Emergency Neurologic Clinical Trials Network

Emergency Neurologic Clinical Trials Network (ENCTN)

Final Report

March 17-18, 2004
Bethesda, MD

A NATIONAL INSITUTIE OF NEUROLOGICAL DISORDERS AND STROKE SYMPOSIUM

Emergency Neurologic Clinical Trials Network (ENCTN)

Contents

Preamble

Arthur M. Pancioli, M.D.

William G. Barsan, M.D.

Robin A. Conwit, M.D.

Executive Summary

 

Task Force Reports

How Does One Go About Doing EM Research?

Arthur Pancioli, M.D., Claudia Robertson, M.D., Ian Stiell, M.D., MSC, FRCPC,
Gretchen Tietjen, M.D.

Who Are The People And Institutions In The Network?

Edward Jauch, M.D., Dan Lowenstein, M.D., Stephan Mayer, M.D., John Duldner, M.D., M.S.

What Are The Topics For Research And How Will They Be Determined?

William Barsan, M.D., Guy Clifton, M.D., Judd Hollander, M.D., Donald Gilbert M.D., M.S.

Is The Research More Cost Effective If Done By The Networks?

Lewis Morgenstern, M.D., David Matchar, M.D., F.A.C.P., David Wright, M.D.

How Can Data Management Be Handled Most Efficiently?

Michael Fehlings, M.D., Roger Lewis, M.D., Ph.D., Chelsea Kidwell, M.D., Sidney Starkman, M.D.

What Are The Human Subjects Research Issues?

How Will Minority Recruitment Be Handled?

Michelle Biros, M.D., M.S., Peter Panagos, M.D., James Quinn, M.D., M.S.,

Jeff Saver, M.D.

Preamble

Arthur M. Pancioli, M.D. William G. Barsan, M.D.
Steering Committee Co-Chair Steering Committee Co-Chair
University of Cincinnati University of Michigan
Cincinnati, OH Ann Arbor, MI

Robin Conwit, M.D.
Program Director, Clinical Trials
National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, MD

Many neurologic conditions present emergently and patient outcomes can often be determined by the care provided in the first hours after onset. Some of these conditions occur frequently (stroke, traumatic brain injury, seizures) and are associated with considerable morbidity and costs and some occur less frequently (spinal cord injury) but have very high morbidity and high costs associated with their care. Most of the clinical research in the past has focused on chronic and subacute care and there is often limited evidence to support the acute management of many of these conditions. Multicenter research networks have typically been designed to answer disease specific questions and have not been organized to address other neurological conditions that may also be present in the patient population. Each time a new clinical question is addressed, a new clinical research network is established and is usually terminated after the clinical question is answered. Significant start-up costs are associated with the establishment of these networks and are iterative in each new study.

On March 17 and 18, the National Institute of Neurological Disorders and Stroke sponsored a conference to explore the advisability of establishing a multicenter network designed to perform clinical trials in emergent neurological conditions. The Emergency Neurology Clinical Trials Network (ENCTN) concept was discussed by 25 clinicians and scientists from multiple disciplines. The goal was to improve the overall functional outcome for patients with acute neurological emergencies. The participants discussed various aspects necessary in evaluating the potential of such a network, including the organizational structure, funding, cost effectiveness and clinical conditions to be studied. A neurological emergencies network that is not disease specific would open opportunities for clinical research that would facilitate rapid effective treatment of emergent conditions and lead to improved patient outcomes. In addition, the cost savings realized through economies of scale of such a network would allow more research to be performed at a lower cost. The network would have at least 3 clinical trials selected at the time of network activation which could be either phase II or phase III clinical trials. To maintain the network and take advantage of economies of scale, new projects would be continuously developed. The conference participants identified multiple potential clinical trials which could be performed by the network and answer important clinical questions. By facilitating high priority, inter-disciplinary, multi-institutional research into the diagnosis and treatment of neurological emergencies such a network will ultimately lead to new therapies for our patients. We are grateful to the many outstanding and knowledgeable individuals from a variety of professions and medical specialties that contributed to this conference.

Executive Summary

Neurological emergencies constitute a number of serious conditions leading to significant morbidity and mortality yearly in the U.S. Many of these conditions are relatively uncommon, and there is a paucity of evidence from randomized trials to support clinical decisions, particularly in the emergency setting. Neurologists and neurosurgeons are not present during the emergency presentations of many of these disorders, and traditional neurological research has seldom addressed these conditions in the prehospital or emergency department settings.

One significant barrier to performing the necessary studies of many neurological emergencies is the low frequency of presentation to any given institution. Therefore, single institutions or small groups of institutions lack patients for adequately powered studies. A second significant barrier is the tremendous time and resources required to establish networks for multi-center trials. Prior studies of acute neurologic conditions (e.g. the NINDS study of t-PA in stroke) required the development of unique multicenter groups which disband after the study. New studies then require new investigators to construct new networks.

One solution to both of these problems is to create a large, scaleable network of institutions capable of performing research on multiple different acute neurologic emergencies. Thus, multiple relatively uncommon processes which present as acute neurological emergencies could be studied simultaneously, achieving substantial economies of scale, and allowing for adequate sample sizes. The heavy startup costs for the network would only be incurred once and support for each of the individual studies would correspondingly be decreased.

It is the intent of this manuscript to propose the concept of a large, scalable network capable of performing clinical research on acute neurological emergencies. In order to do so, a group of clinical investigators came together to evaluate this concept and generate a potential model of such a network. The following is an outline for the process of evaluation and potential design for such a network.

1. Define the mission

2. Define the scope

3. Design the governance and access

4. Design the structure

5. Describe the type of studies to be performed by this network

6. Evaluate the financial implications of such a network

7. Evaluate the potential data sharing structure of the network

8. Define the necessary human subject protection for all potential patients

 

Mission

The mission of the Emergency Neurology Clinical Trials Network (ENCTN) is to optimize care and improve outcomes for patients who seek care in Emergency Departments for acute neurological disorders.

The ENCTN will achieve its mission by facilitating high priority, inter-disciplinary, multi-institutional research into the diagnosis and treatment of neurological emergencies.

Scope

The ENCTN will encompass a broad range of high morbidity and mortality conditions for which treatment outcomes are poor or little rigorous evidence guides treatment decisions. This network will be built upon a "Hub and Spoke" model. This will be a Scaleable Network of multiple Hubs each responsible for multiple spokes.

ScopeGovernance And Access

The ENCTN should act in a manner similar to a public utility. It must have clearly defined governance and access. It must be designed to foster a broad range of investigations and to insure that the investigations meet the highest scientific standards. Participation in the research by both academic and private practitioners in multiple disciplines will foster acceptance of results to redefine the standard of care. The ENCTN should be open to investigator initiated proposals as well as those from industry.

ScopeEssential Elements In The Governance Of A Clinical Trial Network

Most successful clinical trials networks include a Clinical Coordinating Center and a Data Coordinating Center. An Executive or Steering Committee governs a series of Subcommittees with charges ranging from protocol review to publications.

The ENCTN Steering committee, built on this successful "Board of Directors Model," would have a chairperson, likely the Director of the Clinical Coordinating Center, and other critical members, including a predetermined number of principal investigators from the Hubs who rotate on and off the steering committee at preset intervals. Also included are the Director of the Data Coordinating Center, one or more members from NINDS, and the Chairs of each Subcommittee.

The ENCTN will require a manual of operations to guide the Steering and Subcommittees. The organization will require a geographic home. This could house the Clinical Coordinating Center or the Data Coordinating Center or both, although these may be at separate locations.

Access and Relationship to Funding (NIH and Industry)

Studies to be performed within this network can enter the network via at least two routes.

One route is for investigator-initiated studies, which would be reviewed by an Ad Hoc Review group or Study Section. This review group/study section would include members from the Executive Committee of the network. This would insure peer review from experts in the specific disease state as well as input from administrators with knowledge of the network structure and capabilities.

Studies to be funded by non-governmental entities such as industry can be proposed directly to the Executive Committee. The inclusion of NINDS staff on the Executive Committee insures the infrastructure being funded by NIH/NINDS provides the appropriate scientific return on the investment.

Who Are The People And Institutions In The Network?

Given the breadth of projects that the ENCTN could potentially conduct, a model for the ENCTN network must provide great flexibility, scalability, and institutional breadth. An essential ingredient will be Emergency Medicine and Neurology/Neurosurgery collaboration in the participating institutions. In many institutions this will represent a significant paradigm shift since there has not been a long history of close ties between these specialties. Fostering the development of collaboration between emergency and neurological specialists will be an essential task for the Clinical Coordinating Center.

To enable a scalable network, a hub and spoke model is being proposed. The hubs will be the backbone of the network and all clinical projects will be conducted at the hub centers. For studies requiring larger sample sizes, the network can be "scaled up" by the addition of spokes. Regional hubs will provide research and clinical infrastructure for the spokes, i.e. nearby collaborating hospitals with investigators but perhaps without full time research staff or advanced care capabilities. A hub is not synonymous with the coordinating center. The network will be comprised of dozens of hubs each supporting on average two to five regional spoke institutions. Hubs will likely be regional academic medical centers, the local tertiary care facilities in the region.

