Critical Resources for Hospital Surge Capacity : An Expert Consensus Panel

Background: Hospital surge capacity (HSC) is dependent on the ability to increase or conserve resources. The hospital surge model put forth by the Agency for Healthcare Research and Quality (AHRQ) estimates the resources needed by hospitals to treat casualties resulting from 13 national planning scenarios. However, emergency planners need to know which hospital resource are most critical in order to develop a more accurate plan for HSC in the event of a disaster. Objective: To identify critical hospital resources required in four specific catastrophic scenarios; namely, pandemic influenza, radiation, explosive, and nerve gas. Methods: We convened an expert consensus panel comprised of 23 participants representing health providers (i.e., nurses and physicians), administrators, emergency planners, and specialists. Four disaster scenarios were examined by the panel. Participants were divided into 4 groups of five or six members, each of which were assigned two of four scenarios. They were asked to consider 132 hospital patient care resourcesextracted from the AHRQ’s hospital surge modelin order to identify the ones that would be critical in their opinion to patient care. The definition for a critical hospital resource was the following: absence of the resource is likely to have a major impact on patient outcomes, i.e., high likelihood of untoward event, possibly death. For items with any disagreement in ranking, we conducted a facilitated discussion (modified Delphi technique) until consensus was reached, which was defined as more than 50% agreement. Intraclass Correlation Coefficients (ICC) were calculated for each scenario, and across all scenarios as a measure of participant agreement on critical resources. For the critical resources common to all scenarios, Kruskal-Wallis test was performed to measure the distribution of scores across all scenarios. 1 PLOS Currents Disasters Results: Of the 132 hospital resources, 25 were considered critical for all four scenarios by more than 50% of the participants. The number of hospital resources considered to be critical by consensus varied from one scenario to another; 58 for the pandemic influenza scenario, 51 for radiation exposure, 41 for explosives, and 35 for nerve gas scenario. Intravenous crystalloid solution was the only resource ranked by all participants as critical across all scenarios. The agreement in ranking was strong in nerve agent and pandemic influenza (ICC= 0.7 in both), and moderate in explosives (ICC= 0.6) and radiation (ICC= 0.5). Conclusion: In four disaster scenarios, namely, radiation, pandemic influenza, explosives, and nerve gas scenarios; supply of as few as 25 common resources may be considered critical to hospital surge capacity. The absence of any these resources may compromise patient care. More studies are needed to identify critical hospital resources in other disaster scenarios.


Introduction
Hospital surge capacity is dependent on the ability to increase or conserve resources, in response to an influx of patients in a disaster situation 1 .Adequate surge capacity depends on the fundamental understanding of which hospital resources are critical to patient care 2 .Surge capacity has four conceptual components, three of which categorize resources; space, staff, and supplies 3 , 4 , 5 .The fourth key component defining surge capacity is the "system" category, which denotes organization, processes, policies and procedures that govern and organize the allocation and conservation of the first three components.
In 2009, the Agency for Healthcare Research and Quality (AHRQ) developed a Hospital Surge Model 6 that forecasted the hospital resources required to treat casualties resulting from 13 National Planning Scenarios 7 .The AHRQ's planning tool was discontinued on June 30, 2011 6 .The objective of our study was to identify which of the AHRQ hospital resources were the most critical to care for patients in four of the National Planning Scenarios: pandemic influenza, radiation event, explosion, and nerve gas attack 7 .This information is essential for hospital preparedness planners to make supply-chain driven decisions based on the number of patients treated.In addition, we wished to determine which of these critical hospital resources were common across multiple types of events, and distinguish resources that are critical only for specific scenarios.

Methods
We organized and conducted a facilitated national expert consensus panel held during the spring of 2011.
Twenty three expert panelists from various disciplines in the health care sector (Table 1) took part in a one day consensus-building set of exercises.The majority of the participants were purposefully chosen to be healthcare providers, because the decision of what constitutes critical hospital resource for patient care, and issues related to standards of care are primarily clinical.However, we did include allied professionals to ensure other

Introduction
Hospital surge capacity is dependent on the ability to increase or conserve resources, in response to an influx of patients in a disaster situation 1 .Adequate surge capacity depends on the fundamental understanding of which hospital resources are critical to patient care 2 .Surge capacity has four conceptual components, three of which categorize resources; space, staff, and supplies 3 , 4 , 5 .The fourth key component defining surge capacity is the "system" category, which denotes organization, processes, policies and procedures that govern and organize the allocation and conservation of the first three components.
In 2009, the Agency for Healthcare Research and Quality (AHRQ) developed a Hospital Surge Model 6 that forecasted the hospital resources required to treat casualties resulting from 13 National Planning Scenarios 7 .The AHRQ's planning tool was discontinued on June 30, 2011 6 .The objective of our study was to identify which of the AHRQ hospital resources were the most critical to care for patients in four of the National Planning Scenarios: pandemic influenza, radiation event, explosion, and nerve gas attack 7 .This information is essential for hospital preparedness planners to make supply-chain driven decisions based on the number of patients treated.In addition, we wished to determine which of these critical hospital resources were common across multiple types of events, and distinguish resources that are critical only for specific scenarios.

