- Email: [email protected]
Minneapolis bridges falling down: Emergency transfusion preparedness
Transfusion and Apheresis Science 49 (2013) 403–407
Contents lists available at SciVerse ScienceDirect
Transfusion and Apheresis Science journal homepage: www.elsevier.com/locate/transci
Minneapolis bridges falling down: Emergency transfusion preparedness Jed B. Gorlin a,b,⇑, John L. Hick c,1 a
Memorial Blood Centers, St. Paul, MN, United States Hennepin County Medical Center, Transfusion Service, United States c Emergency Medicine MC 825, Hennepin County Medical Center, 701 Park Ave. S, Minneapolis, MN 55415, United States b
a r t i c l e
i n f o
Keywords: Disaster response Transfusion preparedness Integrated networks
a b s t r a c t The 7/1/2007 bridge collapse into the Missisippi River was instructional from both a disaster response and a mass casualty transfusion response perspective. It is a well cited example of how community disaster response coordination can work well, especially following systematic preparation of an integrated response network. The blood center is and should be an integral part of this diaster response and should be included in drills where appropriate. We give personal perspectives on both the hospital and transfusion service response to this particularly dramatic event. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction August 1, 2007: Nobody predicted the fall of the I-35 bridge into the Mississippi River. Built in 1967 it had been rated in 2006 as ‘‘structurally deficient, but not in need of immediate replacement.’’ In short, it was no worse than myriad other similar structures of aging infrastructure around the US. Its 2000 foot span crossed a 390 foot wide section of the Mississippi about 7 feet deep. The bridge deck was about 115 feet above the water, the center of the structural arch 64 feet. About 141,000 cars/day traversed the span, far more than when it was conceived and built. The collapse occurred at 6:05 pm as cars were in a relative stand still either on a homeward commute or travelling into the city for a 7PM Minnesota Twins baseball game, immediately adjacent to the first exit from the bridge travelling South over the river. Fortuitously, also adjacent to both the stadium and the bridge is the city’s busiest level one trauma center, Hennepin County Medical
⇑ Corresponding author at: Hennepin County Medical Center, Transfusion Service, United States. Tel.: +1 651 332 7284. E-mail addresses: [email protected] (J.B. Gorlin), [email protected] (J.L. Hick). 1 Tel.: +1 612 873 4908. 1473-0502/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.transci.2013.05.034
Center, where one of us (JG) is co-medical director of the Transfusion Service. The bridge collapsed in sections over seconds, as sequential portions failed (see Fig. 1). There was no redundant capacity as it was a steel bridge single arch bridge with two major supporting beams, built to avoid pillars in the river which would be problematic for the upriver lock and dam. Gusset plates, which hold the girders together, were of insufficient thickness when installed, and over time rusted and buckled. Under the extra weight of building materials placed on the bridge and rush hour traffic at a virtual standstill, one of these plates failed, triggering the collapse. The standstill traffic was, in retrospect, a blessing as cars fell straight down when the bridge collapsed, as opposed to careening off at highway speeds. The hazards were too numerous to recount, but included the fact that the collapse was ongoing, so that in some cases victims survived the fall, cushioned in their cars but later succumbed when additional bridge sections or debris fell on them as they escaped their vehicles. The response can be summarized as follows: Collapse to last patient transported: Initial clearing of all sectors: 1 h 35 min. Last EMS transport: 2 h 6 min.