Spokes will vary in institutional format and will include academic centers, community hospitals with academic affiliation and community hospitals. Incorporating spokes which are not based at tertiary care centers will increase enrollment and expand patient access to studies while allowing evaluation of the study intervention in a real-world practice setting, providing an estimate of the intervention's clinical effectiveness. Depending upon the complexity of the study, spokes may either enroll and study patients locally, utilizing the hub only for general guidance, or may identify and enroll patients at the spoke prior to transfer to the hub for specialty care. Unlike many hub and spoke models, the ENCTN should also include local practice-based neurologists. These specialists would provide a referral base for potential patients and would serve as co-investigators supporting longitudinal outcome evaluations.

Participants in each spoke will vary by study. The core members at each institution will be emergency physicians with neurological collaboration. Given the existing specialty physician shortages, not all centers will have in-house access to neurologists or neurosurgeons. In these settings, access to the hub's neurologic expertise may be sufficient. In more complex therapeutic studies, lack of neurologist or neurosurgeon specialists will mandate transfer to hub institutions for continuance of the study.

Each hub and spoke will require various support staff commensurate with the studies being conducted at each institution. To some degree a research nurse or research coordinator should be identified and funded to support the ENCTN efforts at each institution. The level of training and commitment will vary by institution.

What Are The Topics For Research And How Will They Be Determined?

The Emergency Neurology Clinical Trials Network (ENCTN) should augment the efficiency and productivity of the Nation's clinical research enterprise. It should provide the capability to conduct, more rapidly and efficiently, multiple high-quality clinical studies and trials in emergency neurology. The full report contains sections addressing critical issues related to 1) the Research Agenda of the ENCTN; 2) the possible mechanisms for targeting and prioritizing specific neurologic conditions for the ENCTN; 3) possible mechanisms for submitting, soliciting, and reviewing proposals for this network, and 4) initial research questions that the ENCTN would be ideally and uniquely positioned to perform.

Research Agenda of the ENCTN

In general terms, the mission of the ENCTN is to optimize care and improve outcomes of patients who seek care in Emergency Departments for acute events affecting the nervous system. The ultimate measure of success for clinical trials is the reduction of mortality and disability in persons who present to the Emergency Department after suffering acute neurologic events.

In specific terms, the research agenda of the ENCTN will involve two categories of diagnoses. The first category comprises high prevalence neurologic diagnoses like ischemic stroke. Despite the high prevalence of these neurologic diagnoses, many diagnostic and therapeutic questions are unanswered. Treatment protocols vary widely among Emergency Departments, and little evidence supports current diagnostic and therapeutic practices. Clinical trials demonstrating even modest treatment effects for these high prevalence diagnoses would result in large absolute benefit in reduction of human suffering and economic losses. Accurately estimating critical, though small, treatment effects requires large absolute numbers of study participants. The ENCTN would have the large number of centers required to perform this research. The involvement of a large number of centers should also greatly enhance dissemination of research findings into the community, accelerating the pace at which new, effective treatments improve outcomes nationally.

The second category is low prevalence but high morbidity and mortality neurologic diagnoses. These diagnoses are not seen at single centers in sufficient numbers to allow adequately powered clinical trials. Yet, from a standpoint of human suffering and societal burden, they merit careful clinical trials. Much clinical research to date on these problems has been piecemeal and inconclusive. For these lower prevalence diagnoses, even multi-center trials can fall short of necessary sample sizes. Thus the ENCTN would be ideally suited to involve the large number of hospitals, both university and community based, required to enroll sufficient numbers of patients.

Examples of high and low prevalence neurologic diagnoses are found in table 1.

High prevalence neurologic diagnoses	Low prevalence, high morbidity/mortality neurologic diagnoses
Stroke Intracerebral hemorrhage Subarachnoid hemorrhage Traumatic Brain Injury Seizure/ Status epilepticus	Global Brain Ischemia Venous Sinus Thrombosis Spinal Cord Injury Meningitis/encephalitis

Criteria for targeting and prioritizing specific neurologic conditions for ENCTN study

Proposals for clinical trials through the ENCTN should be evaluated for significance to human health and scientific merit. With regard to significance to human health, we propose these factors be considered in prioritizing clinical studies and trials:

1. Prevalence of the condition.

2. Mortality.

3. Morbidity, including lost Quality Adjusted Life Years (QALYs).

4. Economic impact.

5. Relevance to the NIH Roadmap goals and appropriateness for ENCTN.

Other Research Considerations

While the main thrust of the ENCTN will be interventional studies, there will be an excellent opportunity to incorporate observational studies into the larger context of these interventional trials. The ability of the Network to collect observational data will be of critical importance in evaluating new areas for interventional studies and providing data regarding numbers of patients and outcomes with conventional therapy prior to designing interventional studies.

The ENCTN will be best utilized if there are multiple simultaneous studies in different disease areas. The simultaneous conduct of multiple studies will enable greater economies of scale and the projects can be phased in such a way that there is no "idling time" present in the network.

 

Is The Research More Cost Effective If Done By The Networks?

 

The cost analysis in the full report is organized into three subsections. In the first subsection, a simplified theoretical framework for identifying and calculating the major components of cost is described. In the second subsection, we consider the factors that may make a network a more efficient strategy for setting up new multicenter clinical trials. In the final section, we use this framework to create an approximate estimate of the relative steady state costs of the ENCTN.

The task force determined that beyond the potential value of developing the ENCTN in promoting the social welfare, we estimate that developing such a network would make economic sense. Specifically, for a fixed portfolio of three large clinical trials, we estimate that for an initial investment of approximately $1.5 million, a network would result in a net savings of nearly $8 million. This savings would result primarily from efficiencies in the operation of trials, since modest decreases in per subject costs is multiplied over a large number of subjects. In order to more precisely predict the cost impact of research performed in the ENCTN, it would be essential to specify the projects in some detail. Despite the highly speculative nature of the current estimates, they do suggest that the ENCTN could be economically feasible.

How Can Data Management Be Handled Most Efficiently?

In order to facilitate multi-center clinical trials in emergency neurological conditions, it is critical that data be collected in an efficient, cost effective, and readily accessible manner that facilitates merging data from a variety of sources and takes into consideration current HIPAA compliance issues. The full report of this task force addresses critical issues related to 1) the type of data entry forms; 2) enrolling and tracking patients; 3) merging data from a variety of sources; 4) centralized data safety monitoring; 5) protocol compliance; 6) the need to generate regular (e.g. monthly) data reports; 7) compliance with current HIPAA regulations. Based on current advances in web-based technologies, electronic data forms and availability and access to the internet, it is apparent that the management of the ENCTN would be greatly facilitated by an efficient web-based data management system.

To maximize the efficiency of the ENCTN and the quality of collected data, a centralized electronic data collection strategy is required. The use of web-based approaches has several advantages over non-web-based systems. Capability to incorporate complex data from a variety of sources will need to be built in. This approach could facilitate the activities of a Data Safety Monitoring Board. HIPAA compliance issues are important to consider when designing the centralized data collection strategy. Several examples of successful application of web-based strategies to run clinical trials have been reported in the literature and attest to the utility and practicality of this approach.

What Are The Human Subjects Research Issues?

While there are many ethical aspects to research, the Emergency Neurology Clinical Trials Network as a whole will most likely be concerned with those related to the rights of human subjects in research, especially those made vulnerable by a devastating disease or injury. These include, but are not limited to, the issues of meaningful consent for research participation, privacy of medical records, and the adequate recruitment of minority patients into clinical trials. There are also unique aspects of clinical trials research and out of hospital research that pose special challenges for emergency neurology researchers.

Issues associated with informed consent and privacy will be common for all centers participating in the network. The goals of the Emergency Neurology Clinical Trials Network will be to inform and educate patients and to protect patient privacy. Site investigators will need to be aware of local and state regulations. The ENCTN will work with local sites to enhance patient protections as well has patient recruitment and enrollment. An expert panel, developed from within and outside of the ENCTN, will advise investigators on ethical, privacy and access issues.

SUMMARY

The NINDS conference brought together a diverse group of clinician/scientists to evaluate the advisability and feasibility of establishing a network to conduct clinical trials in emergent neurologic conditions. The attendees developed a structural framework for the network and identified potential clinical trials to be performed in the network. The network would be built on a hub and spoke model and would incorporate academic as well as community hospitals, leading to better generalizability of the research findings to the community at large. The conduct of multiple simultaneous clinical trials in the network should lead to economies of scale and reduced costs in performing clinical research. Establishment of a clinical research network has the potential to significantly impact the morbidity and mortality from acute neurological disorders and lead to improved outcomes and cost savings in the future.

TASK FORCE REPORT

How Does One Go About Doing Emergency Neurological Research?

Arthur Pancioli, M.D.

Associate Professor and Vice Chairman

University of Cincinnati

Department of Emergency Medicine

Cincinnati

Claudia Robertson, M.D.

Professor

Baylor College of Medicine

Department of Neurology

Houston

Ian Stiell, M.D., MSC, FRCPC

Professor

Ottawa Health Research Institute

Department of Emergency Medicine

Ottowa

Gretchen Tietjen, M.D.

