Home Current Issue Previous Issues Collections Podcasts Blogs CME For Authors Journal Info
Skip Navigation LinksHome > January 2013 - Volume 106 - Issue 1 > Vertical Hospital Evacuations: A New Method
Southern Medical Journal:
doi: 10.1097/SMJ.0b013e31827caef4
Healthcare System Preparedness

Vertical Hospital Evacuations: A New Method

Iserson, Kenneth V. MD, MBA

Free Access
Article Outline
Collapse Box

Author Information

From the University of Arizona and the AZ-1 Disaster Medical Assistance Team, Tucson.

Correspondence to Dr Kenneth V. Iserson, Professor Emeritus, University of Arizona, Tucson. Email: kvi@u.arizona.edu

K.V.I. has received royalty payments and grants for publications from McGraw-Hill.

Accepted April 11, 2012.

Collapse Box


Abstract: Rarely are hospitals forced to evacuate their nonambulatory patients; however, when a disaster occurs, evacuating nonambulatory patients, particularly from multilevel facilities, represents a major logistical hurdle. Hospital disaster plans often rely on outside agencies and limited equipment to perform vertical evacuations. This article describes a novel method using readily available materials (patient mattresses and bedsheets) to effect a rapid, safe vertical evacuation. This method also can be used in nonhealthcare facilities for less-than-fully ambulatory individuals.

Key Points

* Hospitals are rarely forced to evacuate their nonambulatory patients.

* Vertical evacuations from multilevel facilities represent a major logistical hurdle.

* A novel method using readily available materials (patient mattresses and bedsheets) to effect a rapid, safe vertical evacuation is described.

* This method also can be used in nonhealthcare facilities for less-than-fully ambulatory individuals.

Hospitals rarely are forced to evacuate their nonambulatory patients, but when a disaster occurs, evacuating nonambulatory patients, particularly from multilevel facilities, represents a major logistical hurdle. As Schultz et al pointed out, “Traditionally, hospitals [have been] viewed as sources of medical treatment. However, as has been repeatedly proven, they are also vulnerable to disasters: the hospital becomes the patient.”1

As repeatedly demonstrated during major catastrophes, hospitals are vulnerable to both external and internal disasters. The most common reasons for evacuations are external weather-related events, in particular, hurricanes. Floods, tsunamis, and tornados comprise other less common weather-related causes. Non–weather-related external disasters affecting hospitals include area-wide fires, earthquakes, and military actions. Internal disasters that may precipitate an evacuation include fire, gas or chemical leaks, biological threats, terrorist events, and loss of power, water, and ventilation.

High-rise hospitals most commonly exist in the developed world, as do other multilevel structures, such as apartments or office buildings, in which less-than-fully ambulatory people may reside or work. They are also present in many large cities in the least developed countries. Evacuations of single-level hospitals, which primarily exist outside the most developed countries, pose significantly fewer difficulties than those of high-rise structures.

Back to Top | Article Outline

Historical Context

Comparable to attitudes regarding the “unsinkable” RMS Titanic, when it was necessary to emergently evacuate, many hospitals in the developed world have believed themselves to be impervious to disaster. For example, given their hospitals’ state-of-the-art backup generators, administrators were surprised when these failed, often because they were flooded. A prime example is the University of Texas Medical Branch at Galveston. Although planners had situated the backup generator on an upper floor, a direct hit from Hurricane Ike in 2008 eliminated their power source, forcing administrators to evacuate the huge complex. I subsequently staffed their emergency department, the only part of the medical complex that was operating, with a Disaster Medical Assistance Team (DMAT) using generators supplied by the federal government.

This was not my first such experience: In June 1972, Tropical Storm Agnes flooded the new primary backup generator at the University of Maryland Hospital. The streets surrounding the hospital were flooded, which eliminated the possibility of evacuating patients, even if it was possible to transport them from the upper-floor intensive care units (ICUs) and wards. We scrambled up dark stairwells to bag ventilator-dependent patients for the many hours that the power remained out.