404
J.B. Gorlin, J.L. Hick / Transfusion and Apheresis Science 49 (2013) 403–407
Fig. 1. Northern portion of the 35W bridge collapse. The majority of critically injured victims were extricated from the riverbank and area below the overhanging section of the bridge where the blue van came to rest. Overall debris field was >1 mile in length. Photo courtesy of the Minneapolis StarTribune – used with permission. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
50 patients transported by EMS; (HCMC 23, Univ. MN 11, North Memorial, another level one trauma hospital 9 and a community hospital Abbott Northwestern 6. All of the critically injured were brought to HCMC. 8–13 casualties via other vehicle, usually good Samaritans who stopped by with pick up trucks, but patients were accompanied by EMTs or other emergency personnel. Over 100 patients treated in 24 h. 13 deaths. No serious injuries to first responders. 29 ambulances used in first 4 h. The response has been reviewed in the disaster medicine literature [4,5] and has been acknowledged as an example of the success of integrated disaster planning and response [2]. It is fortuitous that the nearby large level one trauma hospital also houses the busiest EMS service in the state. In addition, it houses the Regional Hospital Resource Center (RHRC) that coordinates care among participating hospitals (and is activated on an event-specific basis) and the Medical Resource Control Center (MRCC) which coordinates routine as well as incident-specific regional EMS agency communications and response to mass casualty incidents. MRCC is the gatekeeper for the webbased statewide ‘‘MNTRAC system’’ which in the metro keeps live information flowing about ER status, bed availability, patient numbers and patient destinations. 2. Disaster planning and response Despite nearly 1.5 miles of response zone, the incident went relatively smoothly because a plan was in place AND drilled regularly. The plan included notification of off-duty personnel, a shared incident response plan for the 24 EMS agencies, and practiced integration with Minneapolis Fire. The ability to expand/contract as needed was built into the system and exercises tested the surge capacity and systems. Through the auspices of the local disaster preparedness unit, there had been drills of various
disasters (although a bridge collapse had not been among them!), that had previous coordinated responses across jurisdictions (Minneapolis and St. Paul police) across departments, fire and police and EMS and other services. John Hick MD, Medical director for emergency preparedness at Hennepin County Medical Center, quipped, ‘‘The middle of an emergency is a lousy time to be exchanging business cards.’’ Because they had prior preparation and many of the responders had previously worked together the collaboration across jurisdictions and services was commendable. A surge capacity model and rapid size up tool for health care facilities during the initial response has been summarized as the CO-S-TR framework [6]. Co(4) Command, Control, communication, coordination. S(4) Space, staff, Stuff, Special (logistics). Tr(4) Triage, Treatment, Transport and Tracking. Focusing on the primary receiving hospital (HCMC) response, the initial information about the bridge collapse was received at about 6:10 pm when the hospital was near capacity. Only 5 intensive care unit beds were available. In the emergency room, the most critically ill patients brought in by ambulance are initially seen and stabilized in a four bed room (‘‘stabilization room’’) – two of those beds were already occupied at the time of the collapse. For immediate access to additional information, the charge nurse turned on the television in patient area and in consultation with senior medical staff an ‘‘Alert Orange’’ was declared at 6:15 pm which automatically triggers several key functions. The alert automatically recalls key personnel, holds on duty personnel, including paging all ED staff to the hospital. In addition, patient reception areas were cleared, ED beds opened, and a hospital command center is established in a pre-designated area staffed by key administrative and clinical personnel identified by premade vests. It was not until 6:40 that the first critically ill patients actually arrived at the hospital where they were received by an overwhelming number of staff who descended upon the ER to ‘‘see if they could help.’’ In short, it was both staff and patients that needed initial assignment to areas in the ED. Over the ensuing 2 h, 25 critically ill patients were received, one dead on arrival (who expired enroute to the hospital from non-survivable thoracic injuries). 6 were intubated, 5 went directly to the operating room after their ED stabilization including one for an emergent Caesarian section for fetal distress (with a good outcome for both mother and baby). Of the 24 patients, 60% were admitted. In the 45 min after the Orange alert was declared the surge capacity plan was activated and 20 additional ICU beds were made available (by calling in additional staff or transferring stable patients to the floor), 10 operating rooms were opened and staffed (instead of the two typically maintained all evening) and all three CT scanners were up and running (normally, only the one in the ER is kept open after usual business hours). Two of the HCMC emergency department physicians are also fellowship-trained EMS physicians and responded to the collapse site per system protocol, liaisoning with the
J.B. Gorlin, J.L. Hick / Transfusion and Apheresis Science 49 (2013) 403–407