Professor and Chair

Medical College of Ohio

Department of Neurology

Toledo

Neurological emergencies constitute a number of serious conditions leading to significant morbidity and mortality yearly in the U.S. Many of these conditions are relatively uncommon and there is a paucity of evidence from randomized trials to support clinical decisions, particularly in the emergency setting. Neurologists and neurosurgeons are not present during the emergency presentations of many of these disorders, and traditional neurological research has seldom addressed these conditions in the prehospital or emergency department settings.

One significant barrier to performing the necessary studies of many neurological emergencies is the low frequency of presentation to any given institution. Therefore, single institutions or small groups of institutions lack patients for adequately powered studies. A second significant barrier is the tremendous time and resources required to establish networks for multi-center trials. Prior studies of acute neurologic conditions (e.g. the NINDS study of t-PA in stroke) required the development of a unique multicenter groups which disband after the study. New studies then require new investigators to construct new networks.

One solution to both of these problems is to create a large, scaleable network of institutions capable of performing research on multiple different acute neurologic emergencies. Thus, multiple relatively uncommon processes which present as acute neurological emergencies could be studied simultaneously, achieving substantial economies of scale, and allowing for adequate sample sizes. The heavy startup costs for the network would only be incurred once and support for each of the individual studies would correspondingly be decreased.

It is the intent of this manuscript to propose the concept of a large, scaleable network capable of performing clinical research on acute neurological emergencies. In order to do so, a group of clinical investigators have come together (see list of participants) to evaluate this concept and generate a potential model of such a network. The following is an outline for the process of evaluation and potential design for such a network.

1. Define the mission

2. Define the scope

3. Design the governance and access

4. Design the structure

5. Describe the type of studies to be performed by this network

6. Evaluate the financial implications of such a network

7. Evaluate the potential data sharing structure of the network

8. Define the necessary human subject protection for all potential patients

 

Mission

The mission of the Emergency Neurology Clinical Trials Network (ENCTN) is to optimize care and improve outcomes of patients who seek care in Emergency Departments for acute events affecting the nervous system.

The ENCTN will achieve its mission by facilitating high priority, inter-disciplinary, multi-institutional research into the diagnosis and treatment of neurological emergencies

Scope

The ENCTN will encompass a broad range of high morbidity and mortality conditions for which treatment outcomes are poor or little rigorous evidence guides treatment decisions. This network will be built upon a "Hub and Spoke" model. This will be a Scaleable Network of multiple Hubs each responsible for multiple spokes.

Governance And Access

The ENCTN should act in a manner similar to a public utility. It must have both governance and access that are clearly defined. It must be designed to foster a broad range of investigations and to insure that the investigations meet the highest scientific standards. Participation in the research by both academic and private practitioners in multiple disciplines will foster acceptance of results to redefine the standard of care.

In order to facilitate discussion of the potential governing structure of the ENCTN, our group has reviewed and summarized the organizational structure of a series of representative successful clinical trial networks, as well as the organization of one intervention-specific large-scale multicenter trial.

Examples

The Pediatric Emergency Care Applied Research Network (PECARN)

The goal of this network is to conduct high priority multi-institutional research into the prevention and management of acute illnesses and injuries in children and youth of all ages. The PECARN network structure features four regional multi-institutional "Nodes" and a coordinating data center. Each "Node" has 25 affiliated Hospital Emergency Department Affiliates (HEDAs). Each Node works with the funding agencies (MCHB/HRSA) to initiate, implement, and administer network research. The PCARN network is governed by a Steering Committee that formulates and monitors policies and procedures guiding all research activities, and reviews and approves research proposals. All major scientific and operational decisions are made by majority vote. Five subcommittees carry out specific tasks identified by the Steering Committee.

The Asthma Clinical Research Network (ACRN)

This network was established in 1993 by the Division of Lung Diseases (DLD) and the National Heart, Lung and Blood Institute (NHLBI). The objectives of this multi-center program are to conduct multiple well designed clinical trials for rapid evaluation of new and existing therapeutic approaches to asthma and to disseminate laboratory and clinical findings to the health care community.

Administratively, the main governing body of the ACRN is its Steering Committee. The Steering Committee consists of the Principal Investigator from each Clinical Center (there are now 10 Centers), the Principal Investigator from the Data Coordinating Center (DCC), a Chairman who is not an Investigator, and a Project Scientist from the NHLBI.

Access to the network follows a standard protocol. Prior to implementation, an ACRN protocol must be approved by its' Protocol Review Committee (PRC). The PRC consists of clinical scientists, basic scientists, and biostatisticians who are not associated with any other ACRN activity. After a protocol has been approved, patients are recruited, enrolled, and monitored at the Clinical Centers. The Data Coordinating Center provides overall scientific, data management, and administrative coordination during the progress of the trial.

The Parkinson's Study Group (PSG)

The PSG has carried out cooperative therapeutic research since 1986, beginning with the NIH-sponsored DATATOP (Deprenyl and Tocopherol Antioxidative Therapy of Parkinsonism) clinical trial. This group has carried out more than 34 multi-center trials examining the symptomatic and neuroprotective effects of experimental interventions in Parkinson's disease. The PSG has partnered with numerous pharmaceutical companies and The National Institutes of Health (NIH) in bringing five new drugs for Parkinson's disease to the market. The PSG now includes more than 350 active investigators, coordinators and scientists from approximately 85 PSG sites located throughout the United States and Canada.

The PSG is governed by a constitution and bylaws and an elected executive committee that is primarily responsible for the direction and oversight of its' research projects and activities. The PSG has a Clinical Trials Coordination Center (CTCC) that is responsible for the overall coordination of the research activities, including data management for PSG clinical trials. About 70 staff members, including postdoctoral fellows in experimental therapeutics, are involved in the implementation of study protocols, site communications, and data management. The Biostatistics Center of the PSG is comprised of three principal biostatisticians, an 11-member faculty, and three postdoctoral fellows. A group of 12 programmers and research associates are responsible for the analysis of all PSG databases.

Emergency Medicine Cardiac Research and Education Group (EMCREG)

 

The Emergency Medicine Cardiac Research and Education Group (EMCREG™) was founded in 1989 to conduct multi-center clinical research trials on serum markers for the early diagnosis of acute myocardial infarction (AMI). The EMCREG mission has promoted the collaboration between emergency medicine and other specialties in the arenas of both academic research and clinical practice. Structurally, EMCREG is a collective group of emergency physicians with expertise in cardiovascular and neurovascular emergencies that design and implement multicenter trials as well as coordinate seminars on advances in cardiovascular and neurovascular emergencies. Thus it is an inter-disciplinary group which seeks to advance care at the intersection of multiple disciplines including emergency medicine, cardiology, and neurology.

The group has a president as well as executive committee and a series of subcommittees. Access to the research infrastructure comes from member sponsorship and is primarily industry supported. The group currently performs multiple diagnostic studies, runs the largest chest pain registry in existence, and has a strong educational and CME reputation.

American Brain Injury Consortium (ABIC)

The ABIC was established in April 1993 as a non-profit, academic, research based organization with the primary mission of designing, implementing and evaluating the results of clinical trials directed toward improving outcome of brain injury patients.

The concept was to have a group of experienced clinical traumatic brain injury (TBI) centers who are willing and capable of contributing patients to a drug trial continuously available and a central data coordinating center. The individual centers would have a trained nurse coordinator who has experience in collecting the core data set. The centers would also have an interested primary investigator, a large TBI population, and clinicians who are willing to follow the head injury management guidelines. The central data coordinating center would have available accurate information on patient characteristics and patient volume of each individual center, which could be used to design an optimal clinical trial. This organization of continuously available clinical centers was intended to facilitate the conduct of drug trials funded by pharmaceutical companies or by NIH through a partnership. ABIC would provide the expertise for designing and conducting clinical trials in TBI patients, and the pharmaceutical companies would provide the drugs for testing and the financial support.

The group has two co-directors, the ABIC chairman who leads the clinical centers, and the technical director who runs the central data coordinating center located at Medical College of Virginia. The group has an executive committee, consisting of 13 senior investigators from the clinical centers. The group has additional working sub-committees.

Organizational Structure of the NIH Renal Artery Stent Trial (CORAL)

The Renal Artery Stent Trial (CORAL) is a multi-center (65 sites to start, expansion to 85 sites) NIH-sponsored trial (NHLBI) comparing medical therapy to interventional + medical therapy. The trialists include nephrologists, cardiologists, interventional radiologists, and hypertension specialists. It was submitted as a "cluster" application rather than the traditional single R01 (with one PI) with multiple subcontracts. A cluster application is composed of a cluster of interdependent R01s, each with a separate key function, funding and PI. Advantages of the cluster over the R01 include: 1) each R01 has its own direct and indirect costs, 2) Each PI has a 25 page limit, rather than 25 page limit for the entire grant, 3) This arrangement fosters innovation. Disadvantages are that the grant score is that of the weakest link and the coordination is complex and challenging.

The Study has a Chair and Co-Chair who report to the Project Officer and Deputy Project Officer at the NHLBI. The Executive Committee is composed of the Study Chairmen, the PIs of the separate RO1s, the Core lab PIs, and key members (2 interventionalists and 2 clinical trialists). This committee has the responsibility for running the study, the monthly conference-call, and yearly meeting. Committees reporting to them include:

1) The Study Committees: Interventional Committee, Hypertension/Renal Failure Committee, Statistics Committee, Site Selection Committee.

