All official NCVHS documents including meeting transcripts are posted on the NCVHS website (http://ncvhs.roseliassociates.com)

Department of Health and Human Services

NATIONAL COMMITTEE ON VITAL AND HEALTH STATISTICS

Subcommittee on Populations

February 14, 2007

Hubert Humphrey Building
200 Independence Avenue
Washington, D.C.

Meeting Summary

Data Needs for Assessing Surge Capacity and Hospital Preparedness


The National Committee on Vital and Health Statistics Subcommittee on Populations Workshop on Data Needs for Assessing Surge Capacity and Hospital Preparedness was convened on February 14, 2007 in Washington, D.C.

All official NCVHS documents are posted on the NCVHS Website.

Present

Committee Members

  • Donald M. Steinwachs, Ph.D., Chair (by telephone)
  • Larry A. Green, M.D.
  • Garland Land, MPH
  • William J. Scanlon, Ph.D.
  • C. Eugene Steuerle, Ph.D.
  • Kevin C. Vigilante, M.D., MPH

Absent

Staff and Liaisons

  • Dale Hitchcock, ASPE, Lead Staff
  • Nancy Breen, Ph.D., NCI, NIH
  • Audrey Burwell, OMH
  • Justine Carr, M.D., Chair, Workgroup on Quality
  • Leslie Cooper, Ph.D., NCI, NIH
  • Miryam Granthon, OPHS
  • Marjorie Greenberg, NCHS/CDC
  • Suzanne Haynes, Ph.D., OPHS
  • Debbie M. Jackson, NCHS
  • Cille Kennedy, Ph.D., ASPE
  • Jacqueline Lucas, NCHS
  • Edna Paisano, HIS
  • Susan Queen, Ph.D., HRSA
  • James Scanlon, OSDP, ASPE
  • Harvey Schwartz, Ph.D., AHRQ

Others

  • Terry A. Adirim, M.D., M.P.H.
  • Erik Auf der Heide, MD, MPH, FACEP; CDC, ATSDR
  • Douglas Boenning, M.D., ASPE
  • John Donohue, MIEMSS
  • Roland Gamache, Ph.D., MBA, Indiana State Health Dept.
  • Chris Handley, MIEMSS
  • Til Jolly, M.D., DHS
  • David Marcozzi, M.D., HHS Office of Preparedness and Emergency Operations
  • Andrew Miller, ARES Corporation
  • Patrick Oliver, Medical College of Virginia
  • Greg A. Pane, M.D., Washington DC Dept. of Health
  • Beverly Pritchett, Washington DC Dept. of Health
  • Melissa Sanders, HRSA (by telephone)
  • Janet Schiller, HRSA/HSB/DHP
  • Jack Snyder, M.D., NLM

ACTIONS

Action Steps

Dr. Vigilante plans to: process the information of the meeting; take key points to the Committee; determine next steps; and reestablish contact with the meeting participants.

Dr. Boenning requested input via email or telephone calls about next steps from participants in order to move forward from this “pilot” meeting.

EXECUTIVE SUMMARY

NOTE: Please refer to Detailed Summary, Transcript, and PowerPoint presentations for further information and examples.

Overview of the Data Challenge James Scanlon, OSDP, ASPE

Noting the impact of September 11, 2001, anthrax scares, the response efforts for Hurricanes Katrina and Rita, an ensuing emphasis on preparedness for all threats and hazards, and the recent focus on pandemic preparedness, HHS is reviewing the entire national response and preparedness apparatus in the United States. The National Response Plan’s (NRP) first response is handled at the community level, followed by help from the state, the region, and the federal government (in that order) for areas of health care, public health, medical support, and human services, as needed. HHS has conducted a literature review about what constitutes hospital preparedness and has awarded cooperative grants to states for hospital preparedness. The GAO has a study to assess performance measurement in preparedness arenas. Consideration of a certification process was mentioned.

PANEL I FEDERAL PERSPECTIVE

Ongoing Federal Surveys and Available Data

Catharine Burt, Ed.D., Richard Niska, M.D. NCHS

NCHS’s nationally representative annual survey (which allows for national estimates) is a good planning tool for modeling disaster scenarios. The National Hospital Ambulatory Medical Care Survey (NHAMCS) has a 700 hospital sample. Two facility survey supplements have added questions about staffing capacity, ambulance diversion, bioterrorism, and mass casualty preparedness relative to emergency department [ED] and outpatient department [OPD] visits. While approximately 80 percent of the surveyed hospitals had mass casualty emergency response plans, there was a disparity between the plans and the memorandums of understanding (MOUs). Questions were posed about plans for stockpiling antibiotics and supplies and about how best to distribute resources to urban verses rural areas. Techniques for converting use of space or activating unused space were mentioned. It was noted that a third of the hospitals and somewhat less than half of the rural hospitals did not know their National Disaster Medical Systems (NDMS) status. While demand on emergency departments and crowding have increased in the last ten years, the number of EDs has declined (note June 2006 IOM report). Diversion status statistics were presented. The projection of treatment spaces/beds provides a model that helps to determine needed beds and the impact of different expected scenarios.

Conclusions (as noted in PowerPoint presentation)

  • Crowding in EDs currently limits the ability to handle an influx of cases
  • Disparity exists across regions in ability to handle a surge due to natural or man-made disasters
  • Surge capacity planning for specific elements found in ¼ – ¾ of hospitals
  • Observed disparity favors urban hospitals in emergency planning and resources: urban hospitals tended to have more written plans and resources for dealing with disasters compared to non-metropolitan hospitals.

Discussion

Survey data can improve hospital surge capacity. Ambulance diversion and the “burden of reporting” were discussed. It is important to understand the purpose of collecting specific data. The complexities of including a community perspective in survey questions were described. Since the 2003-2004 survey data, there has been a large funding increase for the National Hospital Bioterrorism program, resulting in more current data in such areas as decontamination expenditures and personal protective equipment (PPE). Ms. Schiller reported that a system is being developed that collects data twice/year, analyzes and makes it available for decision-making.