Tropical Storm Allison caused catastrophic flooding and loss of power, water, and communications at Houston’s Memorial Hermann Hospital in June 2001 when the main power system went offline and switches for the backup generators, located in the basement, were under floodwaters. Consequently, the hospital evacuated more than 500 patients.2 This vertical evacuation required that “patients be carried out on backboards down four flights of stairs to the hospital’s general transfer holding area.”2 One nurse wrote, “Ten people carried a hand-ventilated patient on a stretcher down the stairs on their shoulders. Every person’s face showed the strain.”2

During Allison, the Children’s Hospital was forced to evacuate their seventh-floor neonatal unit. An easier task in general than evacuating adults, it required moving infants in critical condition four flights of stairs to the helipad or down seven flights to waiting ambulances. As a member of the nursing staff later wrote, “Babies were secured on backboards with tape, identification stickers covering them. Staff members strapped blankets, equipment, and oxygen tanks to the boards and carried them down seven flights of stairs. We had no power; we monitored the old-fashioned way—with assessment skills and a stethoscope.”2

Charity Hospital in New Orleans and Chalmette Regional Medical Center in adjacent St Bernard Parish, Louisiana, experienced the same fate, and permanently so, after Hurricane Katrina in 2005. In both instances, physicians related that some patients were eventually lowered by bedsheets to boats: “No patients suffered injuries from the experience, but that was probably as much due to luck as it was to skill.”3 Following Superstorm Sandy in October 2012, multiple New York City hospitals also had to evacuate all of their patients as a result of infrastructure failures.

Back to Top | Article Outline

Current Requirements/Planning

Most countries do not require hospitals to prepare for external or internal disasters, including formulating and practicing patient evacuation plans.4,5 In the United States, however, The Joint Commission (formerly known as the Joint Commission for the Accreditation of Healthcare Organizations) requires “that all hospitals prepare contingency plans for the possible scenarios in the event of systems failures.”6 As of 2008, this has meant that each hospital’s “emergency operations plan identifies [sic] the capabilities and established response efforts when [it] cannot be supported by the local community for at least 96 hours in the six critical areas,” including staged evacuation and total evacuation.6

The Joint Commission’s requirement does not ensure that practicable evacuation measures are in place. To comply with The Joint Commission, California’s Hospital Evacuation Plan, an updated version of New York’s plan, directs users to notify the local fire department and emergency medical service when considering evacuation.7 Although hospital evacuation plans exist, their vertical evacuation components rely on personnel (eg, fire departments that may be unable to respond, at least immediately, in a widespread disaster environment) and equipment that may not be available (eg, specially made evacuation equipment that may require special training to use and under the best circumstances will be available for only one patient at a time).

The US government considers the goal in hospital evacuations as optimizing “the time to empty the building”—that is, the time required to move patients from their location inside the hospital (eg, their room) to a staging area from which they can be loaded into ambulances and other vehicles for transport to another hospital. The staging area may be the lobby of the hospital, the emergency department (which has ambulance bays), or a parking lot across the street from the hospital.”8 The assumption is that once patients are in a staging area, appropriate shelter or transportation will be available.

Back to Top | Article Outline

Current Approaches to Vertical Evacuation

As was noted in the discussion of the examination of hospital evacuations after Tropical Storm Allison hit Houston, “A hospital disaster preparation plan is usually directed toward handling an external event that creates mass casualties, not an internal event that requires complete evacuation of the institution. Evacuation of a hospital is rare, but has been reported in cases of riot, earthquake, internal flooding, and hazardous material spill.”9 For disaster planners, the vertical evacuation of high-rise hospitals represents either an insurmountable obstacle or an element left to local fire departments. Nonambulatory patients may need to be moved down stairways or lowered out of windows when a building must be evacuated. These techniques can be dangerous and strenuous, especially when performed by personnel who have not been trained in such systems.4 Participants in the Houston hospital evacuation noted: “[It] required multiple persons to carry each patient. Volunteers were numerous and eager at the beginning of the day, but by midnight they were tired. In retrospect, it would have been prudent to bring patients from upper floors when help was fresh and available rather than keeping patients on the upper floors until ready to evacuate. The evacuation of patients was slow and eventually halted until morning to decrease risk of potential injury to either patients or volunteers.”9

Moving ICU patients within a hospital with all systems functioning often presents logistical challenges. This level of complexity “would be unrealistic in the event of an unplanned evacuation of ICU patients, especially if it was necessary to move patients down flights of stairs (i.e., vertical evacuation).”10 Multiple methods have been suggested for vertical evacuations, including using a five-person canvas (Reeves) stretcher with grips, a one-person method using a mattress pad with belt fasteners attached to secure the patient and to provide grips, Stokes (wire mesh) baskets with inserted backboards, an air mattress or the mattress from a patient’s bed, various forms of blanket drags, one- or two-person carries, and expensive proprietary devices.2,11,12 All of these methods are useful in specific situations, although those used for mass vertical evacuations require considerable effort and sufficient personnel.