2) The Operations Committee, which is responsible for the day-to-day operations, the

members of whom are also on the Executive Committee, has a conference call on a weekly

basis. The committees which report to it include: Clinical Coordinating Center (coordinates

the investigative sites and deals with the drug and device sponsors) and the Data

Coordinating Center (to whom the investigative sites report).

 

Essential Elements In The Governance Of A Clinical Trial Network

These successful clinical trials networks include a Clinical Coordinating Center and a Data Coordinating Center. An Executive or Steering Committee governs a series of Subcommittees with charges ranging from protocol review to publications.

The ENCTN Steering committee, built on this successful "Board of Directors Model," would have a chairperson, likely the Director of the Clinical Coordinating Center, and other critical members, including a predetermined number of principal investigators from the Hubs who rotate on and off the steering committee at preset intervals. Also included are the Director of the Data Coordinating Center, one or more members from NINDS, and the Chairs of each Subcommittee.

Sub-Committees include:

Protocol Subcommittee

Publications Subcommittee

Data Management and Analysis Subcommittee

Safety and Regulatory Subcommittee

Quality Assurance Subcommittee

Budget, Finance and Funding Subcommittee

Human Subjects/ Ethics Subcommittee

The ENCTN will require a constitution and bylaws to drive the organization and to guide the Steering and Subcommittees. The organization will require a geographic home. This could house the Clinical Coordinating Center or the Data Coordinating Center or both although these may be at separate locations.

Access and Relationship to Funding (NIH and Industry)

Studies to be performed within this network can enter the network via at least two routes.

One route is for investigator-initiated studies, which would be reviewed by an Ad Hoc Review group or Study Section. This review group/study section would include members from the Executive Committee of the network. This would insure peer review from experts in the specific disease state as well as input from the administrators with knowledge of the network structure and capabilities.

Studies to be funded by non-governmental entities such as industry can be proposed directly to the Executive Committee. The inclusion of NINDS staff on the Executive Committee insures the infrastructure being funded by NIH/NINDS provides the appropriate scientific return on the investment.

Note that there is a considerable span of access among current research networks. Some networks run trials exclusively designed by members of the network, others facilitate studies brought to the network from investigators outside of the network or from industry.

TASK FORCE REPORT

Who Are The People And Institutions In The Network?

 

Edward Jauch, M.D.

Assistant Professor

University of Cincinnati

Department of Emergency Medicine

Cincinnati

 

Dan Lowenstein, M.D.

Professor of Neurology

University of California - San Francisco

Department of Neurology

San Francisco

 

Stephan Mayer, M.D.

Associate Professor

Columbia University

Department of Neurology

New York

 

John Duldner, M.D., M.S.

Research Director and Assistant Professor

Akron General Medical Center

Department of Emergency Medicine

Akron

 

Introduction

Given the breadth of projects that the ENCTN could potentially conduct, a model for the ENCTN network must provide great flexibility, scalability, and institutional breadth. Key to these features is to encourage Emergency Medicine and Neurology/Neurosurgery collaboration in the participating institutions. In many institutions this will represent a significant paradigm shift. In order for this to occur, visionary physician leaders in these specialties must lead and move the process forward within their institutions. Additional leadership can come from the regional "hubs" where such collaboration already exists.

The breadth of potential studies will likely require a hub and spoke model for the ENCTN. Regional hubs will provide research and clinical infrastructure for the spokes, i.e. nearby collaborating hospitals with investigators but perhaps without full time research staff or advanced care capabilities. The network will be comprised of dozens of hubs each supporting from two to perhaps twenty regional spoke institutions. Hubs will likely be regional academic medical centers, the local tertiary care facilities in the region.

Spokes will vary in institutional format and will include academic centers, community hospitals with academic affiliation, community hospitals, and to some degree practice-based settings (neurology clinics). Incorporating spokes which are not based at tertiary care centers will not only increase enrollment and expand patient access to studies but will allow evaluation of the study intervention in a real-world practice setting, providing an estimate of the intervention's clinical effectiveness. Depending upon the complexity of the study, spokes may either enroll and study patients locally, utilizing the hub only for general guidance, or may identify and enroll patients at the spoke prior to transfer to the hub for specialty care. Unlike many hub and spoke models, the ENCTN should also include local practice-based neurologists. These specialists would provide a referral base for potential patients and serve as co-investigators for low-risk intervention studies and for longitudinal outcome evaluations.

 

Figure 1. Potential Network Components

 

Participants in each center will vary by study. The core members at each institution will be emergency physicians. At most centers, neurologists and neurosurgeons will participate equally. Given the existing specialty physician shortages, not all centers will have in-house access to neurologists or neurosurgeons. In these settings access to the hub's neurologic expertise may be sufficient. In more complex therapeutic studies, lack of neurologist or neurosurgeon specialists will mandate transfer to hub institutions for continuance of the study.

Each hub and spoke will require various support staff commensurate with the studies being conducted at each institution. To some degree a research nurse or research coordinator should be identified and funded to support the ENCTN efforts at each institution. The level of training and commitment will vary by institution.

Additional resources at hub and spoke institutions will include:

 

Table 2 Network resources

Hub Components	Spoke Components
Physicians Emergency Medicine Neurology Neurosurgery Neurointensivists / Intensivists Neuroradiology Support Clinical nurse coordinator Site (spoke) monitors Central pharmacy IRB / Centralized IRB for spokes Blood / biologic sample repository	Physicians Emergency Medicine Neurology (hospital and practice based) Neurosurgery   Support Clinical nurse coordinator IRB Pharmacy

Examples of Hub / Spoke networks by study type

1) Management of Hypertension in Subarachnoid Hemorrhage

Background

At present, evidence-based management of blood pressure in the hyperacute setting of subarachnoid hemorrhage does not exist. A prospective, randomized trial is needed to guide care provided by emergency professionals. The proposed trial would be randomized treatment of patients with SAH in the Emergency Department with nicardipine to achieve target blood pressure reduction. Treatment would begin within 4 hours of ictus.

Process

Treatment of subarachnoid hemorrhage requires a comprehensive care continuum. The primary centers (hubs) that will provide definitive management are those centers that have the resources to provide a comprehensive approach.

The network is activated based on the type of hospital to which the patient presents as the first medical contact. If a patient presents to a community hospital emergency department (spoke), then transfer to a tertiary center (hub) is mandated. In the event that a patient presents as the first medical contact to a "hub", no additional measures are necessary.

The key to the network (for this and for similar trials) lies in the level or degree of involvement of the community hospital. Two options for the spoke hospitals exist, they are (1) to rapidly evaluate, screen and transfer patients who satisfy inclusion & exclusion criteria, or (2) to evaluate, screen, enroll, initiate therapy and then transfer patients. In either scenario, this would require prearranged cooperation between hub and spoke hospitals. The role a particular spoke hospital will define for itself will be dictated by willingness and resources present at the spoke hospitals. It is conceivable that depending on the nature of the intervention, the spoke hospital may be able to enroll and initiate treatment prior to transfer to the hub. In trials where a time-to-intervention element is critical any efforts by the spoke hospitals would optimize enrollment. Key personnel resources for this type of study will include Emergency Medicine, Neurology / Neurocritical Care, Interventional Neuroradiology, Neurosurgery, and Critical Care Transport Services (aeromedical or ground).

Issues

A potential limitation in the hub and spoke model is the link between these two elements. The primary means of transport between these elements in subarachnoid hemorrhage is rotorcraft. Dedicated services providing inter-hospital transport between the hub and spoke will also need to agree to provide the link under the auspices of the protocol.

Summary

Applying the model of acute management of hypertension in SAH to the hub and spoke model is feasible. Presentation to a hub would be the most direct route of entry into a trial. Because of the temporal requirements in the time to treatment (4 hours), presentation to a spoke would afford either mechanism (screen and transfer, or screen, enroll and transfer). Moreover, the protocol care continuum could be employed in the transport link.

 

2) New Protocols for the Treatment of Status Epilepticus in the Emergency Department

Background

Generalized, convulsive status epilepticus (defined, for practical purposes, as seizures lasting more than 5 minutes) remains a common and important neurological emergency seen in the emergency department setting. New approaches for the treatment of status epilepticus have been advocated in recent years, including greater reliance on longer-acting benzodiazepines such as lorazepam, rapid sequencing of therapy to propofol, midazolam or pentobarbital if seizures do not respond to front-line agents, and the use of new intravenous agents such as valproate. Nonetheless, there have been no large-scale clinical trials looking at the relative risks and benefits of these various, newer strategies. The ENCTN would be an obvious system for carrying out such studies.

Process

Patients will arrive at tertiary care facility. Transfer may not be eligible depending upon allowed medication use prior to enrollment. What exactly does this mean? Early notification of research team, primarily neurology investigators, will facilitate early EEG. This will likely be required for enrollment and monitoring. The study will transition to the ICU possibly involving the intensivist (neuro-intensivist). This study will also be fairly labor intensive.