A brief discussion followed about the value of a distance to care measure. It was recommended that the group distinguish between prospective data collection (e.g., ambulance diversion) and the ability to create real time situational awareness, critical for surge capacity and a facility’s ability to forecast what may be coming. Some real time questions will be asked on the 2007 NCHS survey. NCHS uses JCAHO hospital emergency response plan requirements and input from experts to formulate survey questions. It was suggested that future data collection be tied to CDC and HRSA guidance on bioterrorism. Collaboration between hospitals and other community facilities was another topic. Hospital preparedness data (from 2003-2004) represent baseline data prior to federal intervention. It was noted that funders (e.g., HRSA) generally elicit a different response than NCHS, which does not provide grants. A discussion about accreditation ensued, noting that fewer unaccredited hospitals were likely to have done preparedness drilling and training. Some data collection difficulties arise because response plans are developed to meet specific needs of particular communities that are not comparable. The National Health Security Strategy, which requires a unifying preparedness strategy from states and hospitals, was delineated.

DHS Office of Health Affairs and Our Role in Interagency Medical Planning

Terry Adirim, M.D., MPH Office of Health Affairs, Department of Homeland Security

A brief history as well as mission and goals of the Office of Health Affairs (OHA) within DHS were presented. OHA’s three divisions were described (i.e., weapons of mass destruction; biodefense; and occupational health) as was its efforts to gather data about early warnings for infectious disease outbreaks. The National Biosurveillance Integration System (NBIS), to be housed at OHA-DHS, collects data in one central location. In general, OHA-DHS is a policy and planning unit that serves as an integrator among agencies and coordinator of medical disaster planning. How to locate and obtain relevant data are key issues for OHA-DHS because the Office resides outside of HHS. See PowerPoint presentation for more information about weapons of mass destruction (WMD) and biodefense; medical readiness; data needs; and OHA’s contributions to the DHS mission.

Case Study: Value of the Electronic Health Record in the VA

Michael Vojtasko, VHA

The working relationship between the Department of Veteran Affairs (VA) and the Department of Defense (DoD) in wartime was described, as was the National Disaster Medical System [NDMS] (see transcript for examples of how the system worked with Hurricanes Katrina and Rita). Additionally, the Joint Patient Tracking Application (JPTA) and federal coordinating centers were portrayed. A 2003-2004 survey indicated that a third of the hospitals did not know that they were NDMS facilities although a new MOU (updated in late 2006) between DHS, HHS, VA, and DoD requires the execution of new memoranda with all NDMS hospitals. Primary FCC responsibilities and the Global Patient Movement Requirements Center (GPMRC) were delineated. A bed casualty reporting system called Transcom Regulating and Command & Control Evacuation System (TRAC2ES) is used with JPTA to track patients through an evaluation system to hospitals. The need for a fully integrated system that can track patients being evacuated to their reception facility, to include coordination of arrival via various modes and systems at the local level was emphasized.

Comments by Discussant Art Kellermann, M.D. Emory University

Dr. Kellermann noted that the best way to ensure a system’s ability to function under extreme circumstances is to ensure efficient and effective functioning on a daily basis (which is not currently the case). Data and metrics should manage the system. See transcript for examples that illustrate these points and the consequences to patients of systems that lack capacity. He suggested a pilot program of Ambulance Traffic Control Centers. He cited Maryland and Northern Virginia as good examples of preparedness systems that are largely effective. It was noted that the regional hospital coordinating center in Northern Virginia began as a medical communications clearinghouse called MedCom.

Discussion

Dr. Kellermann indicated the need for a uniform platform, set of functions, and data collection from the proposed Ambulance Traffic Control Centers that would involve community partnership, local funding, and a possible partnership with the state. The ability of such centers to control patients during a mass casualty surge was raised. The proposed system would work for severely ill or injured patients requiring transport, noting that Israel has developed this capacity.

The need for real time data was again mentioned, noting that current data shows most disasters in the United States to be characterized by a maldistribution of casualties. Especially for critically ill or injured people transported by ambulances, it is important to know which hospitals are overloaded or incapacitated in order to avert ambulance traffic to those sites (e.g., Los Angeles example). The need to educate the hospital EMS community and patients about patient transfer was supported.

PANEL II STATE AND LOCAL PERSPECTIVE

Regional Trauma Center Preparedness and HAvBED

Dan Hanfling, M.D., INOVA Fairfax Hospital, Virginia

A real time surge capacity hospital plan (i.e., Virginia’s 1999 ice storm) was described. Situational awareness provides hospitals with the ability to make necessary choices during times of increased demand for patient care and to match available medical resources to patient needs. Hospitals that do not adhere to situational awareness will be “closed for business.” Matching resources are critical for sustained events like biological attack or emerging infectious disease.

The Northern Virginia Hospital Alliance, which coordinates emergency response disciplines of all regional hospitals under the HRSA division, was described. Borrowing ideas from the Maryland Institute for Emergency Medical Services Systems (MIEMSS) and the Washington DC Hospital Association, Virginia developed a regional clearinghouse following 9-11. Northern Virginia’s Ambulance Diversion program, which includes information about the use of the Medical Surge Capacity and Capability (MSCC) Management System, was described. The area’s regional preparedness MOU was delineated.

The question of how hospital capability information becomes available in real time was raised. The HAvBED Project was described. A model for alternative care delivery may need to be developed if demand cannot be estimated and supply calibrated or improved within the current structure. The need to communicate and get information to communities and citizens using local radio, national cable networks, satellite radio, and television networks was stressed.

Discussion

DoD’s involvement in preparedness discussions was recommended for their expertise in real time data management. It was suggested that the concepts of “emergency department” and “emergency beds” be expanded. The integration of alternative care facility development into emergency/incident management was suggested. A systems perspective is needed.

Consideration of surge capacity in community-based non-hospital medical facilities was recommended because of the “overwhelming” number of casualties that could be treated in non-hospital settings. To decrease hospitalizations, people must be reunited with their routine sources of medical care; and the needs of evacuees must be anticipated when doing surge capacity. To that end, data about the prevalence of chronic diseases and ongoing diseases within communities can help to determine needed resources for people who lose access to routine care. The caregiver is the key metric and the “rate limiting step” in the ability to surge. It was suggested that post-disaster data that determines how many have gone to hospitals due to inaccessible routine care, would also be useful.