Hospital staff used potentially dangerous evacuation techniques, such as sheets, to lower patients into boats in the aftermath of Hurricane Katrina. Hospital personnel should not “attempt to lower patients unless absolutely necessary! The basic technique, without normal litters and ropes, involves securing the patient as much as possible into sheet slings and using additional sheets as ‘ropes.’ Bed-sheet ropes can be anchored to hospital beds; or the rigid ‘standpipes’ in stairwells can be used to belay patients. However, under the best circumstances, this method involves a great deal of risk.”4

Experience has shown that most hospital disaster plans expect vertical patient evacuation to be handled using fire department stretchers and multiple teams of physically fit people to clear the way and carry patients individually or use backbreaking multiple-person carries.11,13 These are extremely slow, labor-intensive processes that are mainly unsuccessful.

The questions that disaster planners and participants should consider before implementing a traditional firefighter-managed vertical evacuation from a hospital include the following: What are the firefighters’ fitness levels when wearing full turnout gear (eg, how many patients could they evacuate before becoming exhausted)? What method is optimal for vertical evacuations? What minimal patient safety level, including monitoring, is optimal or required? How many firefighters would be needed to efficiently evacuate all nonambulatory patients?11

Another possible scenario is the hospital’s experiencing a “drop everything and go” evacuation, in which patients and staff are in immediate danger and must exit the hospital as quickly as possible. Yet, according to the Agency for Healthcare Research and Quality, “Depending on the type of disaster, there would likely be staff shortages. It would be helpful to attempt to pre-estimate the attrition rate of a hospital’s workforce during a disaster, as many employees may themselves become victims of the disaster, or have family responsibilities that interfere with their ability to staff the hospital (e.g., evacuating dependent children.)”8

One vertical evacuation drill, under optimal conditions, averaged 8 minutes per patient executed by 15 well-rested, physically fit firefighters, 12 nurses, 12 physicians, and 2 respiratory therapists, who evacuated 12 simulated ICU patients down 4 floors.10 A similar drill using a Stokes basket demonstrated that “a four firefighter extraction team along with an accompanying nurse and respiratory therapist would be able to evacuate one patient at a rate of 3.75 minutes per floor,” nearly double the time of the prior study.12 Both studies were conducted using slender volunteers, who are markedly unlike most patients in an ICU. They were also conducted “in well-lit stairwells that were free of smoke, and from a unit that had not lost electrical power. If the power was out and the stairwells dark, one person would have had to lead and light the way, possibly move debris, or communicate with other rescuers.”10 These less than optimal conditions would also slow the evacuation flow.14

Back to Top | Article Outline

Mattress–Sheet Method4

In contrast to the various evacuation techniques described above, the following vertical descent method can be used/improvised in nearly any situation, is much faster, takes only a few people of normal strength to operate, and requires no specialized equipment or personnel. (This method, however, is much harder to use in an ascent, such as to a helipad.) In 2008, its use was illustrated by rapidly evacuating volunteer “patients” from the top floor of a five-story hospital. Videotape of the exercise showed that <1 minute per floor elapsed for a slow, controlled descent. A DMAT that had no prior experience with the technique (AZ-1, Tucson, Arizona) demonstrated that it was fast, safe, and easily implemented. Hospital staff subsequently used this technique successfully in multiple drills.

The technique involves two parts: two- to three-person teams prepare multiple patients for evacuation, while another team clears the stairwell(s) to be used and lines them with mattresses from patient beds. Personnel should follow the instructions below for preparing a patient’s room:

1. Untuck the patient’s sheet(s) from the bed. If the patient is lying on only one sheet, insert another sheet under him or her.