Methods
We organized and conducted a facilitated national expert consensus panel held during the spring of 2011.
Twenty three expert panelists from various disciplines in the health care sector (Table 1) took part in a one day consensus-building set of exercises.The majority of the participants were purposefully chosen to be healthcare providers, because the decision of what constitutes critical hospital resource for patient care, and issues related to standards of care are primarily clinical.However, we did include allied professionals to ensure other

Introduction
Hospital surge capacity is dependent on the ability to increase or conserve resources, in response to an influx of patients in a disaster situation 1 .Adequate surge capacity depends on the fundamental understanding of which hospital resources are critical to patient care 2 .Surge capacity has four conceptual components, three of which categorize resources; space, staff, and supplies 3 , 4 , 5 .The fourth key component defining surge capacity is the "system" category, which denotes organization, processes, policies and procedures that govern and organize the allocation and conservation of the first three components.
In 2009, the Agency for Healthcare Research and Quality (AHRQ) developed a Hospital Surge Model 6 that forecasted the hospital resources required to treat casualties resulting from 13 National Planning Scenarios 7 .The AHRQ's planning tool was discontinued on June 30, 2011 6 .The objective of our study was to identify which of the AHRQ hospital resources were the most critical to care for patients in four of the National Planning Scenarios: pandemic influenza, radiation event, explosion, and nerve gas attack 7 .This information is essential for hospital preparedness planners to make supply-chain driven decisions based on the number of patients treated.In addition, we wished to determine which of these critical hospital resources were common across multiple types of events, and distinguish resources that are critical only for specific scenarios.

Methods
We organized and conducted a facilitated national expert consensus panel held during the spring of 2011.
Twenty three expert panelists from various disciplines in the health care sector (Table 1) took part in a one day consensus-building set of exercises.The majority of the participants were purposefully chosen to be healthcare providers, because the decision of what constitutes critical hospital resource for patient care, and issues related to standards of care are primarily clinical.However, we did include allied professionals to ensure other perspectives.The study was approved by an Institutional Review Board on Human Subjects Research, and all study participants provided appropriate consent.

Trauma physician 1
Prior to the exercises, the investigators conducted a detailed review of the literature and produced a list of all patient care related resources (space, staff, supplies) that may be required for routine use and for specific disaster scenarios based on four of the national planning scenarios chosen for these exercises (pandemic influenza scenario, radiation exposure, explosives, and nerve gas).The radiation scenario postulates a radiological dispersal device or "dirty bomb" detonated in 3 different sites moderate-to-large cities causing fatalities; 270 injuries; 20,000 detectable contaminations (at each site).The pandemic influenza scenario posits a 15% attack rate: 87,000 fatalities; and 300,000 hospitalizations.In the nerve gas scenario, six spray dissemination devices and releases Sarin vapor into the ventilation systems of three large commercial office buildings in a metropolitan area causing 6,000 fatalities (95% of building occupants).In the explosion scenario, improvised explosive devices (IEDs) were detonated a sports arena, an underground public transportation concourse, and in a parking facility near the entertainment complex causing 100 fatalities and 450 hospitalizations.
Participants were provided key background literature, as well as the URL to the original AHRQ model, and a list of the resources to review two weeks prior to the meeting.The finalized resource list was provided to all participants at a meeting held the evening prior to the day of the exercises.To ensure uniform application of the planned methods the next day, two separate mock exercises (using different national planning scenarios) were conducted at the evening session with the entire group of expert panelists.During that session, participants were also briefed on the logistics of the consensus panel the following day.On the day of the formal exercises, participants were divided into 4 groups of five or six members.Each group was assigned two of the four scenarios., i.e. each of the four scenarios was assigned to two separate groups, in various combinations.Each group had a facilitator to guide the proceedings and discussion.Weeks prior to the study, the PI and the Project Manager chose and trained four experienced facilitators.Panelists in each group were asked to consider 132 hospital patient care resources -extracted from the AHRQ's hospital surge model and the literature -in order to identify the ones that would be critical to patient care.Both pediatric and adult patients were to be considered.The operational definition of a "critical" hospital resource was as follows:

Trauma physician 1
Prior to the exercises, the investigators conducted a detailed review of the literature and produced a list of all patient care related resources (space, staff, supplies) that may be required for routine use and for specific disaster scenarios based on four of the national planning scenarios chosen for these exercises (pandemic influenza scenario, radiation exposure, explosives, and nerve gas).The radiation scenario postulates a radiological dispersal device or "dirty bomb" detonated in 3 different sites moderate-to-large cities causing fatalities; 270 injuries; 20,000 detectable contaminations (at each site).The pandemic influenza scenario posits a 15% attack rate: 87,000 fatalities; and 300,000 hospitalizations.In the nerve gas scenario, six spray dissemination devices and releases Sarin vapor into the ventilation systems of three large commercial office buildings in a metropolitan area causing 6,000 fatalities (95% of building occupants).In the explosion scenario, improvised explosive devices (IEDs) were detonated a sports arena, an underground public transportation concourse, and in a parking facility near the entertainment complex causing 100 fatalities and 450 hospitalizations.
Participants were provided key background literature, as well as the URL to the original AHRQ model, and a list of the resources to review two weeks prior to the meeting.The finalized resource list was provided to all participants at a meeting held the evening prior to the day of the exercises.To ensure uniform application of the planned methods the next day, two separate mock exercises (using different national planning scenarios) were conducted at the evening session with the entire group of expert panelists.During that session, participants were also briefed on the logistics of the consensus panel the following day.On the day of the formal exercises, participants were divided into 4 groups of five or six members.Each group was assigned two of the four scenarios., i.e. each of the four scenarios was assigned to two separate groups, in various combinations.Each group had a facilitator to guide the proceedings and discussion.Weeks prior to the study, the PI and the Project Manager chose and trained four experienced facilitators.Panelists in each group were asked to consider 132 hospital patient care resources -extracted from the AHRQ's hospital surge model and the literature -in order to identify the ones that would be critical to patient care.Both pediatric and adult patients were to be considered.The operational definition of a "critical" hospital resource was as follows:

Trauma physician 1
Prior to the exercises, the investigators conducted a detailed review of the literature and produced a list of all patient care related resources (space, staff, supplies) that may be required for routine use and for specific disaster scenarios based on four of the national planning scenarios chosen for these exercises (pandemic influenza scenario, radiation exposure, explosives, and nerve gas).The radiation scenario postulates a radiological dispersal device or "dirty bomb" detonated in 3 different sites moderate-to-large cities causing fatalities; 270 injuries; 20,000 detectable contaminations (at each site).The pandemic influenza scenario posits a 15% attack rate: 87,000 fatalities; and 300,000 hospitalizations.In the nerve gas scenario, six spray dissemination devices and releases Sarin vapor into the ventilation systems of three large commercial office buildings in a metropolitan area causing 6,000 fatalities (95% of building occupants).In the explosion scenario, improvised explosive devices (IEDs) were detonated a sports arena, an underground public transportation concourse, and in a parking facility near the entertainment complex causing 100 fatalities and 450 hospitalizations.
Participants were provided key background literature, as well as the URL to the original AHRQ model, and a list of the resources to review two weeks prior to the meeting.The finalized resource list was provided to all participants at a meeting held the evening prior to the day of the exercises.To ensure uniform application of the planned methods the next day, two separate mock exercises (using different national planning scenarios) were conducted at the evening session with the entire group of expert panelists.During that session, participants were also briefed on the logistics of the consensus panel the following day.On the day of the formal exercises, participants were divided into 4 groups of five or six members.Each group was assigned two of the four scenarios., i.e. each of the four scenarios was assigned to two separate groups, in various combinations.Each group had a facilitator to guide the proceedings and discussion.Weeks prior to the study, the PI and the Project Manager chose and trained four experienced facilitators.Panelists in each group were asked to consider 132 hospital patient care resources -extracted from the AHRQ's hospital surge model and the literature -in order to identify the ones that would be critical to patient care.Both pediatric and adult patients were to be considered.The operational definition of a "critical" hospital resource was as follows: absence of the resource is likely to have a major impact on patient outcomes, i.e., high likelihood of untoward event, possibly death.
By design, morbidity, mortality and the number of victims for various scenarios have no bearing on the outcomes of interest since the main objective of the study was to determine which specific hospital resources were considered critical to care for patients of all levels of acuity, regardless of the clinical load.In this regard, two major assumptions were communicated to the panelists.First, they were asked to assume that basic hospital/unit operations, such as infrastructure, food, security, housekeeping, laundry, etc., remained intact, and that the hospital building was not degraded structurally in any way.Second, it was assumed that clinical standards of care would be maintained, i.e. panelists were asked to perform the exercise without considering the availability of alternate or compensatory resources that did not maintain equivalent standards of care.The PI of the study and the Project Manager continuously rotated among panels to clarify definitions and methods, and to ensure that each panel was adhering to the protocols, but did not espouse any particular point of view.
For items with any disagreement in ranking within each group, a facilitated discussion (modified Delphi technique) was conducted until consensus (defined as more than 50% agreement) was reached.To help mitigate potential dominance by one or more panel members, rankings by each individual panelist was done separately and in a blinded fashion.Each group was also encouraged to add potential hospital resources not already on the list.Panel members were asked to recuse themselves from exercising an opinion and score on resource items with which they were not familiar.