EMS supervisors and assisting in rescue and triage operations. Many additional medical personnel spontaneously came to the scene (from nearby hospitals and home). While their Good Samaritan intent was appreciated, their involvement sometimes created safety concerns as they were ill-equipped and ill-trained to be working in such a hazardous environment. As an aside, Minnesota is a leading state in signing up medical personnel for the Medical Reserve Corps, a national system of local corps designed to pre-credential medical personnel to assist in the event of disasters, though not for actual scene response. Furthermore the local agencies provide training on mass casualty, mass public health initiative (vaccination, drug dispensing), and psychological support of victims and providers during disasters. 3. Lessons learned Before getting to the transfusion specific issues, it is worthwhile to review other lessons learned. Auf der Heide [1] summarized common assumptions that are often incorrect during a disaster response: (1) Dispatchers will hear of disaster and send appropriate emergency units to the scene. (2) Trained emergency personnel will carry out field search and rescue. (3) Trained EMTs will carry out triage, provide first aid or stabilizing before transport. (4) Casualties will be transported to hospitals by ambulance. (5) Casualties will be transported to hospitals appropriate for their needs and proportionate to their ability to acutely care for them. (6) Authorities will assure prompt notification of hospitals including numbers, types and severities of casualties. (7) The most serious casualties will be transported to hospitals first. In the case of the I-35 Bridge collapse, myriad services heard and many different police, fire and EMS units were sent. Many of the initial rescuers were bystanders – as noted above. At least 5 of the seriously ill were transported in the back of pick-up trucks because normal access roads to one of the casualty collection areas were blocked by portions of the bridge that fell on them. The final point is perhaps most germane. It takes time to extract and stabilize the most critically injured, so the first showing up at the ER are the ‘‘walking wounded’’ that self-transport, usually to the closest facility. Therefore, the initial wave is not a good predictor of severity and gives no perspective on what is to follow; though they sometimes are the first information that the facility may have about an incident that has occurred. 4. Communication Cell phone circuits were overwhelmed for the first 90 min of the event until capacity shifting was performed
405
by the cell carriers. Hospital switchboards close to the incident were jammed by staff and patient inquiries, which is typical for a major incident. 800 mHz communication for responders experienced no significant issues. The most useful communication tool between EMS and the hospitals was MRCC which facilitated triaging the most critically ill accident victims to HCMC while diverting the less critically ill, or those with unrelated emergencies to neighboring hospitals who were better staffed to deal with them and issued alerts on the MnTrac system, providing 29 updates over the course of the evening. Finally, archaic solutions, such as runners within the hospital provided redundant simple means of communication, for example, JG walked between the command center and the hospital transfusion service to ensure staff could handle the increasing patient volume. Communications to the outside world was equally challenging. There was intense national attention. The hospital quickly realized it needed to designate a few spokespersons to systematically provide information and interviews on a scheduled basis. This generally meant relieving those persons from direct patient care assignments, which allowed the perception of consistent and focused information. Early in the event well-intentioned media provided significant amount of misinformation. For example, they requested any medically trained person to go to the bridge to help, recalled all HCMC personnel whether they were needed or not and instructed community members to go to hospitals to donate blood (problematic as all blood collection fixed sites are not located in the hospitals). It was also after hours for Memorial Blood Centers, so we retained staff to encourage donors to return in the morning.
5. Patient tracking considerations Patient tracking was difficult and patients (even from the same family) were often taken to different hospitals. Fortunately information sharing issues had been addressed prior to this event by inter-hospital compact that allowed for sharing of information for public safety tracking and for reunification. Information that was shared was not HIPAA protected by design. Multiple agencies did request information that could not be provided, however, such as questions about specific injuries to specific persons. Questions about sharing information with other agencies are being further clarified by Minnesota Department of Health. A Family Assistance Center was rapidly set up to address the needs of those seeking loved ones and was invaluable in taking some of these pressures off the hospital.
6. Supplies Select ED supplies became temporarily exhausted. For example, the stabilization room ran out of simple gauze pads and other consumables, particularly intubation drugs but OR and pharmacy personnel quickly obtained these items. Hospitals may wish to have selected disaster supplies brought to ED by default and need to have a good replacement mechanism in place. Stockpiles of commonly
406
J.B. Gorlin, J.L. Hick / Transfusion and Apheresis Science 49 (2013) 403–407
needed items should be available based on guidance by departments of health and preparedness program efforts.
7. Transfusion specific issues 7.1. Distribution Recognizing that the extent of injuries was unknown and that transportations issues would likely get worse, the hospital service manager at Memorial Blood Centers took the liberty of simply contacting the hospitals proximate to the accident site who we serve and might conceivably receive patients. We then send additional group O cells in proportion to what we knew they had in their inventory and potential additional need. Like most prior major physical disasters relatively few units were required immediately. In fact, only 13 units of red cell concentrates were used that evening all at HCMC and only about 50 total products for the 24 critically injured patients by the end of the week. Since the exact blood need and transfusion service work load was not known at the time of receipt of additional units, in consultation with staff, we called in additional technologists, initially pulling staff from the core lab and subsequently a transfusion service staff arrived from home. Transfusion specific steps following an alert orange: Once staff were notified, there is a Standard operating procedure (SOP) for a Laboratory Disaster Plan and a SOP for a Transfusion Service Disaster Response. The lab wide