2. Wrap the patient in the two sheets, leaving the patient’s face uncovered.

3. Tie tightly both ends of each sheet individually at the head and foot ends with a square knot (Fig. 1). Wrapping the patient in two sheets provides extra support and, for safety, allows two rescuers (usually) at the head end each to hold a different sheet while removing the patient from the bed. The rescuer at the foot end holds both sheets in one hand. If only one sheet is on the bed, retrieve another sheet from storage or use a bedcover as the second sheet.

Fig. 1
Fig. 1
Image Tools

4. Move the patient’s bed so that there is sufficient room for the patient to lie lengthwise on the floor perpendicular to the bed.

5. Move the patient off the bed by rotating the mattress so that it is 90° to the long axis of the bed, with the head portion completely on the bed and the foot portion off the bed and leading to the floor (Fig. 2).

Fig. 2
Fig. 2
Image Tools

6. Slide the patient off the mattress and onto the floor gently, using the sheets for control, with one or two people holding the sheets on each end (Fig. 3).

Fig. 3
Fig. 3
Image Tools

7. Slide the patient down the hall to the stairwell, with one person on each end of the sheets. On most smooth hospital floors, two adults can easily slide most normal-size patients to the head of the stairs (Fig. 4). The staff sliding the patient must avoid any debris or obstacles. If the floor is wet, it may be slippery, but this should not unduly slow patient movement.

Fig. 4
Fig. 4
Image Tools

Personnel should follow the instructions below for moving a patient through a stairwell:

1. Check the entire length of the stairwell to be certain that it is clear and that an exit is available at the bottom. For safety, station someone at the bottom of the stairway to ensure that the exit stays open and to warn the team if it does not. The amount of time needed to clear the stairwell and line the stairs with mattresses depends upon the nature of the disaster, but it should not take long in most instances.

2. Place patient mattresses end to end on the stairs. Cover each flight of the stairs (the stairs between a flat landing) with mattresses by placing the mattress even with the top stair and extending the mattresses to the landing. Push all of the mattresses against the wall side of the stairway, leaving a narrow area for personnel to walk and to grasp a railing for balance and safety.

3. Repeat the process for the entire length of stairs to be traversed. Each hospital floor will, in general, have more than enough mattresses to cover at least several flights of the staircase; however, patients removed from their beds may be forced to wait short periods before the stairs are covered completely.

4. Transport each patient down the stairs using a three-person team. The person at the foot of the patient lifts the patient’s feet, using the sheets, onto the end of the mattress at the top of the stairs. If available, for large or obese patients, a slide board may make this process easier. The person holding the foot end begins walking down the stairs, sliding the patient. The person holding the head end of the sheet follows, simultaneously holding onto the stairway handrail and slowing the patient’s slide as necessary. If available, it is helpful for a third person to grasp the top (head) person’s belt to steady him or her (Fig. 5).

Fig. 5
Fig. 5
Image Tools

5. At the landing, slide the patient to the next set of mattresses and repeat the process, continuing to the bottom of the stairs.

6. Once each team clears a landing, the next patient can be started down the mattress ramp from the flight above.

7. Staff can either transport patients down only one flight and return for others or, to maintain continuity with the patients, transport the same patient to the bottom. The vertical distance they travel during the course of the evacuation will be identical in both cases.

When considering this method, questions naturally arise regarding the patient’s weight, accompanying medical equipment, and problems in the stairwell. In the DMAT exercise, despite significant weight and size discrepancies between the simulated patients (up to approximately 240 lb/76 in.) and the staff (approximately 130 lb/65 in.), smaller staff experienced no significant problems moving much larger patients. If necessary, additional staff can be used, which would be most beneficial at the foot end while sliding the patient down the hall and at the head end while helping to control the slide down the stairwell.

Any vertical evacuation method that does not include the use of elevators necessarily limits the equipment that can be used during descent to those that are vital for short-term survival. The mattress–sheet method is no exception. One or two extra personnel may be needed in the stairwell beside the patient to, for example, use a bag-valve mask. Intravenous solutions can be placed under the patient (pressurize the bag), within an inflated blood pressure cuff, or in a portable fluid pressure device.4 There may be no way to evacuate patients who cannot survive being taken off large mechanical assistance devices for a brief time.

Under some circumstances, stairwells may be obscured by smoke or darkness if backup power or battery-powered lights have failed. It is suggested that at least some staff who perform this evacuation method wear headlamps or that other staff members try to light the area.