Analysis
To measure the agreement among all participants on assigning critical resources, intraclass correlation coefficient (ICC) for every item was calculated individually for each scenario and collectively across all scenarios; Kruskal-Wallis test was performed to measure the distribution of scores for each resource across all scenarios.

Ethics Statement
This work was reviewed and approved by the Johns Hopkins School of Medicine Institutional Board and a waiver of consent was approved and documented by the Program Manager.The waiver was given because the research involves no risk to participants as they were invited to participate and could choose at any time not to attend.They were offering their opinion on a subject they are experts in and all of their opinions and all related data were recorded unidentifiable and anonymously.By recording their consent, we would be introducing identifiers.

Results
Of the 132 hospital resources evaluated, 25 were considered critical in all four scenarios by more than 50% of participants (Table 2).There was 90% or more agreement among panelists on 16 of these 25 hospital resources, with agreement ranging from 64.7% to 100%.Crystalloid solution was the only resource that had 100% agreement on being critical, in all four scenarios.
absence of the resource is likely to have a major impact on patient outcomes, i.e., high likelihood of untoward event, possibly death.
By design, morbidity, mortality and the number of victims for various scenarios have no bearing on the outcomes of interest since the main objective of the study was to determine which specific hospital resources were considered critical to care for patients of all levels of acuity, regardless of the clinical load.In this regard, two major assumptions were communicated to the panelists.First, they were asked to assume that basic hospital/unit operations, such as infrastructure, food, security, housekeeping, laundry, etc., remained intact, and that the hospital building was not degraded structurally in any way.Second, it was assumed that clinical standards of care would be maintained, i.e. panelists were asked to perform the exercise without considering the availability of alternate or compensatory resources that did not maintain equivalent standards of care.The PI of the study and the Project Manager continuously rotated among panels to clarify definitions and methods, and to ensure that each panel was adhering to the protocols, but did not espouse any particular point of view.
For items with any disagreement in ranking within each group, a facilitated discussion (modified Delphi technique) was conducted until consensus (defined as more than 50% agreement) was reached.To help mitigate potential dominance by one or more panel members, rankings by each individual panelist was done separately and in a blinded fashion.Each group was also encouraged to add potential hospital resources not already on the list.Panel members were asked to recuse themselves from exercising an opinion and score on resource items with which they were not familiar.

Analysis
To measure the agreement among all participants on assigning critical resources, intraclass correlation coefficient (ICC) for every item was calculated individually for each scenario and collectively across all scenarios; Kruskal-Wallis test was performed to measure the distribution of scores for each resource across all scenarios.

Ethics Statement
This work was reviewed and approved by the Johns Hopkins School of Medicine Institutional Board and a waiver of consent was approved and documented by the Program Manager.The waiver was given because the research involves no risk to participants as they were invited to participate and could choose at any time not to attend.They were offering their opinion on a subject they are experts in and all of their opinions and all related data were recorded unidentifiable and anonymously.By recording their consent, we would be introducing identifiers.

Results
Of the 132 hospital resources evaluated, 25 were considered critical in all four scenarios by more than 50% of participants (Table 2).There was 90% or more agreement among panelists on 16 of these 25 hospital resources, with agreement ranging from 64.7% to 100%.Crystalloid solution was the only resource that had 100% agreement on being critical, in all four scenarios.
absence of the resource is likely to have a major impact on patient outcomes, i.e., high likelihood of untoward event, possibly death.
By design, morbidity, mortality and the number of victims for various scenarios have no bearing on the outcomes of interest since the main objective of the study was to determine which specific hospital resources were considered critical to care for patients of all levels of acuity, regardless of the clinical load.In this regard, two major assumptions were communicated to the panelists.First, they were asked to assume that basic hospital/unit operations, such as infrastructure, food, security, housekeeping, laundry, etc., remained intact, and that the hospital building was not degraded structurally in any way.Second, it was assumed that clinical standards of care would be maintained, i.e. panelists were asked to perform the exercise without considering the availability of alternate or compensatory resources that did not maintain equivalent standards of care.The PI of the study and the Project Manager continuously rotated among panels to clarify definitions and methods, and to ensure that each panel was adhering to the protocols, but did not espouse any particular point of view.
For items with any disagreement in ranking within each group, a facilitated discussion (modified Delphi technique) was conducted until consensus (defined as more than 50% agreement) was reached.To help mitigate potential dominance by one or more panel members, rankings by each individual panelist was done separately and in a blinded fashion.Each group was also encouraged to add potential hospital resources not already on the list.Panel members were asked to recuse themselves from exercising an opinion and score on resource items with which they were not familiar.

Analysis
To measure the agreement among all participants on assigning critical resources, intraclass correlation coefficient (ICC) for every item was calculated individually for each scenario and collectively across all scenarios; Kruskal-Wallis test was performed to measure the distribution of scores for each resource across all scenarios.