plan directs the transfusion service to assess the current blood supply, contact medical director, communicate with Memorial Blood Center, the community blood center to arrange for needed products, and to conserve O negative red cells. The Transfusion service specific plan highlights what to do for specific problem scenarios like: computers are down, phone is down, electric is out, or fridges/freezers for blood product storage don’t work. This plan also tells staff to fill out an assessment tool. The assessment tool documents current inventory, what systems are or aren’t working, what are staffing levels are and helps to assess what actions are needed to ensure the transfusion service can appropriately handle the level of disaster called. Having sufficient group O Rh(neg) units was greatly facilitated by the longstanding policy of reserving the Rh(neg) units for females while emergency uncrossmatched transfusions to males are group O when units are needed before type specific blood can be administered. Since the majority of massive trauma patients are male (suggesting, perhaps that those of male gender are the more unwise of the species?), this prevents the consumption of many O Rh(neg) units. Since 6/7 recipients of the Rh(positive) units are Rh(pos) themselves, this is of no consequence to the majority, and the few surviving Rh(neg) patients who do become sensitized, only become so weeks later. In the event that a patient had been previously sensitized, we review prior history and if none, generally have an antibody screen completed within the first ½ hour after patient arrival. Since all platelets transfused at our institution are collected by apheresis and functionally free of red cells, we do not require Rh matching of platelets which
further simplified obtaining and distributing platelets as needed following the event. While we continue to update and revise our massive transfusion policy, it was activated multiple times that evening for the many patients potentially in need for many blood products. The transfusion service staff monitor patients with large ongoing needs and request frequent labs (Hct/Hgb, plt, PT (INR), PTT, fibrinogen to guide transfusion. While at that time we were not rigidly adhering to a fixed ratio of red cell to plasma units, we do agree that it is prudent not to wait for a dilutional coagulopathy to develop. The default ratio at the time was to dispense about one unit of thawed plasma for every two units of red cells, but to adjust as a function of the iterative coagulation values received. HCMC leaves 2 units of AB plasma thawed at all times. In the setting of multiple trauma victims, this policy does mean that the second victim received in a short period of time does not have pre-thawed plasma available, and we are currently revising the policy to keep 2 units of group A plasma (compatible with both group O and A recipients and hence 83% of our local population) thawed in addition to the 2 AB units. This will facilitate ‘‘keeping ahead’’ when sending out multiple rounds of coolers during a massive transfusion event. This low amount of blood used is entirely consistent with multiple mass trauma events [3]. Schmidt [7] reviewed US civilian disasters and found that virtually all major disasters used less than the daily collection volume of the nearest blood center which generally all maintain 3– 5 days inventory at all times. Event
Date
# RBC used
Skywalk collapse- KC Airliner crash Sioux City Oklahoma City Govt. Bombing Columbine high school 9/11
1981 1989 1995 1999 2001
126 119 131 105 600
Since the bridge collapse we have collaborated with the regional wide disaster planning task force to include transfusion where appropriate into scenario planning. This has included mock drills of mass shooting events, where we included both the trauma center and the blood center to the point of timing how long it took for additional platelets (actually an empty box) to be shipped from the blood center to the laboratory (<30 min), justifying a low stock inventory of platelets at the trauma center. Additionally, we have worked out redundant communications mechanisms in the event that usual telephone services are unavailable. WealsoworkedwiththeMinnesotaDepartmentofHealth on an approach to scarce resources and included single page charts on how we might deal with various levels of shortages for both red blood cells and platelets. (See: http:// www.health.state.mn.us/oep/healthcare/standards.pdf. Note:blood productson page 12–13 of the card set.) In short, there are virtually no disasters that overwhelm the total inventory of blood products available somewhere, but the details of actually providing the exact components
J.B. Gorlin, J.L. Hick / Transfusion and Apheresis Science 49 (2013) 403–407
when and where they are needed are best ensured by drilling multiple scenarios in advance of a disaster event and assuring that the blood bank staff understand their roles and their communication and response options. This ensures better communication and fewer bottlenecks when a real event happens. Finally, it helps identify clinical or agency silos (also wittily described as ‘‘cylinders of excellence’’) that may not communicate well during stressful events for which plans must be made to assure the integrated response in order to best meet the needs of the community. References [1] Auf der Heide E. The importance of evidence-based disaster planning. Ann Emergency Med 2006;47(1):34–49.
407
[2] FEMA. I-35WBridge collapse and response. USFA-TR-166. 2007. US Department of Homeland Security. Technical report series. Ref Type: Report. US Fire Administration. [3] Hess JR, Thomas MJ. Blood use in war and disaster: lessons from the past century. Transfusion 2003;43(11):1622–33. [4] Hick JL et al. Hospital response to a major freeeway bridge collapse. Disaster Med Public Health Preparedness 2008;2:S11–6. [5] Hick JL et al. Emergency medical services response to a major freeway bridge collapse. Disaster Med Public Health Preparedness 2008;2:S17–24. [6] Hick JL et al. Concepts in disaster medicine: surge capacity concepts for health care facilities: the CO-S-TR model for initial incident assessment. Disaster Med Public Health Preparedness 2008;2(Suppl. 1):S51–7. [7] Schmidt P. Blood and disaster-supply and demand. NEJM 2002;346:617–20.