Hospital Preparedness: Many Things to Many People

James Bentley, Ph.D., AHA

Decreased funding for data collection in the healthcare system was discussed. Can demand be matched to resources? A model of demand management is needed. Questions were raised about non-recoverable patients; the difference between a military and a civilian definition of triage; self-care; distributed ambulatory care; shelter capabilities; and hospital resources. Staff rather than beds must be used as the rate limiting step. Utilities and a system for their transport are essential to maintaining capacity. It may be more important to have data about the resource supply chain than about patients. During disasters, people often go to hospitals when they don’t need to. Five observations about data include:

  1. Know why data is being collected
  2. Determine usefulness of data to submitter
  3. Ensure consistency of data collection process with that of the national plan
  4. Data should be a by-product of routine operations
  5. Simpler is better – in most disasters, sophistication is “our enemy”

Discussion

The AHA annual survey was described. The reporting burden of many requests for hospital data was discussed. The Hospital Quality Alliance is working to ensure that consistent data requests, answered once, satisfy all interested parties. The HRSA hospital bioterrorism preparedness program was originally conceived to harmonize preparedness data collection although “the vision of the whole is lost” as a result of state and local modifications. There is also “enormous tension” between what different levels of government want. Two major hospital surveys about quality of patient experience omitted emergency department care, supporting the notion that some of what needs to be measured for preparedness (as well as life and death care) is not taking place.

Most health departments and local hospital associations have measured preparedness for several years (noting New York State’s Critical Asset Survey and a Washington DC survey)

Dr. Pritchett described her role and hospital experience in Iraq, which included weekly MASCAL (mass casualty exercises) and many emergency experiences. Most large cities in the United States are operating at disaster level every day. An overarching system is needed to control the big picture. An intermediate case was described that exists between everyday hospital use and disaster use (e.g., 1996 Atlanta hurricane example). The importance of maximizing the number of people doing self-care was emphasized.

Public Health Informatics Needs at the State Level

Roland Gamache, Ph.D., MBA Indiana State Health Department

A state perspective about surge was introduced relative to surveillance and outbreak management, noting Indiana’s Newborn Screening System and Public Health Emergency Surveillance System [PHESS]). Health Information Exchanges (HIEs), which use electronic clinical and medication data, were described. Indiana’s public health success stories included their newborn screening process; electronic transmission of reportable laboratory data and chief complaint data; and the first phase of Indiana’s National Electronic Disease Surveillance System (NEDSS). The process of coding and analyzing emergency department chief complaint data was described (note PowerPoint presentation). Indiana is working to improve reporting of discharges.

State surveillance systems are helpful to the study of surge capacity. While developing a surveillance system that provides real time diversion to ambulances is very expensive, it allows for diversion in a timely manner and helps with outbreak management. Indiana’s interstate sharing of emergency department data was described as was its work to integrate ambulance with emergency department data. Staffing available beds continues to be a big challenge for the Indiana and ISDH. The use of admissions and discharge hospital systems to monitor census is being explored as a pilot project. Within a year, Indiana will be able to capture pre-hospital electronic clinical information.

Discussion

While Indiana hospitals can currently get outbreak information, not every place has the capability or wants that burden. A protocol for potential outbreak situations was suggested. In Indiana, alerting systems for most local health departments are connected with others in the pipeline. Phone calls are still the best communication tool within emergency departments. It was observed that Dr. Gamache’s presentation provided a good example of syndromic surveillance.

Northern Virginia has a real time capability to talk by radio with hospitals in the District of Columbia and Maryland. The need to produce information in real time and to collect data using two-way radios was stressed because telephones and cell phones are unreliable in disaster situations. SYRIS, a similar system used in the Southwest, was mentioned. Amateur radio services were noted as an additional resource.

Maryland’s communications hub (SYSCOM), which sits at MIEMSS, facilitates communication between the paramedic and the hospital. This process includes an ability to talk from field to hospital and a running status of the County Hospital Alert Tracking System [CHATS]. The Facility Resource Emergency Database (FRED) program was described, noting that the National Disaster Medical System (NDMS) of Maryland, Delaware, and Pennsylvania filters through the FRED program. The EMRC program was also delineated.

The difference between dynamic and static information (such as what appears on their web-based emergency management application [EOC]) was noted. The EOC helps to communicate situational awareness and to assign resources. Patient tracking and practitioner availability are other issues to address. Patient tracking systems were described, with an emphasis on Maryland’s system.

Comments by Discussant Sally Phillips, R.N., Ph.D. AHRQ

A standard is needed for data sharing and retrieval between various data systems. It is important to “tweak all those systems” to allow for data exchange and to draw real time or static data into a common pool. A pool of data must be available for research intended to improve emergency care as it is important to have a minimum data base collection at the time of a disaster. These systems must be used to help allocate resources. A better data system is needed to retrieve information about what a system looked like before a disastrous event and how it was impacted in order to anticipate aftershocks. How to define and measure the “steady state” is a major challenge.

Discussion

Dr. Vigilante reiterated the importance of: finding ways to collect data prospectively in order to understand where vulnerabilities lie; collecting data from a research perspective to bolster the evidence base for best practices; and gathering real time data to guide decision-making. Attempts to reduce that burden on those who report the data must accompany these efforts.

Note: The PowerPoint presentation entitled “The Role of Non-Hospital Medical Practitioners in Disasters” by Erik Auf der Heide, MD, MPH, FACEP (CDC, ATSDR). Dr. Auf der Heide participated in the meeting via telephone.

