Because full-scale evacuations of such buildings are rare, little is known about how readily and rapidly these buildings can be evacuated and what factors serve as facilitators or barriers to the process (1).

In 2002, the Mailman School of Public Health at Columbia University and CDC initiated The World Trade Center Evacuation Study, a multiyear qualitative and quantitative research study designed to assess factors that affected evacuation of the two WTC towers. This report summarizes qualitative data collected from Phase I of the study, which suggested that improved preparedness at the individual, organizational, and building environmental levels can facilitate rapid evacuation. Completion of Phase II of the study, together with other research efforts, should help workers, management, and local authorities develop and evaluate model emergency preparedness programs for high-rise occupancies.

Qualitative data for Phase I of the WTC study were collected from 56 participants during 2003, approximately 18 months after the events of September 11, 2001. Participants were self-selected into the qualitative study in response to a multimedia recruitment campaign; they consisted of 36 persons who were administered in-depth, semi-structured interviews and 20 who participated in five focus groups. The data collected helped guide development of a detailed study questionnaire for the quantitative Phase II* of the study.

The 56 participants ranged in age from 23 to 61 years; the mean and median age was 43 years. A total of 31 (55%) were male; 42 (75%) were white, seven (13%) were black, and one (2%) was Asian. Four (7%) identified themselves as of Hispanic ethnicity. A total of 37 (66%) of the participants were college graduates.

Interview scripts were designed to identify the factors that influenced both the decision-making process, as well as the actual evacuation-related behaviors. Transcripts of the taped responses were read and categorized by two reviewers, with coding themes verified by a third reviewer using a modified Q-sort methodology (2). Inter-rater reliability was high, with >95% concordance.

Individual factors. Participants cited four factors that affected their decision to begin evacuating: 1) perceived ability to walk down multiple flights of stairs (i.e., more than 80 for certain persons); 2) experience in evacuation of a WTC tower, including knowledge of stairwell locations and whether individual stairwells led to street level exits; 3) concern over leaving their work areas without the approval of executives or managers; and 4) information regarding what had occurred, what floors were involved, and how to respond. Direct evidence of the magnitude of the event (e.g., observing an aircraft strike a building, smelling fuel, or feeling a building move) caused some persons to leave immediately.

The qualitative data also suggested that, after a decision to evacuate was made, many persons stopped to attend to last-minute activities (e.g., making telephone calls, shutting down computers, or gathering up personal items). Deciding which route to take (e.g., stairs or elevators) might have delayed evacuation progress for others. Progress was reportedly slowed for some persons because of poor physical condition or inadequate footwear (e.g., high-heeled shoes or "flip-flops"). Some persons also delayed their progress to stop and assist others.

Organizational factors. Two major organizational factors affecting evacuation were identified by participants: 1) workplace preparedness planning and training, including evacuation drills (e.g., when drills were held, the majority reported they never actually entered their designated stairwells) and 2) inadequate risk communication. An announcement broadcast in WTC 2 (South Tower) shortly after the first aircraft had struck WTC 1 (North Tower) urged persons to remain in the building and likely led many to return to their work stations.

Building environmental factors. Three major evacuation factors in the WTC building environment were identified as 1) structural damage that blocked egress routes (e.g., debris on stairs or partially collapsed interior walls); 2) heavy congestion on certain stairways, which in some cases caused evacuees to move back upstairs in hopes of switching to a less congested stairwell; and 3) lack of back-up communication systems (e.g., public address system, elevator telephone system, and telephone system). When these systems failed, communication was severely limited.

Participants' experience with evacuations and emergency training varied by occupation. Service workers and temporary employees were less likely than others to have received fire safety training or been instructed in procedures during an emergency. Temporary workers were at a disadvantage because of their lack of familiarity with building evacuation procedures. Many permanent workers, even those with years of experience in the buildings, also reported they did not know how to evacuate via routes that deviated from their normal paths. Many reported confusion at the sky lobby levels, where transfer to express elevators occurred.

Reported by: RRM Gershon, DrPH, E Hogan, MPH, Dept of Sociomedical Science; KA Qureshi, DNSc, Dept of Epidemiology, Mailman School of Public Health, Columbia Univ, New York, New York. L Doll, PhD, National Center for Injury Prevention and Control, CDC.

Editorial Note:

These qualitative data provide initial insights regarding preparedness planning for total evacuation of other multistory buildings, either business or residential. The data suggest that preparedness planning at the individual, organizational, and structural levels factored into evacuation efforts at the WTC towers.

The findings in this report are subject to at least three limitations. First, the study cannot provide an estimate of the prevalence of these factors among WTC evacuees. Second, the data are subject to recall and response biases, which are common to self-reported qualitative and behavioral data. Finally, because of the study design and small sample size, the generalizability of these data to other high-rise evacuations is unknown.

The high rate of survivability of persons on floors in WTC 1 and 2 that were below the aircraft impacts was attributed to the small percentage (estimated at 20%--30%) of the building occupants at their work stations at the time of the first attack (3), reducing the flow of evacuees on the stairways and hastening evacuation (4). The lessons learned from the 1993 WTC bombing resulted in extensive fire safety improvements (e.g., improved signage and better lighting in stairwells) and also are believed to have facilitated the evacuation process (5). The 1993 bombing led some employers to enhance their emergency preparedness plans, which might have helped evacuate their employees more rapidly (6). In addition, on September 11, 2001, evacuees almost uniformly acted in an orderly and cooperative manner. The large numbers of first responders on the stairwells, in the lobby, and at ground level, might have helped to reassure and motivate evacuees.

High-rise buildings should be prepared for rapid, total building evacuation (7). More detailed information on the impact of these various factors on evacuation behaviors and how these behaviors affected the length of time to evacuate is being collected in Phase II, the quantitative phase of this study. A total of 1,500 randomly selected evacuees are completing detailed survey questionnaires that focus on their individual decision-making processes during the disaster. These data will provide a better understanding of the factors that helped shape evacuation decisions. Data from this phase of the study, as well as from similar initiatives, can help inform builders, developers, insurance companies, employee groups, and emergency planners about risk-reduction strategies. Data from such sources also might be of value to regulators at the federal, state, and local levels. To ensure adequate readiness, further studies should focus on the development and evaluation of model emergency preparedness programs for high-rise buildings.