Resources

The logistical issues would be relatively complex compared to the studies described above. Personnel necessary for such a study would include Emergency Medicine, Neurology, Intensive Care, Clinical Nurse, EEG technician, and pharmacy. First, given that the response to therapy and outcome of status epilepticus are closely linked to etiology, careful diagnostic assessment would be required at all enrollment sites. Second, the study would almost certainly require real-time EEG monitoring, so all emergency departments would need to be equipped with EEG machines, technicians and personnel qualified to interpret the EEG as a guide to treatment (all of which, of course, would need to be available 24 hours a day). It is likely that a neurologist would need to be closely involved with all patient enrollments, both at the time of emergency room management and for subsequent follow-up in-hospital and after discharge. Also, given how difficult it is to follow the clinical course of these patients (in terms of adherence to protocols, response to treatment, hospital course, etc.); an on-site study coordinator would probably be required at each of the participating centers.

3) Emergency Department Evaluation of New Diagnostic Biomarkers for Stroke

Background

Similar to the use of biomarkers for myocardial injury in acute coronary syndromes, biomarkers of brain jury, especially in the setting of acute stroke, are coming closer to fruition. The Emergency Department has vast experience utilizing biomarkers. Additionally, there has been a recent a growth of markers for various conditions, including B-type naturetic protein for congestive heart failure and more accurate d-dimer tests for venous thromboembolism. An important feature of these new tests is the point-of-care (POC) setting. These ELISA based test platforms allows for the test results to be available to the treating physician within 15 minutes. Several companies are in final stages of refining and producing POC stroke marker tests. They will logically be evaluated by Emergency Department staff but Neurology awareness and involvement in the evaluation of these news diagnostics will be critical to building a consensus on if and how these tests will be used in routine practice.

Ideally the setting for evaluating this technology would be in a hospital with a commitment to acute stroke treatment. This will facilitate not only the evaluation of the test performance but also its impact on stroke care (time to CT, time to MD, time to treatment, etc.). Of the several possible studies this hypothetical study will focus on acute stroke (within 4-6 hours of symptom onset), evaluating the markers as a diagnostic aid. The mechanics of the study will likely require serial blood draws to evaluate the acute temporal change in markers in the first 24 hours.

Process

Following the elements of the Stroke Chain of Recovery, prehospital care personnel would notify the receiving hospital of a potential stroke patient arrival. This would facilitate both emergent evaluation of the patients and timely CT but would also mobilize the research infrastructure to evaluate the patient for study enrollment. Due to potential cognitive or language impairment in study subjects, informed consent would frequently be obtained from next of kin. The blood samples would be collected and processed locally on the manufacturer's platform. The study may require additional samples to be shipped to the sponsor or to a core lab for validation. A research nurse or assistant could perform sample collection and case-report-form completion.

Resources

The study would have a local emergency medicine principal investigator and ideally, but not necessary, a neurology co-investigator. The study would likely require research nurse or assistant (10-20% FTE); ideally in center with an existing research infrastructure that could be utilized. No additional personnel would be required and clinical data would be obtained from standard of care evaluation of stroke patients. Data processing and analyses would be performed by the core facilities. No outpatient follow up would be required.

 

Issues

Industry sponsorship an issue related to data management and analysis and would be an issue for the network governance. Given the relative simplicity of such a study it is suitable for a broad spectrum of Emergency Department settings. Given ongoing biomarker research into other diseases there is potential for follow-up trials utilizing same technology (traumatic brain injury, subarachnoid hemorrhage, etc.).

TASK FORCE REPORT

What Are The Topics For Research And How Will They Be Determined?

 

William Barsan, M.D.

Professor and Chair

University of Michigan

Department of Emergency Medicine

Ann Arbor

 

Guy Clifton, M.D.

Professor and Chairman

The University of Texas Medical School at Houston

Department of Neuosurgery

Houston

 

Judd Hollander, M.D.

Professor of Emergency Medicine

University of Pennsylvania

Department of Emergency Medicine

Philadelphia

Donald Gilbert M.D., M.S.

Associate Professor

Children's Hospital of Cincinnati

Division of Neurology

Cincinnati

 

Purpose

The primary goal of this meeting is to define the critical issues related to clinical trials in emergency neurology, and to determine priorities for future clinical and basic research.

Introduction

The Emergency Neurology Clinical Trials Network (ENCTN) should augment the efficiency and productivity of the Nation's clinical research enterprise. It should provide the capability to conduct, more rapidly and efficiently, multiple high-quality clinical studies and trials in emergency neurology. This section will address critical issues related to 1) the Research Agenda of the ENCTN; 2) the possible mechanisms for targeting and prioritizing specific neurologic conditions for the ENCTN; 3) possible mechanisms for submitting, soliciting, and reviewing proposals for this network, and 4) initial research questions that the ENCTN would be ideally and uniquely positioned to perform.

Research Agenda of the ENCTN

In general terms, the mission of the ENCTN is to optimize care and improve outcomes of patients who seek care in Emergency Departments for acute events affecting the nervous system. The ultimate measure of success for clinical trials is the reduction of mortality and disability in persons who present to the Emergency Department after suffering acute neurologic events.

In specific terms, the research agenda of the ENCTN will involve two categories of diagnoses. The first category is high prevalence neurologic diagnoses like ischemic stroke. Despite the high prevalence of these neurologic diagnoses, many diagnostic and therapeutic questions are unanswered. Treatment protocols vary widely among Emergency Departments, and little evidence supports many current diagnostic and therapeutic practices. These high prevalence neurologic diagnoses are associated with irreversible damage to the central nervous system, and thus, the extent to which practice variations result in worse outcomes has devastating results for patients, families, and the healthcare system. Clinical trials demonstrating even modest treatment effects for these high prevalence diagnoses would result in large absolute benefit in reduction of human suffering and economic losses. Accurately estimating critical, though small, treatment effects requires large absolute numbers of study participants. The ENCTN would have the large number of centers required to perform this research. The involvement of a large number of centers should also greatly enhance dissemination of research findings into the community, accelerating the pace at which new, effective treatments improve outcomes nationally.

The second category is low prevalence but high morbidity and mortality neurologic diagnoses. These diagnoses are not seen at single centers in high enough volume to enable efficient clinical trials. Yet, from a standpoint of human suffering and societal burden, they merit careful clinical trials. Much clinical research to date on these problems has been piecemeal and inconclusive. For these lower prevalence diagnoses, even multi-center trials can fall short of necessary sample sizes. Thus the ENCTN would be ideally suited to involve the large number of hospitals, both university and community based, required to produce sufficient numbers of patients, whereas more traditional, investigator initiated studies may lack sufficient subjects.

Examples of high and low prevalence neurologic diagnoses are found in table 1.

High prevalence neurologic diagnoses	Low prevalence, high morbidity/mortality neurologic diagnoses
Stroke Intracerebral hemorrhage Subarachnoid hemorrhage Traumatic Brain Injury Seizure/ Status epilepticus	Global Brain Ischemia Venous Sinus Thrombosis Spinal Cord Injury Meningitis/encephalitis

Criteria for targeting and prioritizing specific neurologic conditions for ENCTN study

Proposals for clinical trials through the ENCTN should be evaluated for significance to human health and scientific merit. With regard to significance to human health, we propose these factors be considered in prioritizing clinical studies and trials:

1. The prevalence of the condition, which can be estimated superficially, although not always accurately, through identifying ED ICD-9 billing codes.

2. Mortality.

3. Morbidity, including lost Quality Adjusted Life Years (QALYs) when this can be calculated. These factors will in some cases be critical for prioritizing lower prevalence neurologic conditions.

4. The economic impact of these diagnoses, when known.

5. Relevance to the NIH Roadmap goals and appropriateness for ENCTN, as opposed to traditional clinical trial research mechanisms.

Other Research Considerations

While the main thrust of the ENCTN will be interventional studies, there will be an excellent opportunity to incorporate observational studies into the larger context of these interventional trials. The ability of the Network to collect observational data will be of critical importance in evaluating new areas for interventional studies and providing data regarding numbers of patients and outcomes with conventional therapy prior to designing interventional studies. The Network will also have the ability to collect surveillance data on important acute neurological conditions.

The ENCTN will be best utilized if there are multiple simultaneous studies in different disease areas. The presence of more than one study will lead to greater economies of scale and the projects can be phased in such a way that there is no "idling time" present in the network.

Possible mechanisms for reviewing and selecting proposals to be performed by ENCTN

Sources of proposals

There are several possible sources of proposals. First, we anticipate that many proposals will come from individual investigators within the ENCTN. The ENCTN will include sites where clinician scientists in the relevant fields, emergency medicine, neurology, and neurosurgery, already have significant experience with phase II and phase III trials in emergency neurology. New and well- established investigators or teams of investigators may submit traditional grants that have an R01 format.

Second, the NIH may wish to consider collaborating with an elected, executive body within the ENCTN in generating traditional RFPs for specific, high impact proposals for Emergency Neurology Research that would be particularly well suited to the structure of the network. This may enhance the development of a specific, successful, and clinically important research agenda during the initial 2-3 years of the Network.

Finally, a governing committee within the ENCTN may wish to commission a group of investigators within the ENCTN to work collaboratively to generate important, high impact clinical trials consistent with the ENCTN research mission. This may be an important mechanism for developing more novel, high-risk proposals or proposals translating basic discoveries to early phase clinical testing.