General Discussion and Plans for the Future

The Subcommittee on Populations needs to identify economically feasible strategies with “operational values” for the different types of data needs. What capability should a community have when dealing with disasters? The importance of defining in advance and assessing systematic readiness or capacity was stressed. Dr. Vigilante wondered if the understanding is on the demand or the supply side. It was noted that advances in health information technology that are interoperable and affordable will improve all health care. The need for retrospective data collection was reiterated. The critical importance of addressing disaster preparedness was emphasized, noting the June 2006 IOM reports about emergency care in the United States. Some would say that the components of a typical mass casualty exist on a daily basis somewhere in the United States. A commitment to measuring is vital to progress.

Dr. Boenning requested further input from participants of this “pilot” meeting via email or telephone.


DETAILED SUMMARY

Wednesday, February 14, 2007

Overview of the Data Challenge James Scanlon, OSDP, ASPE

Noting the impact of September 11, 2001, anthrax scares, the response efforts for Hurricanes Katrina and Rita, an ensuing emphasis on preparedness for all threats and hazards, and the recent focus on pandemic preparedness, HHS is reviewing the entire national response and preparedness apparatus in the United States. Relative to the National Response Plan [NRP] (e.g., the play book and model for how emergency incidents should be handled), Hurricanes Katrina and Rita pointed to the need for more systematic assessment and greater clarity.

A fundamental assumption of the NRP is that first response is handled at the community level. Communities requiring additional resources reach out to the state. If state resources are overwhelmed, help would be requested at the regional and interstate level; and then from the federal government (for such areas as health care, public health, medical support, and human services).

HHS has conducted a literature review about what constitutes hospital preparedness. The GAO has undertaken a study to assess performance measurement in preparedness arenas. HHS also has awarded cooperative grants to states for hospital preparedness. Consideration of a certification process was mentioned.

PANEL I FEDERAL PERSPECTIVE

Ongoing Federal Surveys and Available Data

Catharine Burt, Ed.D., Richard Niska, M.D. NCHS

The National Hospital Ambulatory Medical Care Survey (NHAMCS) is a representative annual survey of non-federal general and short-stay hospitals (excluding military, federal, and VA hospitals) that has a sample of 700 hospitals. Two facility survey supplements (i.e., visits to emergency departments [ED] and outpatient departments [OPD]) done in 2003-2004 added questions about staffing capacity, ambulance diversion, bioterrorism, and mass casualty preparedness. Visit data were for more than 76,000 encounters. Survey response rate was 85 percent. The frame used was the Verispan Hospital market base, of which 92 percent had emergency departments and 68 percent were located in the South or Midwest (the latter of which has more ED capability). NCHS’s nationally representative survey has a complex sample design that allows for national estimates, which is a good planning tool for modeling disaster scenarios.

While demand on emergency departments has increased in the last ten years, the number of departments has declined while crowding has increased (note IOM reports of June 2006 and see transcript for statistics and precise definition of crowding). Metro areas in 2003-2004 had approximately 400 hours in diversion status annually or 500,000 ambulances diverted from the closest emergency room, mostly due to a lack of inpatient beds and the number of ED patients. On average, large volume hospitals (relating to bed size) spent as much as 20 percent of their time on diversion with a mean of 311 beds and average occupancy rate of 81 percent. A discussion of treatment spaces ensued: in 2003-2004, 16 percent of hospitals had expanded ED treatment spaces and 32 percent intended to. Inpatient beds per population weighted up to approximately 137 per hospital with an ED. Such projections provide a model that can help to determine needed beds or the impact of different expected scenarios.

The survey supplements mentioned above asked hospitals to describe their mass casualty emergency response plans (see transcript for findings). While approximately 80 percent of the surveyed hospitals had plans, there was a big disparity between the plans and the actual memorandums of understanding (MOUs). Questions were posed about the need to ramp up plans for stockpiling antibiotics and supplies; and about how many resources should divert to urban verses rural areas. See transcript for details about techniques such as converting post-anesthesia care units to intensive care units or activating unused space.

With regard to the survey question about National Disaster Medical Systems (NDMS), a third of the hospitals and somewhat less than half of the rural hospitals did not know their NDMS status (see transcript for other question findings). The disparity between greater urban and fewer rural resources was raised relative to critical care beds, mechanical ventilators, negative pressure isolation rooms, personal protection gear, and decontamination showers. See transcripts and PowerPoint presentation for specific data.

Conclusions

  • Observed crowding in EDs currently limits an ability to handle an influx of cases;
  • Observed disparity across regions in ability to handle a surge due to natural or man-made disasters;
  • Surge capacity planning for specific elements found in ¼ – ¾ of hospitals;
  • Observed disparity favoring urban hospitals in emergency planning and resources.

Discussion

Dr. Niska believes that hospital surge capacity could be improved with the help of survey data. Dr. Burt noted how difficult it is to know the status of ambulance diversion. In 2005, 50 percent of hospitals surveyed did not answer the ambulance diversion question, while in 2002-2003, they had answered the same question (see transcript for discussion about the “burden of reporting”). Dr. Vigilante believes that this burden is part of the mission of today’s workshop in that it affects the quality of data. Dr. Auf der Heide thinks that data should not be collected without a clear understanding of what will be done with it. Dr. Schiller noted that in recent years, better questions have resulted in better answers. In addition, since the survey data of 2003-2004, there has been a large funding increase for the National Hospital Bioterrorism program, resulting in more current data in such areas as decontamination expenditures and personal protective equipment. Ms. Schiller reported that a system is being created that collects data twice/year, analyzes and makes it available for decision-making.

A brief discussion followed about the value of a distance to care measure. Since home address may not be the departure point to the ED, this measure has not been analyzed. Dr. Hanfling recommended that the group distinguish between prospective data collection (e.g., ambulance diversion) and the ability to create real time situational awareness, which is critical to work on surge capacity limitations and a facility’s ability to forecast what may be coming. Dr. Burt noted that some real time questions would be asked on the 2007 NCHS survey. Dr. Niska noted that it would be interesting to tie future data collection to CDC and HRSA guidance on bioterrorism.