Back to Top | Article Outline

Conclusions: Where Do We Go from Here?

Why is there still a need for improvised methods for vertical hospital evacuation, particularly in the most-developed countries? Multiple reasons exist. The most obvious reason is that the evacuation of hospitals is rarely necessary. Because this represents a low-probability/high-risk situation, hospital administrators can, with some confidence, adopt an “it won’t happen to us” attitude. Planners, saddled with practical financial constraints, perceive vertical evacuation as beyond their capabilities, relying instead on local fire departments. Unfortunately, few fire departments routinely practice the simultaneous evacuation of multiple nonambulatory patients. The limited evacuation equipment hospitals occasionally purchase is woefully inadequate for a mass evacuation.

Vertical hospital evacuations involve transporting nonambulatory patients down stairwells, because elevators will be nonoperable or dangerous to use. Ideally, as Nates wrote, “Effective design and systems for vertical evacuation of patients could help to reduce evacuation time, personnel requirements, potential injuries, or deaths. This includes updated engineering and architectural design of stairwells, as well as special evacuation equipment.”4 (Nates’s “special equipment” is expensive proprietary evacuation equipment that can be used by only one patient at a time.)

Few significant advances have occurred in developing useable, cost-effective, and available techniques to accommodate a large simultaneous vertical hospital evacuation of nonambulatory patients. At present, the mattress–sheet method, if practiced, offers a widely available method. Future research funding may expand the options.

Back to Top | Article Outline

Sue A. Philpott, RN, BSN, training officer, AZ-1 DMAT and emergency management coordinator, Oro Valley Hospital, Oro Valley, AZ, is acknowledged for helping to develop and test the mattress–sheet evacuation method.

Back to Top | Article Outline


1. Schultz CH, Koenig KL, Auf der Heide E, et al.. Benchmarking for hospital evacuation: a critical data collection tool. Prehosp Disaster Med 2005; 20: 331–342.

2. Verklan MT, Kelley K, Carter L, et al.. The day the rain came down: stranded in the NICU by tropical storm Allison. Am J Nurs 2002; 102: 24AA–24II.

3. Iserson KV. Improvised Medicine: Providing Care in Extreme Environments. New York, McGraw-Hill, 2012: 279–281.

4. Nates JL. Combined external and internal hospital disaster: impact and response in a Houston trauma center intensive care unit. Crit Care Med 2004; 32: 686–690.

5. Kai T, Ukai T, Ohta M, et al.. Hospital disaster preparedness in Osaka, Japan. Prehosp Disaster Med 1994; 9: 29–34.

6. Alert: change in compliance expectations for emergency management standards. Jt Comm Perspect 2008; 28:1, 3–6.

7. California Hospital Association. Hospital evacuation checklist. www.calhospitalprepare.org/post/hospital-evacuation-checklist. Accessed April 9, 2012.

8. Zane R, Biddinger P, Gerteis J, et al. Hospital assessment and recovery guide. http://archive.ahrq.gov/prep/hosprecovery/hosprecovery.pdf. Accessed April 9, 2012.

9. Cocanour C, Allen S, Mazabob J, et al.. Lessons learned from the evacuation of an urban teaching hospital. Arch Surg 2002: 137: 1141–1144.

10. Manion P, Golden IJ. Vertical evacuation drill of an intensive care unit: design, implementation, and evaluation. Disaster Manag Response 2004; 2: 14–19.

11. Gretenkort P, Harke H, Blazejak J, et al.. Interface between hospital and fire authorities—a concept for management of incidents in hospitals. Prehosp Disaster Med 2002; 17: 42–47.

12. Gildea JR, Etengoff S. Vertical evacuation simulation of critically ill patients in a hospital. Prehosp Disaster Med 2005; 20: 243–248.

13. Alvarez AR, Russell MT. Emergency evacuation: removal of the critically ill patient. Focus Crit Care 1987; 14: 18–22.

14. Pauls J. Vertical evacuation in large buildings: missed opportunities for research. Disaster Manag 1994; 6: 128–132.


disasters; hospital evacuation; improvisation; vertical evacuation

© 2013 Southern Medical Association


Search for Similar Articles
You may search for similar articles that contain these same keywords or you may modify the keyword list to augment your search.