Ethics Statement
This work was reviewed and approved by the Johns Hopkins School of Medicine Institutional Board and a waiver of consent was approved and documented by the Program Manager.The waiver was given because the research involves no risk to participants as they were invited to participate and could choose at any time not to attend.They were offering their opinion on a subject they are experts in and all of their opinions and all related data were recorded unidentifiable and anonymously.By recording their consent, we would be introducing identifiers.

Results
Of the 132 hospital resources evaluated, 25 were considered critical in all four scenarios by more than 50% of participants (Table 2).There was 90% or more agreement among panelists on 16 of these 25 hospital resources, with agreement ranging from 64.7% to 100%.Crystalloid solution was the only resource that had 100% agreement on being critical, in all four scenarios.Out of the 132 hospital resources, the number considered critical by more than 50% of participants varied from one scenario to another; 58 (44%) for the pandemic influenza scenario, 51 (39%) for radiation exposure, 41 (31%) for explosives, and 35 (27%) for nerve gas scenario (Table 3).There were an additional 10 scenariospecific hospital resources considered critical by all participants (100% agreement), that were not considered as generally critical for all scenarios.These 10 resources were distributed as follows: 5 in the pandemic influenza scenario (isolation room/cohorting, respiratory therapist, facemasks, antiviral agents for influenza, and dialysis); 4 in the radiation scenario (radiation specialist, potassium iodide, Geiger counter, and decontamination capability); 1 in the explosives scenario (standard radiograph), and none specific for nerve agents.There were a number of resources that were specific to a single scenario (Table 3) as follows: pandemic influenza (isolation room/cohorting, respiratory therapist, facemasks, antiviral agents for influenza, and dialysis); radiation (radiation specialist, potassium iodide, Geiger counter, decontamination capability and isotope chelating Out of the 132 hospital resources, the number considered critical by more than 50% of participants varied from one scenario to another; 58 (44%) for the pandemic influenza scenario, 51 (39%) for radiation exposure, 41 (31%) for explosives, and 35 (27%) for nerve gas scenario (Table 3).There were an additional 10 scenariospecific hospital resources considered critical by all participants (100% agreement), that were not considered as generally critical for all scenarios.These 10 resources were distributed as follows: 5 in the pandemic influenza scenario (isolation room/cohorting, respiratory therapist, facemasks, antiviral agents for influenza, and dialysis); 4 in the radiation scenario (radiation specialist, potassium iodide, Geiger counter, and decontamination capability); 1 in the explosives scenario (standard radiograph), and none specific for nerve agents.There were a number of resources that were specific to a single scenario (Table 3) as follows: pandemic influenza (isolation room/cohorting, respiratory therapist, facemasks, antiviral agents for influenza, and dialysis); radiation (radiation specialist, potassium iodide, Geiger counter, decontamination capability and isotope chelating Out of the 132 hospital resources, the number considered critical by more than 50% of participants varied from one scenario to another; 58 (44%) for the pandemic influenza scenario, 51 (39%) for radiation exposure, 41 (31%) for explosives, and 35 (27%) for nerve gas scenario (Table 3).There were an additional 10 scenariospecific hospital resources considered critical by all participants (100% agreement), that were not considered as generally critical for all scenarios.These 10 resources were distributed as follows: 5 in the pandemic influenza scenario (isolation room/cohorting, respiratory therapist, facemasks, antiviral agents for influenza, and dialysis); 4 in the radiation scenario (radiation specialist, potassium iodide, Geiger counter, and decontamination capability); 1 in the explosives scenario (standard radiograph), and none specific for nerve agents.There were a number of resources that were specific to a single scenario (Table 3) as follows: pandemic influenza (isolation room/cohorting, respiratory therapist, facemasks, antiviral agents for influenza, and dialysis); radiation (radiation specialist, potassium iodide, Geiger counter, decontamination capability and isotope chelating agents); explosion (surgeon, packed red blood cells, Silvadene cream, gauze pads, fresh-frozen plasma, enteral feeding tubes); and nerve agent (Atropine and 2-PAM).
Agreement on all 132 resources was strong for the pandemic influenza (ICC= 0.