Dr. Gamache mentioned that communities in Indiana were asked to not stockpile. More hospitals are now collaborating with other community facilities to meet inventory and other needs. Personal protective equipment (PPE) may also be part of a community plan. Developing survey questions that include a community perspective is complicated because more communities are creating their own strategies to combine and coordinate resources. Dr. Niska emphasized that hospital preparedness data collected in 2003-2004 represents baseline data prior to federal intervention. He suggested that those being surveyed respond one way to funders (e.g. HRSA) and another way to NCHS, which does not provide grants.

The Commission on Accreditation of Healthcare Organizations (JCAHO) has specific requirements for hospital emergency response plans. NCHS surveys used these requirements as well as input from experts to formulate questions but they did not ask to see hospital response plans. Fewer unaccredited hospitals were likely to have done preparedness drilling and training (see transcript for specifics on accreditation). Data collection difficulties stem from the fact that response plans are developed to meet specific needs of particular communities and as such, cannot be compared to others. Dr. Marcozzi mentioned a bill entitled the National Health Security Strategy that passed in December 2006 (with AHA support) that requires a unifying preparedness strategy from states and hospitals that will be submitted to HHS and then to Congress every four years. The bill mandates situational awareness capacity (similar to BioSense at CDC) that mitigates an event and its management, all of which impacts surge capacity.

DHS Office of Health Affairs and Our Role in Interagency Medical Planning

Terry Adirim, M.D., MPH Office of Health Affairs, Department of Homeland Security

A brief history as well as mission and goals of the Office of Health Affairs (OHA) within DHS were presented. The Chief Medical Officer*, who serves as the HHS Secretary’s principal medical advisor, oversees the coordination of DHS’s biodefense and medical preparedness activities. The Office of Health Affairs (the point of contact for federal, state, local, and private sector with regard to medical and public health issues) has three divisions: weapons of mass destruction; biodefense; and occupational health. The OHA-DHS gathers data about early warnings for infectious disease outbreaks, using BioWatch (an environmental monitoring and detection program) and a small component of Bioshield (at ASPR). The National Biosurveillance Integration System (NBIS), to be housed at OHA-DHS, is collecting data in one central location. In general, OHA-DHS is a policy and planning unit that serves as an integrator among agencies and coordinator of medical disaster planning (e.g., pandemic planning with HHS).

How to locate and obtain relevant data are key issues for OHA-DHS because the Office resides outside of HHS while serving an integrative function. See PowerPoint presentation for more information about weapons of mass destruction (WMD) and biodefense; medical readiness; data needs; and OHA’s contributions to the DHS mission.

*Dr. Jeffrey W. Runge was the Chief Medical Officer through March 2007 when he became the Assistant Secretary for Health Affairs.

Case Study: Value of the Electronic Health Record in the VA

Michael Vojtasko, VHA

The twenty-five year old Department of Veteran Affairs (VA) fulfills its legal requirement to support the Department of Defense (DoD) in wartime by surging its beds based on a reordering of care for veterans. When requested by DoD, active duty patients get bed priority; and early discharge and transfers to secondary support hospitals may occur. Operationalized since 1984, the National Disaster Medical System (NDMS) is a public-private partnership employs a response component with the Disaster Medical Assistance Teams (DMATs) and other response teams as well as an evacuation component and the use of NDNS hospitals that have signed up for the system (see transcript for examples of how the system worked with Hurricanes Katrina and Rita). Coordination with DoD is a critical component.

A system used in Iraq by the DoD called the Joint Patient Tracking Application (JPTA) was described (see transcript for specifics). Federal coordinating centers (used to receive and distribute casualties) cover most major metropolitan areas of the United States. Yet, a 2003-2004 survey indicated that a third of the hospitals did not know that they were NDMS facilities. Mr. Vojtasko noted that a new MOU between DHS, HHS, VA, and DoD (developed in October 2005 and updated in late 2006) requires the execution of new memoranda with all NDMS hospitals. Primary Federal Communications Commission (FCC) responsibilities include: recruiting local private sector hospitals; developing local community plans; exercising local plans; developing patient reception process to include local transportation; reporting local NDMS patient bed availability; and receiving patients evacuated to the area and transported to appropriate NDMS hospitals. Bed/casualty reporting goes from the VA and DoD medical treatment facilities to a federal coordinating center, and then to a function at Scott Air Force Base called the Global Patient Movement Requirements Center (GPMRC), which also receives information about patient casualty and movement requirements (for medical/surgical; critical care; burns; psychiatry; and pediatric needs). A bed casualty reporting system called Transcom Regulating and Command & Control Evacuation System (TRAC2ES) is used along with JPTA to track patients through an evaluation system to hospitals able to care for them. The use of these systems translates into patient capability and capacity – the ability to care for patients. There must be a system to move patients to beds and a mechanism at the local level to coordinate arrival and transfer to local hospitals (see PowerPoint presentation and transcripts for further elaboration).

Comments by Discussant Art Kellermann, M.D. Emory University

Dr. Kellermann noted that the best way to ensure a system’s ability to function under extreme circumstances is to ensure efficient and effective functioning on a daily basis (which is not currently the case). Data and metrics should manage the system and facilitate its daily functioning. See transcript for examples that illustrate these points and the consequences to patients of systems that lack capacity. He suggested a pilot program of five or six regional Ambulance Traffic Control Centers (to be located in the most problematic cities for mass casualty events) that would continually monitor occupancy rates, emergency department crowding, the increasing challenge of on-call specialty coverage; and diversion status for those who are critically ill and injured. Such centers would generate shift-to-shift metrics and create situational awareness that would allow for the management of mass casualty events. He cited Maryland and Northern Virginia as good examples of systems that are largely effective in these areas.

Discussion

Dr. Kellermann indicated the need for a uniform platform, set of functions, and data collection from the proposed Ambulance Traffic Control Centers that would involve community partnership and local funding (and a possible partnership with the state). This system would combine communications and decision-making capacity. Dr. Bentley believes that Dr. Kellermann’s idea would improve routine care but expressed doubts about its ability to control patients during a mass casualty surge. He noted that every hospital is overloaded. Dr. Kellermann restated that his proposed system would work for severely ill or injured patients requiring transport, noting that Israel has developed this capacity (i.e., inpatient corridors become inpatient wards, which double as a military hospital system, as needed).