Mechanisms for grant development and review

The efficiency and productivity of the ENCTN may be enhanced by collaborative effort within the Network. Committees within the network may bring together multi-disciplinary clinical knowledge, clinical trials design expertise, and statistical analysis to assist investigators early in the process of grant development. Critical approaches to overcoming unique obstacles to Emergency Department research, including enrollment, consent, and rapid implementation of research protocols, could be developed collaboratively.

There are several well-established models of collaborative research networks, which should be considered for this process. The first model would be based on existing consortia for disease specific research, for example the Parkinson Study Group (PSG) (http://www.parkinson-study-group.org/). The PSG, established in 1986, is governed by an elected executive committee, which is responsible for the direction and oversight of its research. The infrastructure includes a Clinical Trials Coordination Center and a Biostatistics Center, each providing expertise in preparation of proposals, prior to submission, and execution of NIH or industry funded proposals. Each Center is involved in Clinical Research Workforce Training.

Following the PSG model, the ENCTN could set up multiple disease-based committees for common diagnoses. For example, for ischemic stroke, clinician researchers from neurology, emergency medicine, neuroradiology, and rehabilitation medicine could comprise a disease-based research committee to propose and/or review clinical trials. Separate committees and infrastructure could provide support in research design, outcomes measurement, and statistical analysis.

A second model would be based on existing networks like the Neonatal Intensive Care Unit Network, established by the National Institute for Child Health and Development, or the Children's Oncology Group (COG), supported by the National Cancer Institute. The COG structure includes a foundation, the National Childhood Cancer Foundation, that receives and manages COG grants from the NCI and provides grants and contracts administrative personnel and technical services.

COG develops and coordinates cancer clinical trials at over 200 member universities and teaching hospitals (http://www.childrensoncologygroup.org/). Studies are managed through a web-based data entry system. COG has approximately 100 ongoing studies, enrolls 5,000 subjects per year, and follows 35,000 childhood cancer survivors. COG has many disease specific committees. Fields useful to all clinical trials, for example, epidemiology, biostatics, pathology, translational research, and allied disciplines, each have their own committees which then can support and collaborate with the disease specific committees. COG committees review all clinical trials for scientific merit and patient safety prior to review by the National Cancer Institute.

This model is attractive in that it allows for much of the disease specific work on clinical trials to be performed within committees comprised of national expert clinician researchers, while experts within disciplines like biostatistics which are not specific to any one condition are available for use by multiple disease committees. This model has allowed for careful assessment of incremental changes in doses and combinations of therapy, resulting in continuous reductions in morbidity and mortality of childhood cancers over the past decade. If employed in a similar fashion in the ENCTN, therapeutic protocols that show benefit, like TPA for ischemic stroke, may be modified in a similar fashion.

The large oncology trials through NCI have yet a different structure. Within the largest consortium, the Southwest Oncology Group, NCI provides a certain amount of funding over a 7-year period and the trials to be conducted are selected by committees within the group. There is a Clinical Coordinating Center for the group that coordinates the selection and conduct of the clinical trials. The vast majority of the money is spent on per patient costs for the trials to be conducted within the overall budget. Peer review occurs at the end of each 7 year period when the success of the group is assessed and decisions to continue funding are made at the NCI. The NCI separately funds the Data Coordinating Center for the group.

A suggested structure for the ENCTN might use elements of these other networks to form a unique operation. NINDS could establish a special study section specifically for peer review of trials for the ENCTN separate from other clinical trials study sections. This study section would have representation from both the Clinical Coordinating Center and the Data Coordinating Center for the ENCTN. Within the study section would be experts from different neurological areas with the ability to recruit ad hoc reviewers when needed. Non-industry studies to be conducted by ENCTN would only be those selected by this study section. The Clinical Coordinating Center could also entertain proposals from industry by establishing committees within the ENCTN for the purpose of evaluating these proposals. The vast majority of the funding for clinical trial sites would come from per patient/revenues from conducting the clinical trials.

The ENCTN needs to involve Emergency Departments outside of major academic centers, and we anticipate that elected representatives to the disease-based committees would represent all types of member hospitals. Multiple disease-based committees with rotating memberships should also assist in achieving the ENCTN goals of rapidly conducting multiple high quality clinical trials concurrently and efficiently and disseminating research results widely.

 

Priorities for future clinical and basic research

Although specific clinical trials that can utilize the capability of this network may be proposed by individual investigators, we anticipate, particularly in the first 3 years of the ENCTN, that committees within the ENCTN should play a critical role in determining priorities for clinical trials and proposing specific studies. Examples may include studies where pilot work has already been done but where a definitive trial could best be performed using the ENCTN. Such studies should be ideal candidates for the large pool of hospitals in the network. Early establishment of disease-based committees and working group meetings or symposia to develop research aims for high prevalence, high impact diseases like traumatic brain injury may be an important initial step.

The group discussed a number of potential interventional studies as well as observational studies that could be conducted concomitantly. The following is a list of the Interventional and observational ideas discussed by the group:

Interventional Studies:

1. Traumatic Brain Injury

a. Early administration (within one hour of injury) of an NMDA antagonist

b. Early application of hypothermia (within one hour of injury)

c. Early administration of hypertonic saline resuscitation

d. A randomized trial of prehospital vs. ED airway management

2. Cerebrovascular Disease

Role of early surgery in management of ICH > 3 cm2

a. Low dose t-PA combined with a IIB/IIIA antagonist in patients with acute

ischemic stroke (AIS) within 3 hours

b. Intensive glucose control in patients with AIS

c. Goal directed therapy for AIS, ICH and/or SAH vs. conventional therapy

d. Randomized trial of heparin vs. placebo for dural venous sinus thrombosis

3. Epilepsy/Seizure

a. Neuroprotective agents for patients with Status Epilepticus (SE)

b. High vs. Low dose Lorazepam in SE

c. Rapid sequence treatment of SE

d. Treatment of alcohol related seizures

4. Neuro-Infectious Disease

a. Treatment of Bell's Palsy with steroids/antiviral agents

5. Spinal Cord Injury

a. Early vs. delayed decompression of the injured spinal cord

b. Induced hypertension in SCI

c. Hypothermia in SCI

d. Neuroprotective agents in SCI (Riluzole, Erythropoietin, Mincocycline)

6. Global Brain Ischemia

a. Implementation of pathways for induction of hypothermia

b. Catheter vs. surface cooling for hypothermia

The following are suggested topics for Observational studies which could be conducted by the network to answer important clinical questions and gain insights into the need for future clinical trials:

TBI

a. Predictors of poor outcome in moderate TBI (minimal CT abnormalities and

GCS 10-13)

b. Predictors of poor outcome in mild TBI (minimal CT abn. And GCS 13-15)

c. Utility of CT in unresponsive patients

Cerebrovascular Disease

a. Long term deficits in patients with mild stroke (NIHSS of 4 or less and/or rapid

improvement

b. Predictors of hemorrhage after t-PA

Epilepsy

a. Need and timing of diagnostic workup in new onset seizures

b. Practice variations in alcohol related seizures

 

Headache

Predictors of meningitis/SAH/dural sinus thrombosis in acute headache

 

Spinal Cord Injury

a. Role of early vs. late decompression

b. Relationship of blood pressure to neurologic outcome

Recommended Studies for the ENCTN

After discussion, the following studies were felt to be good candidates for initial study in the ENCTN:

I. Rapid sequential treatment of generalized convulsive status epilepticus

Design: Randomized, masked

Intervention: Comparison of rapid sequential antiepileptic (AED) drug treatment regimen vs. standard sequence AED regimen in generalized convulsive status epilepticus (SE).

Rationale: Barbiturates are less effective after failure of benzodiazepines and phenytoin, and GABA receptor composition is modified over the 1^st hour of generalized convulsive SE and less responsive to GABA agonists.

Primary outcome: Neurological status at one month

Secondary outcome: Time to cessation of status, time to discharge, ICU days

II. NMDA receptor antagonist and emergency airway management within one hour of TBI (EMS study)

Design: 2 X 2 factorial design

Intervention A: Randomization to prehospital intubation of no intubation

Intervention B: Randomization to receive NMDA antagonist vs. placebo

Rationale A: Prehospital airway management may delay delivery of patient to the hospital with no outcome benefit

Rationale B: Previous studies of neuroprotective agents have been given too late to be effective

Primary outcome: Neurological outcome at 90 days

Secondary outcome: Aspiration pneumonia, hypoxia, ICU days, hospital days, cost

III. Combination approach to thrombolysis utilizing eptifibatide and rt-PA (CLEAR trial)

Design: Randomized, double-blind, dose escalation

Intervention: Combined eptifibatide/rt-PA vs. standard dose rt-PA

Rationale: Despite proven benefit of rt-PA in AIS, 3 month outcomes demonstrate 50% of patients with physical disability. New combination approaches to improve speed and success of early arterial recanalization are necessary.

Primary outcome: Neurological outcome at 90 days

Secondary outcome: Symptomatic hemorrhage

 

Conclusion

Development of an ENCTN would have broad use in the study of both high prevalence and low prevalence, high morbidity neurological disorders. The ability to further determine small benefits from therapeutic interventions in the high prevalence disorders would have significant impact on the treatment of patients with neurological emergencies. Likewise the ability to identify and enroll patients with low prevalence high morbidity disorders in an ENCTN can lead to diagnostic and therapeutic advances that would otherwise not be possible.