Discussion
In the face of natural and man-made disasters, hospitals are at the forefront providing medical care for patients at all levels of acuity.While there has been considerable work related to surge capacity  35 , detailed accounting of individual hospital resources that are critical for response to specific types of disasters is lacking.Hospital administrators and disaster managers must identify the critical medical resources in all likely scenarios, and design robust storage and supply-chain protocols to maintain adequate supplies of these resources.
In 2009, Hick et al. categorized surge capacity according to three levels of graded response: "conventional," "contingent," or "crisis" level management.[2] A conventional response implies that routinely available resources can address the requirements imposed by the event, and standards of care can be maintained.
Contingent response requires mobilization of additional measures to those routinely in place, while standards of care are essentially maintained.Crisis response requires considerable alteration in routine operations and available measures to meet the patient surge requirements, and standards of care cannot be maintained.We believe our study builds upon the model proposed by Hick et al., by identifying resources considered critical within the boundaries of conventional and contingency surge capacities.
The 25 resources identified as critical -by consensus -for all 4 scenarios (pandemic influenza, radiation, explosives, and nerve agents) cover the three categories of hospital surge capacity resources; supplies ( . In the supplies category, intravenous crystalloid solution was the only resource unanimously ranked by participants as critical in all 4 scenarios.The rationale behind this categorization may be related to the ubiquitous use of crystalloid solutions in acute medical conditions resulting from various disaster scenarios.Other common supplies with consensus as being "critical," include intubation equipment, ventilators, oxygen, sedatives, IV catheters, needles, gloves, and blood pressure cuffs.As expected, some resources are highly specific and critical for response to a certain event.For example, for a disaster with radiation exposure, potassium iodide and a radiation expert were considered critical by 100% of the participants.In the space category, adult and pediatric intensive care and medical/surgical beds were identified as critical resources for all four scenarios.These represent the main physical areas accommodating critical and moderately injured patients requiring admission.In the staff category, critical care physicians and nurses, in addition to non-critical care nurses, were considered critical across all scenarios.Interestingly, non-critical care physicians (e.g.internal medicine physicians) were considered critical by less than 50% of panelists in all four study scenarios.We note that our study panelists placed emphasis on specialized care in disaster response.This is understandable given the orientation of health care practice in the U.S., even during disasters.However, in many countries, primary care resources may play a larger role and would likely be judged as reaching critical hospital resource status in many disaster scenarios.

Discussion
In the face of natural and man-made disasters, hospitals are at the forefront providing medical care for patients at all levels of acuity.While there has been considerable work related to surge capacity The 25 resources identified as critical -by consensus -for all 4 scenarios (pandemic influenza, radiation, explosives, and nerve agents) cover the three categories of hospital surge capacity resources; supplies (18), space (4), and staff (3).In the supplies category, intravenous crystalloid solution was the only resource unanimously ranked by participants as critical in all 4 scenarios.The rationale behind this categorization may be related to the ubiquitous use of crystalloid solutions in acute medical conditions resulting from various disaster scenarios.Other common supplies with consensus as being "critical," include intubation equipment, ventilators, oxygen, sedatives, IV catheters, needles, gloves, and blood pressure cuffs.As expected, some resources are highly specific and critical for response to a certain event.For example, for a disaster with radiation exposure, potassium iodide and a radiation expert were considered critical by 100% of the participants.In the space category, adult and pediatric intensive care and medical/surgical beds were identified as critical resources for all four scenarios.These represent the main physical areas accommodating critical and moderately injured patients requiring admission.In the staff category, critical care physicians and nurses, in addition to non-critical care nurses, were considered critical across all scenarios.Interestingly, non-critical care physicians (e.g.internal medicine physicians) were considered critical by less than 50% of panelists in all four study scenarios.We note that our study panelists placed emphasis on specialized care in disaster response.This is understandable given the orientation of health care practice in the U.S., even during disasters.However, in many countries, primary care resources may play a larger role and would likely be judged as reaching critical hospital resource status in many disaster scenarios.

Discussion
In the face of natural and man-made disasters, hospitals are at the forefront providing medical care for patients at all levels of acuity.While there has been considerable work related to surge capacity The 25 resources identified as critical -by consensus -for all 4 scenarios (pandemic influenza, radiation, explosives, and nerve agents) cover the three categories of hospital surge capacity resources; supplies (18), space (4), and staff (3).In the supplies category, intravenous crystalloid solution was the only resource unanimously ranked by participants as critical in all 4 scenarios.The rationale behind this categorization may be related to the ubiquitous use of crystalloid solutions in acute medical conditions resulting from various disaster scenarios.Other common supplies with consensus as being "critical," include intubation equipment, ventilators, oxygen, sedatives, IV catheters, needles, gloves, and blood pressure cuffs.As expected, some resources are highly specific and critical for response to a certain event.For example, for a disaster with radiation exposure, potassium iodide and a radiation expert were considered critical by 100% of the participants.In the space category, adult and pediatric intensive care and medical/surgical beds were identified as critical resources for all four scenarios.These represent the main physical areas accommodating critical and moderately injured patients requiring admission.In the staff category, critical care physicians and nurses, in addition to non-critical care nurses, were considered critical across all scenarios.Interestingly, non-critical care physicians (e.g.internal medicine physicians) were considered critical by less than 50% of panelists in all four study scenarios.We note that our study panelists placed emphasis on specialized care in disaster response.This is understandable given the orientation of health care practice in the U.S., even during disasters.However, in many countries, primary care resources may play a larger role and would likely be judged as reaching critical hospital resource status in many disaster scenarios.
The differences in agreement among participants regarding all 132 resources, when considering specific scenarios, may be grounded in the degree of clinical and operational familiarity with the specific scenario in question.For example, influenza epidemics occur regularly, so both clinicians and hospital administrators are more familiar with the required critical resources (ICC=0.7),thanthose needed in a radiation disaster (ICC=0.5).
The other interesting finding of variation in the number of critical resources among different scenarios (e.g.58 for pandemic influenza and 35 for nerve gas), is likely due to the clinical complexity of the scenario, in addition to the notion of familiarity mentioned above.
To the best of our knowledge, the derivation and importance ranking of 132 discrete potential hospital resources, is the only collective attempt by subject matter experts to stratify the various hospital-based provisions required for the care of victims during the four scenarios considered in this study.The expert panelists noted which of these hospital resources were critical, and that their absence represented a major risk of an untoward medical outcome.This information enables planners to further examine, describe and delineate their surge capacity beyond simple bed availability.