Contents lists available at SciVerse ScienceDirect
Transfusion and Apheresis Science journal homepage: www.elsevier.com/locate/transci
Minneapolis bridges falling down: Emergency transfusion preparedness Jed B. Gorlin a,b,⇑, John L. Hick c,1 a
Memorial Blood Centers, St. Paul, MN, United States Hennepin County Medical Center, Transfusion Service, United States c Emergency Medicine MC 825, Hennepin County Medical Center, 701 Park Ave. S, Minneapolis, MN 55415, United States b
a r t i c l e
i n f o
Keywords: Disaster response Transfusion preparedness Integrated networks
a b s t r a c t The 7/1/2007 bridge collapse into the Missisippi River was instructional from both a disaster response and a mass casualty transfusion response perspective. It is a well cited example of how community disaster response coordination can work well, especially following systematic preparation of an integrated response network. The blood center is and should be an integral part of this diaster response and should be included in drills where appropriate. We give personal perspectives on both the hospital and transfusion service response to this particularly dramatic event. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction August 1, 2007: Nobody predicted the fall of the I-35 bridge into the Mississippi River. Built in 1967 it had been rated in 2006 as ‘‘structurally deficient, but not in need of immediate replacement.’’ In short, it was no worse than myriad other similar structures of aging infrastructure around the US. Its 2000 foot span crossed a 390 foot wide section of the Mississippi about 7 feet deep. The bridge deck was about 115 feet above the water, the center of the structural arch 64 feet. About 141,000 cars/day traversed the span, far more than when it was conceived and built. The collapse occurred at 6:05 pm as cars were in a relative stand still either on a homeward commute or travelling into the city for a 7PM Minnesota Twins baseball game, immediately adjacent to the first exit from the bridge travelling South over the river. Fortuitously, also adjacent to both the stadium and the bridge is the city’s busiest level one trauma center, Hennepin County Medical
⇑ Corresponding author at: Hennepin County Medical Center, Transfusion Service, United States. Tel.: +1 651 332 7284. E-mail addresses: [email protected] (J.B. Gorlin), [email protected] (J.L. Hick). 1 Tel.: +1 612 873 4908. 1473-0502/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.transci.2013.05.034
Center, where one of us (JG) is co-medical director of the Transfusion Service. The bridge collapsed in sections over seconds, as sequential portions failed (see Fig. 1). There was no redundant capacity as it was a steel bridge single arch bridge with two major supporting beams, built to avoid pillars in the river which would be problematic for the upriver lock and dam. Gusset plates, which hold the girders together, were of insufficient thickness when installed, and over time rusted and buckled. Under the extra weight of building materials placed on the bridge and rush hour traffic at a virtual standstill, one of these plates failed, triggering the collapse. The standstill traffic was, in retrospect, a blessing as cars fell straight down when the bridge collapsed, as opposed to careening off at highway speeds. The hazards were too numerous to recount, but included the fact that the collapse was ongoing, so that in some cases victims survived the fall, cushioned in their cars but later succumbed when additional bridge sections or debris fell on them as they escaped their vehicles. The response can be summarized as follows: Collapse to last patient transported: Initial clearing of all sectors: 1 h 35 min. Last EMS transport: 2 h 6 min.
