly open core shelters, deploy mass care resources, and gather information for critical decision-making. The Hubs include public facilities large enough to serve as shelters and facilities to warehouse supplies and personnel. Evacuation Information Centers are roadside facilities located at highway rest stops that are designed to assist evacuees leaving or returning to a community. They are supposed to provide information to the public and to shelter staff deploying resources to the affected area. These Centers have been planned to provide restroom facilities, immediate emergency assistance, and fuel, but not food and shelter.

The storm

Hurricane Rita was the second-most powerful hurricane of the season (behind Hurricane Wilma and ahead of Katrina) measured by barometric pressure, and the fourth most intense hurricane ever to be recorded in the Atlantic Basin.

Rita raged through the Gulf with sustained winds of up to 175 miles per hour. Although originally projected to hit the Houston/Galveston area, Rita took an easterly turn while still in the Gulf, a shift in direction that spared these metropolitan areas a direct hit. The storm made landfall on Sunday, September 24, 2005, near the Louisiana/Texas border. It struck land with winds of up to 120 miles per hour, raking the coast with a fifteen-foot storm surge and dropping a foot of rain within hours. Beaumont, the economic center of a geographically large rural area, was not as fortunate as its larger neighbors to the west; Hurricane Rita forced 110,000 residents to flee and the storm left many homes and businesses in shambles. The rural areas from Port Arthur northward along the Texas/Louisiana border were similarly affected. Water and sewage services were not restored for many weeks. More than two million people found themselves without electrical power. Many houses and other structures were completely destroyed and await demolition, or remain covered with FEMA-provided blue tarps while awaiting repairs (“operation blue roof”). Five months after the storm, it appeared that Rita contributed to the deaths of 119 people, 113 of those in Texas. The important point for this article is that only six of those 113 deaths are “directly” attributed to the storm (wind, water, surge); the other 107 deaths were caused “indirectly,” primarily by activities (e.g., accidents) related to the evacuation process. Rita also caused an estimated $9.4 billion in damage, making it the ninth costliest storm in U.S. history.

Despite planning efforts, the evacuation conducted prior to Hurricane Rita's landfall encountered numerous problems that emergency officials will try to avoid or minimize in the future, including:

Traffic problems.

Harris County emergency evacuation models had predicted 800,000 to 1.2 million people evacuating from the affected areas; instead, approximately 2.5 million chose to flee the largely urban/suburban areas in Southeastern Texas prior to Hurricane Rita⁶. They traveled to the largely rural communities to the north.

Coastal communities successfully evacuated massive numbers of residents prior to Hurricane Rita's arrival. The evacuation process, however, was complicated by a number of factors. First, the late change in the hurricane's course meant that many more people than necessary were evacuated. Second, further congestion was created along evacuation routes when inland counties were also put under mandatory evacuation. Third, lack of adequate traffic control in rural areas contributed to the massive traffic problems. Fourth, the vivid warnings issued by the National Weather Service and media outlets, Texas Emergency Management, and FEMA officials motivated residents to evacuate. Finally, Texans were heavily influenced by the very recent experience of Hurricane Katrina. They knew that many New Orleans residents were unable to or refused to obey an evacuation order, and that a number of them subsequently died. A large proportion of the Katrina evacuees had found their way to Texas, and to the Houston area in particular. They provided Texans with a vivid, real-life demonstration of the value of evacuations.

The large number of residents fleeing from Hurricane Rita overwhelmed the infrastructure of many rural East Texas communities. On September 22, 2005, in one rural county alone, it was estimated that 150,000 vehicles sat bumper-to-bumper on four lanes of a 30-mile stretch of Interstate 45 north of Houston. The congested roadways prevented emergency medical workers from quickly responding to the medical emergencies of evacuees, including dialysis, oxygen, insulin, births, and deaths. Extended evacuation times caused major fuel shortages. Vehicles of every type ran out of gas and became stranded along the evacuation routes, worsening the congestion. A trip that usually takes three and a half hours became a 24-hour drive during the evacuation. When evacuees did reach their rural destinations, their huge demand for goods and services such as food, water, ice, and restroom facilities soon overwhelmed supply. Temperatures soared to 100 degrees and humidity hovered at 94%. Evacuees were forced to turn off their car air conditioners to conserve fuel or to keep engines from overheating. Lack of adequate restrooms along evacuation routes forced evacuees to use blankets and towels as privacy screens to construct makeshift facilities along the roadside. This unsanitary disposal of human waste created potential public health hazards such as the spread of infectious diseases and the contamination of the ground water supply.

Failure of Shelter Hubs and Evacuation Information Centers.