Limitations
There are several limitations to our study.First, the study is based on an expert panel consensus and not on a specific functional exercise or disaster event.Second, the size of each group (sub-panel) of experts was not very large (five or six members).Still, two separate panels, i.e., 50% of the available panel members, considered each scenario.Third, group composition may not be comprehensive in representing the numerous specialties and sub-specialties that would typically take part in responding to the selected disaster scenarios.
Representation from every type of hospital may also be lacking (e.g.rural hospitals were poorly represented).
While we were thoughtful and deliberate regarding the composition of each group -seeking to optimize representation from a broad national pool of expert practitioners -this limitation may affect the generalizability of our finding.Finally, due to the objectives of the study, the results are applicable to only four national planning scenarios.To mitigate this limitation, we are planning on conducting further studies to explore identifying critical resources for the remaining national planning scenarios.

Conclusion
Twenty five hospital resources were found to be critical to maintain continuity medical care in four disaster planning scenarios; namely radiation, pandemic influenza, explosives, and nerve gas scenarios.However, some specific disaster scenarios require additional critical specialized resources necessary for the corresponding type of disaster.Planning for each hospital should be dictated by the hazard vulnerability analysis, gauging their vulnerabilities within the environment, in order to prioritize and maintain adequate supplies of scenario-specific critical hospital resources.Further studies are needed in the field of hospital surge capacity, to validate these findings, determine utilization rates for each of the resources during a surge event, and to identify appropriate alternatives to these critical hospital resources 36 .
The differences in agreement among participants regarding all 132 resources, when considering specific scenarios, may be grounded in the degree of clinical and operational familiarity with the specific scenario in question.For example, influenza epidemics occur regularly, so both clinicians and hospital administrators are more familiar with the required critical resources (ICC=0.7),thanthose needed in a radiation disaster (ICC=0.5).
The other interesting finding of variation in the number of critical resources among different scenarios (e.g.58 for pandemic influenza and 35 for nerve gas), is likely due to the clinical complexity of the scenario, in addition to the notion of familiarity mentioned above.
To the best of our knowledge, the derivation and importance ranking of 132 discrete potential hospital resources, is the only collective attempt by subject matter experts to stratify the various hospital-based provisions required for the care of victims during the four scenarios considered in this study.The expert panelists noted which of these hospital resources were critical, and that their absence represented a major risk of an untoward medical outcome.This information enables planners to further examine, describe and delineate their surge capacity beyond simple bed availability.

Limitations
There are several limitations to our study.First, the study is based on an expert panel consensus and not on a specific functional exercise or disaster event.Second, the size of each group (sub-panel) of experts was not very large (five or six members).Still, two separate panels, i.e., 50% of the available panel members, considered each scenario.Third, group composition may not be comprehensive in representing the numerous specialties and sub-specialties that would typically take part in responding to the selected disaster scenarios.
Representation from every type of hospital may also be lacking (e.g.rural hospitals were poorly represented).
While we were thoughtful and deliberate regarding the composition of each group -seeking to optimize representation from a broad national pool of expert practitioners -this limitation may affect the generalizability of our finding.Finally, due to the objectives of the study, the results are applicable to only four national planning scenarios.To mitigate this limitation, we are planning on conducting further studies to explore identifying critical resources for the remaining national planning scenarios.