404
J.B. Gorlin, J.L. Hick / Transfusion and Apheresis Science 49 (2013) 403–407
Fig. 1. Northern portion of the 35W bridge collapse. The majority of critically injured victims were extricated from the riverbank and area below the overhanging section of the bridge where the blue van came to rest. Overall debris field was >1 mile in length. Photo courtesy of the Minneapolis StarTribune – used with permission. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
50 patients transported by EMS; (HCMC 23, Univ. MN 11, North Memorial, another level one trauma hospital 9 and a community hospital Abbott Northwestern 6. All of the critically injured were brought to HCMC. 8–13 casualties via other vehicle, usually good Samaritans who stopped by with pick up trucks, but patients were accompanied by EMTs or other emergency personnel. Over 100 patients treated in 24 h. 13 deaths. No serious injuries to first responders. 29 ambulances used in first 4 h. The response has been reviewed in the disaster medicine literature [4,5] and has been acknowledged as an example of the success of integrated disaster planning and response [2]. It is fortuitous that the nearby large level one trauma hospital also houses the busiest EMS service in the state. In addition, it houses the Regional Hospital Resource Center (RHRC) that coordinates care among participating hospitals (and is activated on an event-specific basis) and the Medical Resource Control Center (MRCC) which coordinates routine as well as incident-specific regional EMS agency communications and response to mass casualty incidents. MRCC is the gatekeeper for the webbased statewide ‘‘MNTRAC system’’ which in the metro keeps live information flowing about ER status, bed availability, patient numbers and patient destinations. 2. Disaster planning and response Despite nearly 1.5 miles of response zone, the incident went relatively smoothly because a plan was in place AND drilled regularly. The plan included notification of off-duty personnel, a shared incident response plan for the 24 EMS agencies, and practiced integration with Minneapolis Fire. The ability to expand/contract as needed was built into the system and exercises tested the surge capacity and systems. Through the auspices of the local disaster preparedness unit, there had been drills of various
disasters (although a bridge collapse had not been among them!), that had previous coordinated responses across jurisdictions (Minneapolis and St. Paul police) across departments, fire and police and EMS and other services. John Hick MD, Medical director for emergency preparedness at Hennepin County Medical Center, quipped, ‘‘The middle of an emergency is a lousy time to be exchanging business cards.’’ Because they had prior preparation and many of the responders had previously worked together the collaboration across jurisdictions and services was commendable. A surge capacity model and rapid size up tool for health care facilities during the initial response has been summarized as the CO-S-TR framework [6]. Co(4) Command, Control, communication, coordination. S(4) Space, staff, Stuff, Special (logistics). Tr(4) Triage, Treatment, Transport and Tracking. Focusing on the primary receiving hospital (HCMC) response, the initial information about the bridge collapse was received at about 6:10 pm when the hospital was near capacity. Only 5 intensive care unit beds were available. In the emergency room, the most critically ill patients brought in by ambulance are initially seen and stabilized in a four bed room (‘‘stabilization room’’) – two of those beds were already occupied at the time of the collapse. For immediate access to additional information, the charge nurse turned on the television in patient area and in consultation with senior medical staff an ‘‘Alert Orange’’ was declared at 6:15 pm which automatically triggers several key functions. The alert automatically recalls key personnel, holds on duty personnel, including paging all ED staff to the hospital. In addition, patient reception areas were cleared, ED beds opened, and a hospital command center is established in a pre-designated area staffed by key administrative and clinical personnel identified by premade vests. It was not until 6:40 that the first critically ill patients actually arrived at the hospital where they were received by an overwhelming number of staff who descended upon the ER to ‘‘see if they could help.’’ In short, it was both staff and patients that needed initial assignment to areas in the ED. Over the ensuing 2 h, 25 critically ill patients were received, one dead on arrival (who expired enroute to the hospital from non-survivable thoracic injuries). 6 were intubated, 5 went directly to the operating room after their ED stabilization including one for an emergent Caesarian section for fetal distress (with a good outcome for both mother and baby). Of the 24 patients, 60% were admitted. In the 45 min after the Orange alert was declared the surge capacity plan was activated and 20 additional ICU beds were made available (by calling in additional staff or transferring stable patients to the floor), 10 operating rooms were opened and staffed (instead of the two typically maintained all evening) and all three CT scanners were up and running (normally, only the one in the ER is kept open after usual business hours). Two of the HCMC emergency department physicians are also fellowship-trained EMS physicians and responded to the collapse site per system protocol, liaisoning with the
J.B. Gorlin, J.L. Hick / Transfusion and Apheresis Science 49 (2013) 403–407