 

TASK FORCE REPORT

Is The Research More Cost Effective If Done By The Networks?

Lewis Morgenstern, M.D.

Director of the Stroke Program

University of Michigan

Department of Neurology and Emergency Medicine

Ann Arbor

David Matchar, M.D., F.A.C.P.

Director and Professor of Medicine

Duke University

Durham

David Wright, M.D.

Assistant Professor and Assistant Director

Emergency Medicine Research Center

Emory University

Atlanta

Introduction

The decision to set up a network for performing multicenter clinical trials involves balancing the social benefits versus the resource utilization. Other sections of this document cover the social benefits (e.g., creating a cadre of clinicians well familiar with the state-of-the-science). The primary goal here is to describe the cost of research in a network, focusing on the major issues that are likely to influence whether setting up and sustaining a network is a more or less costly compared to the alternative - performing a portfolio of clinical studies without the support of a network.

This section is organized into three subsections. In the first subsection, we provide a simplified theoretical framework for identifying and calculating the major components of cost. In the second subsection, we consider the factors that may make a network a more efficient strategy for setting up new multicenter clinical trials. In the final section, we use this framework to create a "back of the envelope" estimate of the relative steady state costs of an emergency neurology research network.

II. A Framework For Identifying And Calculating Relevant Resource Costs

What Costs Are Particularly Relevant To Decision Making?

All accounting problems begin with a statement of the decision that is being informed by the analysis. In this case, there are at least two possible decisions to consider: (a) what is the likely cost and research productivity of a network versus the current non-network approach to setting up and running emergency neurology trials? and (b) given a fixed portfolio of projects, what resources are required by a network compared to a non-network research strategy? The first question involves speculating about the number and variety of projects that may be performed under network vs. non-network conditions. Not only do costs depend in a non-trivial way on the nature of the research portfolio, an estimate of research productivity depends on predictions of how the very existence of a network might impact the decision to develop and fund potential research projects. Thus, we focus on the latter question: given a fixed portfolio of projects, what resources are required by a network compared to a non-network research strategy?

In considering this question, we sought to estimate two quantities: (a) the cost of starting up a network - the costs that occur in the first year which do not recur in subsequent years - and (b) net steady state costs - the extra cost (or the savings) of performing research in the network vs. performing the same research portfolio without benefit of a network. The net steady-state cost relates to costs of initiating new projects as well as the costs of continuing ongoing projects.

The taxonomy of costs considered in this analysis is listed in Table 1.

Table 1. Taxonomy of costs

Type of cost	Notation
A. Cost of studies without a network
1. Project start up costs	C(NP[DLG1]|No net)
2. Ongoing study costs	C(OP|No net)
B. Cost of studies in a network
1. Initial cost of setting up a network (non-recurrent)	C(Net startup)
2. Network infrastructure costs (recurrent)	C(Net recurrent)
3. Project start up costs	C(NP|Net)
4. Ongoing study costs	C(OP|Net)

 

Calculating Network Startup Costs And Relative Steady State Cost For Network Vs. Non-Network Based Research

Costs of starting up a network are straightforward and involve a simple summation of the initial, non-recurrent costs associated with creating a research network. Absolute steady state cost, whether in a network or not, are the sum of (a) the cost of setting up any new projects (C(NP|Net) and C(NP|No net), and (b) the cost of running an ongoing portfolio of projects (C(OP|Net) and C(OP|No net) for network or no network, respectively) and for network or no network, respectively). For a network-based strategy, there is an additional cost of running the network itself (C(Net recurrent). Thus, the net cost of a network is:

{C(Net recurrent) + C(NP|Net) + C(OP|Net)} - {C(NP|No net) + C(OP|No net)}.

Rearranging,

C(Net recurrent) + {C(NP|Net)} - C(NP|No net)} + {C(OP|Net) - C(OP|No net)}.

Therefore, the net steady state cost of a network vs. non-network based research strategy is the sum of three terms: (a) the steady state cost of running the network for the duration of the trials, (b) the net cost of starting new projects, and (c) the net cost of ongoing projects.

Other Potentially Relevant Calculations

Assuming that the above calculation leads to a net annual cost saving, one can calculate a "time to break even" for the expenses initially incurred by setting up the network (C(Net startup)). That is:

C(Net startup)/net annual cost saving,

where net annual cost saving can be approximated by taking the negative of net total project cost divided by the estimated average duration of trials.

Iii. Key Relative Cost Drivers: What Makes Setting Up And Running Projects More Or Less Efficient In A Network?

To estimate the net steady state cost of network based research, it is important to consider which resources are required, focusing on those resource using activities for which a network may be potentially valuable (in terms of the previous section, activities that are more efficiently performed in the context of a network: Such activities are potential drivers of the net savings of network-based research. A list of potential cost drivers is provided in Tables 2 and 3.

 

Table 1. Potential Drivers Of Cost For Project Start Up In A Research Network

Cost driver	Cost driver modifiers that might be affected by presence/absence of network	Comment
Site identification	Complexity of protocol	A network could increase rate of site identification; this benefit could be greater for more complex protocols
Hiring site personnel	Availability of personnel (either working on current trials or available from the local labor pool)	Ongoing trials via the network would increase the likelihood a site would have appropriate staff and would not have to hire new staff
Training site personnel	Complexity of protocol	Ongoing trials via the network would increase the likelihood a site would have staff experienced in basic trial procedures (e.g., GCP); this could decrease training time
IRB approval	Complexity of consent (e.g., surrogate)	A network IRB could assure that protocols are in ideal shape for local IRBs; for sites participating in regional IRBs the network could maintain an ongoing relationship with the regional IRB and thus hasten approval

Table 3. Potential Drivers Of Cost For Ongoing Projects In A Research Network.

Cost driver	Cost driver modifiers that might be affected by presence/absence of network	Comment
Per patient negotiated cost	Economy of scale	Per patient enrollment costs and negotiated fees with spoke members could be much less than single contractual agreements based on the promise of more studies and the efficiency of having ongoing infrastructure
Enrollment rate	Number of centers enrolling subjects	A large network could increase the number of subject recruitment sites and dramatically reduce the time to study completion, shorter studies would reduce costs and bring discoveries to the clinic sooner.
Site attrition	Vetted and stable group of sites	Study sites vetted by a research network are more likely to be stable and well prepared for study protocols. Attrition from single multicenter clinical trials can be high and add significantly to the overall costs of running the project.
Personnel attrition	Stable personnel environment and pool	Having a network would create a stable environment and reduce attrition of key, highly trained, and site specific, neurologically experienced core study personnel. This would reduce the cost of hiring and rehiring personnel for new studies.
Site performance	Experienced sites with central oversight	Experienced network-based sites would have better performance, enhanced recruitment, and decrease cost of data maintenance/cleaning
Study protocol adherence	Protocol experience	Experienced personnel and previously vetted capabilities will lead to improved protocol adherence. Continuous network training. Standardized Manual of Operations, Standard Operating Procedures, and Methodologies will all lead to enhance protocol compliance.
Consistency of care	Patient management protocols	Building a network would include determining the capabilities and characteristics of each study site. In order to participate, sites would have to agree to follow protocols for patient management and have commitment from the specialists involved. Standardizing the management of patients could reduce the variability in outcome and reduce the number of subjects needed to show a difference.
Data quality	Standardized data management plan, standard data dictionary, personnel experience	Standardized data management plan and data handling process created by the Network Core could greatly increase the data integrity and quality.In addition, a well thought out comprehensive data dictionary could enhance study design, increase comparability between studies, and added increased value to the overall data set.
Database	Number of distinct databases and number of data elements in each database	A standardized neurological emergencies database maintained at a central location could reduce the cost of building multiple smaller databases for individual studies. Centralized maintenance could reduce the cost database support.

 

Iv. Estimate Of Cost Of Network Vs. Non-Network Based Research

Assumptions

Applying the constructs outlined, below we estimate the specific cost elements, net costs, and time to break even for a neurological emergencies network performing a fixed portfolio of research, relative to identical research activities performed without a network. In order to make this comparison, we made several simplifying assumptions:

First, we assumed that the research portfolio consisted of 3 multicenter clinical trials. Though not defined in detail, we considered each trial would require 1,800 patients. This number was derived from previous literature regarding the number needed to show clear clinically significant improvement in outcome (obviously depending on effect size and variance) as well as the reasonable number of recruitable subjects.

Second, we considered a "hub and spoke" network model. This model consists of a series of central core centers that coordinate regional recruitment activities. The spokes are described as smaller clinical research units capable of rapid upstarts due to preset contracts, agreements, study personnel identification, continuous training activities, and central coordination and oversight by the local hubs. The hubs, mostly geographically appropriate high volume academic centers, would be continuously operational and would have permanently funded staff (including a nurse coordinator) and would provide some support for a faculty site investigator even during idle state. Hub activities during idle state would continue spoke recruitment and training, maintaining cooperative agreements, IRB approvals, and coordinate future studies with other hub centers and a central advisory body. For the purpose of calculating costs, the network was assumed to contain 15 hubs, each with 10 spokes.