Dr. Auf der Heide supported the need for real time data, noting that current data shows that most disasters in the United States are characterized by a maldistribution of casualties, due in large part to the transport of patients by non-ambulance vehicles. Especially for critically ill or injured people transported by ambulances, it is important to know which hospitals are overloaded or incapacitated in order to avert ambulance traffic to those sites (e.g., Los Angeles example). Dr. Hanfling noted that the regional hospital coordinating center in Northern Virginia began as a medical communications clearinghouse called MedCom. He noted the need for education of the hospital EMS community as well as patients about decision-making relative to patient transfer.

PANEL II STATE AND LOCAL PERSPECTIVE

Regional Trauma Center Preparedness and HAvBED

Dan Hanfling, M.D., INOVA Fairfax Hospital, Virginia

Note: remarks were focused on medical situational awareness in real time or near real time.

Dr. Hanfling illustrated a real time surge capacity hospital plan during Virginia’s 1999 ice storm. He referred to situational awareness, which provides hospitals with the ability to make necessary choices during times of increased demand for patient care; and to match available medical resources to patient needs (see transcript for specifics). He stressed that hospitals that do not adhere to situational awareness will be “closed for business.”

It was noted that INOVA Fairfax Hospital (a Level One Trauma Center in Northern Virginia) did not receive one patient in response to 9-11 due to a lack of situational awareness in real time about the location of available resources. Matching resources are critical for sustained events like biological attack or emerging infectious disease. Since 2002, the Northern Virginia Hospital Alliance has coordinated emergency response disciplines of all regional hospitals under the HRSA division, which is integrated under a regional hospital coordinating center. A format has been created for regional health care coordination and cooperation that includes information management and a communications capability. Borrowing ideas from the Maryland Institute for Emergency Medical Services Systems (MIEMSS) and the Washington DC Hospital Association, Dr. Hanfling suggested that the state of Virginia develop a regional clearinghouse following 9-11, which it did. To date, Northern Virginia is farthest along in integrating a healthcare coordinating function.

An ambulance diversion program in Northern Virginia began as a fax-based and then an internet-based system purchased from a vendor. Because the rest of the state had a different system, Virginia and a representative from Maryland began to talk about how to coordinate the creation of standards. The Medical Surge Capacity and Capability (MSCC) Management System was used as a framework. The MCSS system includes six coordinated tiers (healthcare asset management; healthcare coalition; jurisdiction incident management; state response and coordination; interstate regional coordination; and federal response (see PowerPoint chart and transcript).

The question of how hospital capability information could become available in real time was raised at the national level by Dr. Sally Phillips at AHRQ via the funding of the HAvBED Project. One critical question had to do with bed availability and patient distribution. Based on voluntary reporting, the HAvBED Project provides a sense of available beds as defined by NDMS categories as well as resident identity information within individual facilities. In northern Virginia, a regional MOU (between all 13 hospital CEOs) stipulates that healthcare facilities update preparedness information on a once per shift basis within an open system.

A model for alternative care delivery may need to be developed if demand cannot be estimated and supply calibrated or improved within the current structure. Dr. Hanfling spoke of “echelons of contact” (or “communication”) relative to getting information to communities that optimizes the use of local radio as well as national cable networks, satellite radio, and television networks (see PowerPoint chart). Citizens need data to make informed decisions about access to available care. See transcript and PowerPoint presentation for further elaboration.

Discussion

Dr. Snyder suggested eliciting DoD help in civilian preparedness discussions because their real time data management and capacity to identify patient status is “truly remarkable.” He wondered about the definition of “hospital bed” and “emergency department bed” and suggested that the concepts of “emergency department” and “emergency beds” be expanded. Dr. Hanfling spoke about the need for alternative care facility development (e.g., buildings of opportunity such as hotels or state fair grounds) that could be integrated into emergency/incident management. One way to accomplish this would be to link data to a real time understanding of what kind of patients are in hospitals and what kind remain in the community. Dr. Auf der Heide pointed to the need for a systems perspective. He urged consideration of surge capacity in community-based non-hospital medical facilities (such as physician offices; pharmacies; assisted living facilities, dialysis centers, and urgent care centers). Because an overwhelming number of casualties are not serious injuries, they could be treated in non-hospital settings. He observed that in a number of recent disasters, the majority of casualties were medical rather than trauma cases (e.g., chlorine release in Mississauga, Ontario: see transcripts for specifics). To decrease hospitalizations, it is important to reunite people with their routine sources of medical care and to anticipate the needs of evacuees when doing surge capacity. To that end, it would be useful to gather information about the prevalence of chronic diseases and ongoing diseases within communities in order to determine needed resources for people who lose access to their routine care.

Dr. Hanfling believes that the caregiver is the key metric and the “rate limiting step” in the ability to surge. Dr. Auf der Heide believes that it would be useful to collect post-disaster data to determine how many have gone to hospitals due to inaccessible routine care. This information could influence where resources are used (note Hurricane Aniki [Hawaii] example of inefficient resource use).

Hospital Preparedness: Many Things to Many People

James Bentley, Ph.D., AHA

Less money is available in the healthcare system for data than ever before due to cuts in administrative positions and competing priorities. It is important to ask what decisions the data are designed to address. Defining “surge” as a process to expand capability to care for people, Dr. Bentley believes that there has been too much of a tendency to think of surges and inventory in the past five years. Can demand be matched to resources? A model of demand management is needed. Also for consideration: 1) Can non-recoverable patients be rapidly identified? The public must understand the difference between a military and a civilian definition of triage. 2) Can self-care be maximized? 3) How can distributed ambulatory care be maximized? 4) How can shelter capabilities be developed for people with special needs such as dialysis or a pregnancy beyond 32 weeks? What resources do hospitals really need?

Dr. Bentley agreed that the focus should be on staff as the rate limiting step rather than beds. Utilities (including water, sewage, electricity, gas, oil, and stand-by generators) and a system for their transport are essential to maintaining capacity. The resource supply chain must function and it may be more important to have data on the supply chain than on patients. During disasters, public safety services come from police, fire departments, houses of worship, and hospitals. Because police, fireman and church or synagogue staff is often unavailable, people go to hospitals even when they don’t need to (see transcript for example).