EMS supervisors and assisting in rescue and triage operations. Many additional medical personnel spontaneously came to the scene (from nearby hospitals and home). While their Good Samaritan intent was appreciated, their involvement sometimes created safety concerns as they were ill-equipped and ill-trained to be working in such a hazardous environment. As an aside, Minnesota is a leading state in signing up medical personnel for the Medical Reserve Corps, a national system of local corps designed to pre-credential medical personnel to assist in the event of disasters, though not for actual scene response. Furthermore the local agencies provide training on mass casualty, mass public health initiative (vaccination, drug dispensing), and psychological support of victims and providers during disasters. 3. Lessons learned Before getting to the transfusion specific issues, it is worthwhile to review other lessons learned. Auf der Heide [1] summarized common assumptions that are often incorrect during a disaster response: (1) Dispatchers will hear of disaster and send appropriate emergency units to the scene. (2) Trained emergency personnel will carry out field search and rescue. (3) Trained EMTs will carry out triage, provide first aid or stabilizing before transport. (4) Casualties will be transported to hospitals by ambulance. (5) Casualties will be transported to hospitals appropriate for their needs and proportionate to their ability to acutely care for them. (6) Authorities will assure prompt notification of hospitals including numbers, types and severities of casualties. (7) The most serious casualties will be transported to hospitals first. In the case of the I-35 Bridge collapse, myriad services heard and many different police, fire and EMS units were sent. Many of the initial rescuers were bystanders – as noted above. At least 5 of the seriously ill were transported in the back of pick-up trucks because normal access roads to one of the casualty collection areas were blocked by portions of the bridge that fell on them. The final point is perhaps most germane. It takes time to extract and stabilize the most critically injured, so the first showing up at the ER are the ‘‘walking wounded’’ that self-transport, usually to the closest facility. Therefore, the initial wave is not a good predictor of severity and gives no perspective on what is to follow; though they sometimes are the first information that the facility may have about an incident that has occurred. 4. Communication Cell phone circuits were overwhelmed for the first 90 min of the event until capacity shifting was performed
405
by the cell carriers. Hospital switchboards close to the incident were jammed by staff and patient inquiries, which is typical for a major incident. 800 mHz communication for responders experienced no significant issues. The most useful communication tool between EMS and the hospitals was MRCC which facilitated triaging the most critically ill accident victims to HCMC while diverting the less critically ill, or those with unrelated emergencies to neighboring hospitals who were better staffed to deal with them and issued alerts on the MnTrac system, providing 29 updates over the course of the evening. Finally, archaic solutions, such as runners within the hospital provided redundant simple means of communication, for example, JG walked between the command center and the hospital transfusion service to ensure staff could handle the increasing patient volume. Communications to the outside world was equally challenging. There was intense national attention. The hospital quickly realized it needed to designate a few spokespersons to systematically provide information and interviews on a scheduled basis. This generally meant relieving those persons from direct patient care assignments, which allowed the perception of consistent and focused information. Early in the event well-intentioned media provided significant amount of misinformation. For example, they requested any medically trained person to go to the bridge to help, recalled all HCMC personnel whether they were needed or not and instructed community members to go to hospitals to donate blood (problematic as all blood collection fixed sites are not located in the hospitals). It was also after hours for Memorial Blood Centers, so we retained staff to encourage donors to return in the morning.
5. Patient tracking considerations Patient tracking was difficult and patients (even from the same family) were often taken to different hospitals. Fortunately information sharing issues had been addressed prior to this event by inter-hospital compact that allowed for sharing of information for public safety tracking and for reunification. Information that was shared was not HIPAA protected by design. Multiple agencies did request information that could not be provided, however, such as questions about specific injuries to specific persons. Questions about sharing information with other agencies are being further clarified by Minnesota Department of Health. A Family Assistance Center was rapidly set up to address the needs of those seeking loved ones and was invaluable in taking some of these pressures off the hospital.
6. Supplies Select ED supplies became temporarily exhausted. For example, the stabilization room ran out of simple gauze pads and other consumables, particularly intubation drugs but OR and pharmacy personnel quickly obtained these items. Hospitals may wish to have selected disaster supplies brought to ED by default and need to have a good replacement mechanism in place. Stockpiles of commonly
406
J.B. Gorlin, J.L. Hick / Transfusion and Apheresis Science 49 (2013) 403–407
needed items should be available based on guidance by departments of health and preparedness program efforts.
7. Transfusion specific issues 7.1. Distribution Recognizing that the extent of injuries was unknown and that transportations issues would likely get worse, the hospital service manager at Memorial Blood Centers took the liberty of simply contacting the hospitals proximate to the accident site who we serve and might conceivably receive patients. We then send additional group O cells in proportion to what we knew they had in their inventory and potential additional need. Like most prior major physical disasters relatively few units were required immediately. In fact, only 13 units of red cell concentrates were used that evening all at HCMC and only about 50 total products for the 24 critically injured patients by the end of the week. Since the exact blood need and transfusion service work load was not known at the time of receipt of additional units, in consultation with staff, we called in additional technologists, initially pulling staff from the core lab and subsequently a transfusion service staff arrived from home. Transfusion specific steps following an alert orange: Once staff were notified, there is a Standard operating procedure (SOP) for a Laboratory Disaster Plan and a SOP for a Transfusion Service Disaster Response. The lab