The Shelter Hub concept was designed to force evacuees at least 150 miles inland from the coastline. During the first few hours of the evacuation from the Houston area, the new Shelter Hubs and Evacuation Information Centers successfully facilitated information flow among state and local officials, response agencies, and the evacuees. However, these hubs quickly became overwhelmed as unexpectedly large numbers of evacuees moved along the designated evacuation routes. Congestion stopped the flow of traffic, rendering them useless. As frustrated evacuees attempted to find alternate routes that circumvented the gridlock, these facilities were bypassed and officials lost the ability to monitor the needs of evacuees, provide fuel and basic necessities, offer timely information to this population, and track shelter availability. The Governor's Task Force on Hurricane Preparedness is currently evaluating the effectiveness of these resources and the possibility of relocating Shelter Hubs at least 400 miles inland.

Communication problems.

Communication was a major concern for evacuees as they were unable to place or receive phone calls for emergency or other purposes during the evacuation stage. Wireless phones were typically the only means of communication for evacuees, but they quickly became unreliable when the volume of calls overwhelmed the networks.

Risk communication was also a serious issue. Consistent and accurate warnings are essential for the public prior to and during evacuations. Texas officials did utilize both television and radio for this purpose. Residents, however, also made use of non-official media outlets including local radio and television news stations and the internet, sources that conveyed inaccurate or at least outdated information that may have stimulated unnecessary or premature evacuations by certain groups.⁷ Official announcements also did not foresee the time period that evacuees would spend on the road, and therefore did not sufficiently stress the need to carry food, water, pharmaceuticals, and other medical supplies.

Inefficient patient evacuations.

Helicopters and fixed-wing aircraft were used to evacuate special needs patients from the Beaumont region. Although many patients were successfully evacuated from the area, several major issues were noted. The Transportation Security Administration required screening of all patients before they could board any evacuation aircrafts, significantly slowing down the process. In addition, airplanes are not regularly configured to carry wheelchairs, stretchers, and patients with cumbersome medical equipment. Seats needed to be removed and areas within multiple aircrafts reconfigured. Motorized wheelchairs and oxygen tanks are just two of many items that constitute a flight hazard and are banned from commercial airplanes. When wheelchairs were confiscated at the gate, patients arrived at shelters without the means to be mobile and replacement equipment was not available.

Hazardous evacuations of nursing homes.

Although authorities in Texas reported improved evacuation efforts, many elderly residents were still placed in life-threatening situations because of nonexistent or untested plans. Busloads of nursing home residents were trapped in the gridlock of the major evacuation routes. Many of the residents spent up to 24 hours traveling, often overheated and without access to restroom facilities, food, water, or medication.

The Houston Chronicle has identified a large number of evacuation-related fatalities from nursing homes in the Houston area.⁸ This includes the 23 residents who perished in a bus fire south of Dallas on September 23. Authorities said the bus carrying 38 residents and six employees of a Houston area nursing home apparently caught fire due to a mechanical problem, and that oxygen tanks caused explosions, engulfing the bus in flames.⁹

Many Texas nursing homes lacked evacuation plans, despite state requirements. A number of the nursing homes with evacuation plans had secured school buses with no air conditioning. Even more troublesome, more than 10 nursing homes utilized plans to use the same bus service to evacuate, resulting in an overall shortage of buses to transport residents. Ambulance companies had also been over-obligated.¹⁰

DISCUSSION

The problems described above only begin to describe the issues to be addressed in evacuation planning. The Gulf hurricanes have underscored the need for careful evacuation planning tailored to a range of threats in all 18 at-risk states identified by the Louisiana State University report. The authors reached the following conclusions concerning evacuations:

• Planning should be a multi-state effort to assure evacuation effectiveness from major metropolitan areas. This planning should be followed by training and exercises.

• Public health officials, few of whom have been involved to date, should learn about evacuation planning, implementation, and evaluation, and get involved at the state and local levels to improve health outcomes.

• Emergency management officials should involve public health officials in evacuation planning, implementation, and evaluation.

• Emergency management and public health staff should strive for a high level of connectivity among all agencies involved in evacuations, including public safety and transportation.

• Rural public health and medical care capacity needs to be strengthened to meet the challenges posed by emergencies, including hurricanes.

Texans encountered the serious challenges described in this paper despite a heightened awareness of hurricane dangers and fairly extensive planning efforts. Most of the other “at-risk” states listed in the LSU report have not completed evacuation planning programs anywhere near the level found in Texas. In New England, for example, where Category 4 hurricanes did strike the coast 50 years ago, the authors have not found evidence of interstate agreements and exercises designed to meet the many challenges that would be posed by large numbers of people fleeing the Boston area. The need for greater investment in evacuation planning and execution is a national issue.