Five observations about data include:

  1. Know why data is being collected
  2. Determine usefulness of data to submitter
  3. Ensure consistency of data collection process with that of the national plan
  4. Data should be a by-product of routine operations
  5. Simpler is better – in most disasters, sophistication is “our enemy” (see transcript for Katrina example)

Discussion

The AHA annual survey, which goes to every hospital in the United States, has limited preparedness data. The AHA sometimes distributes special fax-back surveys on relevant topics such as negative pressure rooms. Hospitals are bombarded with requests for data in different formats (that sometimes ask the same questions), which becomes a big reporting burden. The Hospital Quality Alliance is trying to harmonize these measures so that consistent data requests that are answered once satisfy all interested parties. Asked about the need for a process that harmonizes preparedness data collection, Dr. Bentley noted that the HRSA hospital bioterrorism preparedness program was originally conceived to have that role. He added that when states and localities decide what to change or modify, the vision of the whole is lost. In addition, there is an “enormous tension” between what different levels of government want. Dr. Kellermann cited two major hospital surveys about quality of patient experience that omitted emergency department care, which supports his belief that some of what needs to be measured for preparedness (as well as life and death care) is not taking place.

Dr. Pritchett pointed out that most health departments and local hospital associations have been measuring preparedness for several years. She noted New York State’s Critical Asset Survey and a Washington D.C. survey, which are “very, very detailed” with “all the data elements.” Dr. Pritchett described her hospital experience in Iraq, which included weekly MASCAL (mass casualty exercises) and many of the emergency experiences already described. She clarified her role as director of emergency health and medical services and her responsibility for EMS regulatory oversight. She believes that most large cities in the United States are operating at disaster level every day. The most recent statistics for Washington D.C.’s diversion include “one year of man hours of closure” for District hospitals. Additionally, only 60 percent of inpatients are residents while 40 percent come from Virginia and Maryland (just as these states receive patients from Washington D.C.). An overarching system is needed to control the big picture. While regions such as Washington D.C. and Maryland have preparedness plans, there is no such system for the national capital region and the individual systems do not cross-talk. However, Washington D.C. has begun to use a program called HC Standard, which Virginia already uses and which Maryland is willing to use. This easy-to-use program allows for regional data connection that translates data from various manual systems.

Mr. Miller suggested that intermediate scenarios exist between normal hospital functioning and all-out disaster responses. He cited an example of a citywide power outage affecting communications capability only while a majority of the hospital infrastructure remains intact (see transcript for example of 1996 Atlanta hurricane). He stressed the importance of maximizing the number of people doing self-care. Noting a paper by Richard Danzig, Mr. Miller believes that basic self-care is “as much influenced by what it is, as what it isn’t.”

Public Health Informatics Needs at the State Level

Roland Gamache, Ph.D., MBA Indiana State Health Department

Speaking from a state perspective, Dr. Gamache addressed surge relative to surveillance (i.e., Indiana’s Newborn Screening System and Public Health Emergency Surveillance System [PHESS]) and outbreak management. He described functional Health Information Exchanges (HIEs), all of which are breaking even or making money. These exchanges use clinical and medication data that are available electronically. Forty percent of Indiana’s population participate in an HIE. Indiana’s public health success stories include: all newborn screening is sent to HL7 messaging using LOINC codes; all reportable laboratory data sent through IHIE is electronically sent to the Indiana State Department of Health (ISDH); chief complaint data on presentation to emergency departments is sent to the ISDH electronically by more than 70 of 120 hospitals in near real time (which “probably covers about 80 percent of the population”); and the first phase of Indiana’s National Electronic Disease Surveillance System (NEDSS) was expected to go live in March 2007. ISDH is working to share information with hospitals and clinicians.

Coding emergency department chief complaint data involves categorizing complaints into ten buckets (see PowerPoint presentation for specifics) and analyzing the data. In the past six months, three outbreaks were found before providers called (e.g., carbon monoxide poisoning example). While an infrastructure exists to report outbreaks and discharges, the latter are not well reported although the state is working to improve this situation. State surveillance systems can help to examine surge capacity. Indiana’s ED data are being shared with Ohio and CDC (Biosense Program). Other current research includes: immunization reporting though claims-based data and public health messaging systems. In addition, Marion County Health Department is integrating ambulance data with ED data (see transcript for further elaboration). While developing a surveillance system that provides real time diversion to ambulances is very expensive, it allows for diversion within a timely manner and helps with outbreak management. Staffing available beds continues to be a big challenge for the Indiana and ISDH. The use of admissions and discharge hospital systems to monitor census (which uses information from hospital electronic systems that is already collected) is being explored in a pilot project. Within a year, Indiana will be able to capture pre-hospital electronic clinical information: the data, which is already being captured, is not yet being integrated with hospital information, which is the next step. See PowerPoint presentation for further elaboration.

Discussion

Dr. Gamache noted that Indiana hospitals can currently get outbreak information but not every place has the capability or wants that burden. In a potential outbreak situation, he suggests notifying the emergency department, the local and state health departments; and the state epidemiologist who is situated in the affected district. With good follow-up, most of the initial suspected cases are dismissed.

Dr. Hanfling observed that Dr. Gamache’s presentation was a good example of syndromic surveillance (surveillance using health-related data that precede diagnosis and signals a sufficient probability of a case or an outbreak that warrants further public health response). He again stressed the need to produce information in real time. Dr. Gamache reported that the state is working with the Indiana Health Alert Network (IHAN) as well as the Health Information Exchanges (with several thousand physicians signed on). Alerting systems for most local health departments are connected with others in the pipeline. Phone calls still seem to be the best communication tool within emergency departments.