wide plan directs the transfusion service to assess the current blood supply, contact medical director, communicate with Memorial Blood Center, the community blood center to arrange for needed products, and to conserve O negative red cells. The Transfusion service specific plan highlights what to do for specific problem scenarios like: computers are down, phone is down, electric is out, or fridges/freezers for blood product storage don’t work. This plan also tells staff to fill out an assessment tool. The assessment tool documents current inventory, what systems are or aren’t working, what are staffing levels are and helps to assess what actions are needed to ensure the transfusion service can appropriately handle the level of disaster called. Having sufficient group O Rh(neg) units was greatly facilitated by the longstanding policy of reserving the Rh(neg) units for females while emergency uncrossmatched transfusions to males are group O when units are needed before type specific blood can be administered. Since the majority of massive trauma patients are male (suggesting, perhaps that those of male gender are the more unwise of the species?), this prevents the consumption of many O Rh(neg) units. Since 6/7 recipients of the Rh(positive) units are Rh(pos) themselves, this is of no consequence to the majority, and the few surviving Rh(neg) patients who do become sensitized, only become so weeks later. In the event that a patient had been previously sensitized, we review prior history and if none, generally have an antibody screen completed within the first ½ hour after patient arrival. Since all platelets transfused at our institution are collected by apheresis and functionally free of red cells, we do not require Rh matching of platelets which
further simplified obtaining and distributing platelets as needed following the event. While we continue to update and revise our massive transfusion policy, it was activated multiple times that evening for the many patients potentially in need for many blood products. The transfusion service staff monitor patients with large ongoing needs and request frequent labs (Hct/Hgb, plt, PT (INR), PTT, fibrinogen to guide transfusion. While at that time we were not rigidly adhering to a fixed ratio of red cell to plasma units, we do agree that it is prudent not to wait for a dilutional coagulopathy to develop. The default ratio at the time was to dispense about one unit of thawed plasma for every two units of red cells, but to adjust as a function of the iterative coagulation values received. HCMC leaves 2 units of AB plasma thawed at all times. In the setting of multiple trauma victims, this policy does mean that the second victim received in a short period of time does not have pre-thawed plasma available, and we are currently revising the policy to keep 2 units of group A plasma (compatible with both group O and A recipients and hence 83% of our local population) thawed in addition to the 2 AB units. This will facilitate ‘‘keeping ahead’’ when sending out multiple rounds of coolers during a massive transfusion event. This low amount of blood used is entirely consistent with multiple mass trauma events [3]. Schmidt [7] reviewed US civilian disasters and found that virtually all major disasters used less than the daily collection volume of the nearest blood center which generally all maintain 3– 5 days inventory at all times. Event
Date
# RBC used
Skywalk collapse- KC Airliner crash Sioux City Oklahoma City Govt. Bombing Columbine high school 9/11
1981 1989 1995 1999 2001
126 119 131 105 600
Since the bridge collapse we have collaborated with the regional wide disaster planning task force to include transfusion where appropriate into scenario planning. This has included mock drills of mass shooting events, where we included both the trauma center and the blood center to the point of timing how long it took for additional platelets (actually an empty box) to be shipped from the blood center to the laboratory (<30 min), justifying a low stock inventory of platelets at the trauma center. Additionally, we have worked out redundant communications mechanisms in the event that usual telephone services are unavailable. WealsoworkedwiththeMinnesotaDepartmentofHealth on an approach to scarce resources and included single page charts on how we might deal with various levels of shortages for both red blood cells and platelets. (See: http:// www.health.state.mn.us/oep/healthcare/standards.pdf. Note:blood productson page 12–13 of the card set.) In short, there are virtually no disasters that overwhelm the total inventory of blood products available somewhere, but the details of actually providing the exact components
J.B. Gorlin, J.L. Hick / Transfusion and Apheresis Science 49 (2013) 403–407
when and where they are needed are best ensured by drilling multiple scenarios in advance of a disaster event and assuring that the blood bank staff understand their roles and their communication and response options. This ensures better communication and fewer bottlenecks when a real event happens. Finally, it helps identify clinical or agency silos (also wittily described as ‘‘cylinders of excellence’’) that may not communicate well during stressful events for which plans must be made to assure the integrated response in order to best meet the needs of the community. References [1] Auf der Heide E. The importance of evidence-based disaster planning. Ann Emergency Med 2006;47(1):34–49.
407
[2] FEMA. I-35WBridge collapse and response. USFA-TR-166. 2007. US Department of Homeland Security. Technical report series. Ref Type: Report. US Fire Administration. [3] Hess JR, Thomas MJ. Blood use in war and disaster: lessons from the past century. Transfusion 2003;43(11):1622–33. [4] Hick JL et al. Hospital response to a major freeeway bridge collapse. Disaster Med Public Health Preparedness 2008;2:S11–6. [5] Hick JL et al. Emergency medical services response to a major freeway bridge collapse. Disaster Med Public Health Preparedness 2008;2:S17–24. [6] Hick JL et al. Concepts in disaster medicine: surge capacity concepts for health care facilities: the CO-S-TR model for initial incident assessment. Disaster Med Public Health Preparedness 2008;2(Suppl. 1):S51–7. [7] Schmidt P. Blood and disaster-supply and demand. NEJM 2002;346:617–20.