Conclusion
Twenty five hospital resources were found to be critical to maintain continuity medical care in four disaster planning scenarios; namely radiation, pandemic influenza, explosives, and nerve gas scenarios.However, some specific disaster scenarios require additional critical specialized resources necessary for the corresponding type of disaster.Planning for each hospital should be dictated by the hazard vulnerability analysis, gauging their vulnerabilities within the environment, in order to prioritize and maintain adequate supplies of scenario-specific critical hospital resources.Further studies are needed in the field of hospital surge capacity, to validate these findings, determine utilization rates for each of the resources during a surge event, and to identify appropriate alternatives to these critical hospital resources 36 .
The differences in agreement among participants regarding all 132 resources, when considering specific scenarios, may be grounded in the degree of clinical and operational familiarity with the specific scenario in question.For example, influenza epidemics occur regularly, so both clinicians and hospital administrators are more familiar with the required critical resources (ICC=0.7),thanthose needed in a radiation disaster (ICC=0.5).
The other interesting finding of variation in the number of critical resources among different scenarios (e.g.58 for pandemic influenza and 35 for nerve gas), is likely due to the clinical complexity of the scenario, in addition to the notion of familiarity mentioned above.
To the best of our knowledge, the derivation and importance ranking of 132 discrete potential hospital resources, is the only collective attempt by subject matter experts to stratify the various hospital-based provisions required for the care of victims during the four scenarios considered in this study.The expert panelists noted which of these hospital resources were critical, and that their absence represented a major risk of an untoward medical outcome.This information enables planners to further examine, describe and delineate their surge capacity beyond simple bed availability.

Limitations
There are several limitations to our study.First, the study is based on an expert panel consensus and not on a specific functional exercise or disaster event.Second, the size of each group (sub-panel) of experts was not very large (five or six members).Still, two separate panels, i.e., 50% of the available panel members, considered each scenario.Third, group composition may not be comprehensive in representing the numerous specialties and sub-specialties that would typically take part in responding to the selected disaster scenarios.
Representation from every type of hospital may also be lacking (e.g.rural hospitals were poorly represented).
While we were thoughtful and deliberate regarding the composition of each group -seeking to optimize representation from a broad national pool of expert practitioners -this limitation may affect the generalizability of our finding.Finally, due to the objectives of the study, the results are applicable to only four national planning scenarios.To mitigate this limitation, we are planning on conducting further studies to explore identifying critical resources for the remaining national planning scenarios.

Conclusion
Twenty five hospital resources were found to be critical to maintain continuity medical care in four disaster planning scenarios; namely radiation, pandemic influenza, explosives, and nerve gas scenarios.However, some specific disaster scenarios require additional critical specialized resources necessary for the corresponding type of disaster.Planning for each hospital should be dictated by the hazard vulnerability analysis, gauging their vulnerabilities within the environment, in order to prioritize and maintain adequate supplies of scenario-specific critical hospital resources.Further studies are needed in the field of hospital surge capacity, to validate these findings, determine utilization rates for each of the resources during a surge event, and to identify appropriate alternatives to these critical hospital resources 36 .

Table 1 .
Expert panel composition (in alphabetical order)

Table 1 .
Expert panel composition (in alphabetical order)

Table 2 -
Critical hospital resources common to all four scenarios *Significant difference in score distribution among the four scenarios

Table 2 -
Critical hospital resources common to all four scenarios *Significant difference in score distribution among the four scenarios

Table 2 -
Critical hospital resources common to all four scenarios *Significant difference in score distribution among the four scenarios

Table 3 -
Scenario-Specific Critical Hospital Resources 1 , 2 , 3 , 4 , 5 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , and national recommendations to itemize critical healthcare resources35, detailed accounting of individual hospital resources that are critical for response to specific types of disasters is lacking.Hospital administrators and disaster managers must identify the critical medical resources in all likely scenarios, and design robust storage and supply-chain protocols to maintain adequate supplies of these resources.In 2009, Hick et al. categorized surge capacity according to three levels of graded response: "conventional," "contingent," or "crisis" level management.[2]Aconventionalresponse implies that routinely available resources can address the requirements imposed by the event, and standards of care can be maintained.Contingent response requires mobilization of additional measures to those routinely in place, while standards of care are essentially maintained.Crisis response requires considerable alteration in routine operations and available measures to meet the patient surge requirements, and standards of care cannot be maintained.We believe our study builds upon the model proposed by Hick et al., by identifying resources considered critical within the boundaries of conventional and contingency surge capacities.
1 , 2 , 3 , 4 , 5 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , and national recommendations to itemize critical healthcare resources35, detailed accounting of individual hospital resources that are critical for response to specific types of disasters is lacking.Hospital administrators and disaster managers must identify the critical medical resources in all likely scenarios, and design robust storage and supply-chain protocols to maintain adequate supplies of these resources.In 2009, Hick et al. categorized surge capacity according to three levels of graded response: "conventional," "contingent," or "crisis" level management.[2]Aconventionalresponse implies that routinely available resources can address the requirements imposed by the event, and standards of care can be maintained.Contingent response requires mobilization of additional measures to those routinely in place, while standards of care are essentially maintained.Crisis response requires considerable alteration in routine operations and available measures to meet the patient surge requirements, and standards of care cannot be maintained.We believe our study builds upon the model proposed by Hick et al., by identifying resources considered critical within the boundaries of conventional and contingency surge capacities.