Dr. Hanfling stated that in Northern Virginia, there is a real time capability to talk by radio with hospitals in the District of Columbia and Maryland. In addition to collecting data and having a data management capability, it is important to have a means of alert notification. Dr. Auf der Heide believes that real time data collection must be based on two-way radios because telephones and cell phones are unreliable in disaster situations. Dr. Gamache noted that the two-way radio is their point of contact and their “most redundant communication spot.” Dr. Boenning mentioned a similar system being used in the Southwest (SYRIS) which he suggested that Andrew Miller from the ARES Corporation could demonstrate after the meeting. Amateur radio services were noted as an additional resource although at present, data cannot be encrypted by amateur radio.

Mr. Handley stated that within Maryland, a stand-alone communications backbone or microwave can override a fiber optic background. As such, a paramedic in one location can get physician orders from an ED in another locale, if needed. The communications hub (SYSCOM), which sits at the Maryland Institute of Emergency Medical Services Systems (MIEMSS) facilitates communication between the paramedic and the hospital. The paramedic calls the dispatcher in Baltimore who will direct the helicopter to the best facility based on the patient’s severity of illness or injury. This process includes not only an ability to talk from field to hospital (with communication passing through MIEMSS) but also provides a running status of the County Hospital Alert Tracking System (CHATS) that has yellow and red alerts for hospital in-patient and acute bed capacity. A red alert appears on a publicly available website that maintains the status of alerts for all Maryland hospitals. As such, more ambulances are appropriately diverted and overcrowding is lessened. Another program described was the Electronic Maryland Ambulance Information System (EMAIS): half of Maryland is on an electronic real time patient care reporting system. When a medic comes to an ED, s/he completes an on-line web-based report at the hospital (stored at MIEMSS) that is available to the jurisdiction and to the hospital.

Mr. Handley and Mr. John Donohue described the Facility Resource Emergency Database (FRED) program, a centralized web-based system that allows for efficient communication that identifies available resources within the healthcare environment [such as beds and equipment] during big disasters (see transcript for example of pre-FRED response to a hazardous waste Baltimore disaster situation). The National Disaster Medical System (NDMS) of Maryland, Delaware, and Pennsylvania filters through the FRED program. Information can be sent to different groups geographically or by discipline although MIEMSS controls information access and security. Mr. Donohue also described the EMRC program: during a disaster, the patient must talk to the hospital, then call EMRC, and then get connected to the hospital. Ordinarily, patients contact the ambulances directly but when the ambulances become stressed, patients contact EMRC for information and help.

Mr. Donohue differentiated between dynamic information (e.g., what can be used now) and static information such as what appears on their web-based emergency management application (EOC). The EOC helps to communicate situational awareness and helps to assign resources. Patient tracking and practitioner availability are other issues to address. There is currently no off-the-shelf patient tracking system that provides adequate information about number and location of available beds, number of beds being used, and the location of patients. A patient tracking project in the capital region is somewhat stalled at present but still moving forward. Working to develop Maryland’s statewide patient tracking system, Mr. Donohue believes that patient tracking must be done on a statewide basis while merging with nearby states. Much work must be done to catalogue important data points. Dr. Auf der Heide clarified two kinds of patient tracking while noting the need for retrospective systems that gather data from institutions receiving patients. Mr. Donohue respectfully disagreed, stating that any patient tracking must have the capacity for use by all disciplines. During a declared disaster, hospitals must begin to immediately identify and track patients or some patients will get lost (noting 1980s Amtrak incidence). Mr. Donohue offered to host a meeting to demonstrate Maryland’s emergency tracking system.

Comments by Discussant Sally Phillips, R.N., Ph.D. AHRQ

Dr. Phillips stated the need for a standard that allows for data sharing and retrieval between various data systems. This standard should not create something new or be labor intensive. The manual hospital data entry system should be entered into the larger system to allow other systems to “talk” to it and to pull data from it. AHRQ’s patient tracking project will be made public in May 2007. It is important to “tweak all those systems” to allow for data exchange and to draw real time or static data into a common pool. These systems must be used to help allocate resources that allow for the best care for the most people. In addition to being real time available, a pool of data must be available for research intended to improve emergency care. It is important to have a minimum data base collection at the time of a disaster in order to understand the level and scope of impact upon the healthcare system. As states manage acute situations, it is necessary to understand how different large areas get the data they need to make decisions about moving patients. One can review that data after the fact to build upon lessons learned in order to improve the disaster evidence base. Recovery effort data must help to build recovery (see transcript for Katrina example). A better data system is needed to retrieve information about what a system looked like before a disastrous event and how it was impacted in order to anticipate after shocks. How to define and measure the “steady state” is a major challenge.

Discussion

Dr. Vigilante summarized the Subcommittee’s mandate to process the meeting’s information and discussions relative to NCVHS’s mission. He reiterated the importance of: finding ways to collect data prospectively in order to understand where vulnerabilities lay; collecting data from a research perspective to bolster the evidence base for best practices; and gathering real time data to guide decision-making. Attempts to reduce that burden on those who report the data must accompany these efforts.

Note: The PowerPoint presentation entitled “The Role of Non-Hospital Medical Practitioners in Disasters” by Erik Auf der Heide, MD, MPH, FACEP (CDC, ATSDR). Dr. Auf der Heide participated in the meeting via telephone.

General Discussion and Plans for the Future

Dr. W. Scanlon thinks that the Subcommittee on Populations needs to identify economically feasible strategies with operational values for the different types of data needs. He asked about what capability a community should have when dealing with disasters. Mr. J. Scanlon spoke about the importance of being able to define in advance and assess systematic readiness or capacity. Dr. Vigilante wondered if the understanding is on the demand or the supply side. Mr. J. Scanlon noted that advances in health information technology that are interoperable and affordable will improve all health care. Dr. Auf der Heide again emphasized the need for retrospective data collection to benefit from lessons learned.

Dr. Kellermann stressed the critical importance of addressing disaster preparedness, noting the June 2006 IOM reports about emergency care in the United States. A mass casualty situation exists in many communities across the United States on a daily basis. A commitment to measuring is vital to progress.

Dr. Boenning requested further input from participants of this “pilot” meeting via email or telephone.

See official transcript for full Subcommittee discussion.