Third, to avoid double counting of costs associated with a network vs. costs associated with specific projects in the network, we assigned all costs performed by the network personnel and using network resources to be "network" costs. In practice, many of these costs would be accounted to specific projects.

Finally, for simplicity, we include the cost impact of completing studies more quickly in a network by calculating the total costs of project without considering issues of discounting to present value.

Estimates Of The Relevant Component Costs

Described below are the estimates of the relevant components of cost related to projects performed in and outside of a research network. These estimates are summarized in Table 3.

Cost category	Estimated cost for 3 studies
Cost of studies without a network (over 5 years)
Project start up costs	$0.3 million
Ongoing study costs
Fixed costs	$18 million ($1.2 million per year per project x 3 projects x 5 years)
Variable costs	$43.2 million ($8,000 per subject x 1,800 subjects x 3 projects)
Cost of studies in a network (over 3 years)
Initial costs of setting up a network (non-recurrent)	$1.5 million
Network infrastructure costs (recurrent)	$17.5 million ($3.5 million per year x 5 years)
Project start up costs	$0.15 million (0.5 x $0.30 million)
Ongoing study costs
Fixed costs	$8.1 million (0.75 x $1.2 million per year per project x 3 projects x 3 years)
Variable costs	$32.4 million (0.75 x $8,000 per subject x 1,800 subjects x 3 projects)

Cost Of Studies Without A Network

Project start up costs (C(NP|No net)

These costs include the cost of site identification, training related to each protocol, and so on. In the absence of a network, from previous experience, we estimated these start up costs to be $100,000 per study. In addition, due to the complexity of this task, we estimated site recruitment would take 1 year.

Ongoing study costs (C(OP|No net)

Ongoing study costs include a fixed cost related to non-patient activities required for running a study, and a variable cost associated with patient activities such as enrollment, study intervention, and follow up. The variable component is often implemented as a capitation rate.

From experience with large clinical trials, we estimated the typical fixed component for a multisite study involving 1,800 patient studies to be $1.2 million per year per study. For the variable cost, we estimated the per-patient cost to be $8,000. This figure is typical of the capitated reimbursement for an NIH study and is well below the reimbursement for a pharmaceutical company sponsored clinical trials (in the range of $12,000-$16,000 per patient).

Cost Of Studies In A Network

Initial cost of setting up a network (C(Net startup)

The one time cost to set up a large network with 15 hubs was estimated at $100,000 per hub. The expenses incurred would include hub identification and recruitment of regionally appropriate, high volume, experienced research institutions by a central advisory body and network steering committee. The institutions identified would require experienced investigator(s) with a track record of collaborative neurological research (including emergency, neurological, and involvement of other services) and a host of other necessary skills. Although the process of site identification was not specifically defined, this process would include initial meetings between the funding agencies and the steering committee. Once the hubs and site investigators were identified, research personnel such as qualified study coordinators would be recruited and trained. This whole process would likely require regional meetings and site visits by key personnel. Cooperative agreement between sites and local spokes would also be an important part of this process. The total estimated cost for setting up 15 hubs at $100,000 each would be $1.5 million.

Recurrent Network Costs (C(Net Recurrent))

Once established, recurrent network costs relate to maintaining the functions of the central command center and the hub structure. The central core would consist of a steering committee, study review panel, data management committee, statistical center and other pertinent committees. The command center would maintain the central database, data dictionary, case report forms, data management plans, manual of operations, standard operation procedures, and other central processes that would be accessible to all neurological emergency research studies. The estimated annual cost for network maintenance is $500,000.

In addition, we include in the estimate of the costs attributable to network maintenance to be the cost of maintaining a network in "idle" - with no active projects but with the instantaneous capacity to initiate a new project without having to reinvent core capabilities (i.e., the costs for these capabilities considered here to be "non-recurrent"). This would primarily require personnel at each hub; the minimum for each hub costs would include 1 FTE for a nurse coordinator, .1-.15 FTE for a site investigator, and 1 FTE for a research nurse or assistant. This translates to an annual cost per site of approximately $200,000, or $3 million per year for all sites. Together with the core maintenance functions, this is estimated to be approximately $3.5 million per year for all sites.

 

Project start up costs in a network (C(NP|Net))

For an established network, compared to project startup in a non-network setting, the complexity and thus costs would be greatly reduced. Given that the sites are pre-identified and operationally standardized, the largest cost would be in training the sites for study specific tasks. This is estimated to cost about $50,000 per study and take only 3 months in a network based system. The overall financial savings for the project startup costs is $50,000 (i.e., relative cost = 0.5).

 

Ongoing study costs (C(OP|Net)

Fixed costs. One advantage to a network is that the personnel for basic operations will not have to be duplicated for each individual multicenter clinical trial. For example, since the hubs already have study coordinator, site investigator, and research assistant support from the network, only a limited number of additional personnel would need to be added as more studies are brought on line. Centralized training and consistent experienced personnel will also add to the efficiency, quality, and overall success of the clinical trials performed in the network (as a benefit above simply centralizing the training process). The other projected network savings is from accelerated subject enrollment that would be expected from a large experienced network. We estimated that the relative fixed costs would drop with each additional project as follows: 1.0 for the first, 0.75 for the second, and 0.5 for the third, for a weighted relative fixed ongoing cost of 0.75 for three projects.

Variable costs. The most critical cost benefit for a network will be the ability of the network to negotiate lower per patient costs. Provided that both the hubs and the spoke participants can be "guaranteed" multiple studies, the per-patient costs could be lowered due to the economy of scale. We estimated that a site with three ongoing studies would be willing to reduce their capitated reimbursement by about 25% compared to the rate for a single study. That is, the relative variable ongoing project cost would be 0.75.

Cost Summary

Based on the above inputs, we estimate that the cost of starting up a network would be $1.5 million. Further, once such a network was setup, the relative cost of performing three large multicenter trials in the network relative to a non-network setting is:

C(Net recurrent) +

{C(NP|Net)} - C(NP|No net)} +

{C(OP|Net) - C(OP|No net)}.

or

$17.5 million +

{$0.15 million - $0.3 million} +

{($8.1 million+$32.4 million) - ($18 million + $43.2 million)}

= - $7.85 million.

Note that the estimated savings would be expected to result primarily from network efficiencies in performing ongoing projects.

Further, since the typical project lasts 3 to 5 years, the annual cost savings would be on the order of $1.5 million to $2.5 million per year; this implies that the "time to break even" should be on the order of 1 year.

V. Conclusion

Beyond the potential value of developing a Neurological Emergencies Clinical Trials Network in promoting the social welfare, we estimate that developing such a network would make economic sense. Specifically, for a fixed portfolio of three large clinical trials, we estimate that for an initial investment of approximately $1.5 million, a network would result in a net savings of nearly $8 million. This savings would result primarily from efficiencies in the operation of trials, since modest decreases in per subject costs is multiplied over a large number of subjects.

In order to more precisely predict the cost impact of research performed in a Neurological Emergencies Clinical Trials Network, it would be essential to specify the projects in some detail. Despite the highly speculative nature of the current estimates, they do suggest that a Neurological Emergencies Clinical Trials Network could be economically feasible.

References:

Crowley WF, Jr., Sherwood L, Salber P, et al. Clinical Research In The United States At A Crossroads: Proposal For A Novel Public-Private Partnership To Establish A National Clinical Research Enterprise. JAMA 2004:291(9);1120-26.

Draper ES, Manktelow BN, McCabe C, Field DJ. The Potential Impact On Costs And Staffing Of Introducing Clinical Networks And British Association Of Perinatal Medicine Standards To The Delivery Of Neonatal Care. Arch Dis Child Fetal Neonatal Ed 2004:89(3); F236-40.

Hoffman JR, Wolfson AB, Todd K, Mower WR. Selective Cervical Spine Radiography In Blunt Trauma: Methodology Of The National Emergency X-Radiography Utilization Study (NEXUS). Ann Emerg Med 1998:32(4);461-9

Marmarou A. Conduct Of Head Injury Trials In The United States: The American Brain Injury Consortium (ABIC). Acta Neurochir Suppl (Wien) 1996:66;118-21.

Lee TS, Renaud EF, Hills OF. Emergency Psychiatry: An Emergency Treatment Hub-And-Spoke Model For Psychiatric Emergency Services. Psychiatr Serv 2003:54(12); 1590-1, 1594.

Marmarou A, Nichols J, Burgess J. Effects Of The Bradykinin Antagonist Bradycor (Deltibant, CP-1027) In Severe Traumatic Brain Injury: Results Of A Multi-Center, Randomized, Placebo-Controlled Trial. American Brain Injury Consortium Study Group. J Neurotrauma 1999:16(6);431-444.

McKinley KE, Bryan-Smith L, Dosch TL. A Hub-And-Spoke Model Of Care: Providing Specialty Care In Patients' Own Communities. Jr Comm J Qual Improv 2002:28(10); 574-5, 529.

Birkmeyer NJ, Weinstein JN, Tosteson AN. Design Of The Spine Patient Outcomes Research Trial (SPORT). Spine 2002:27(12); 1361-72.

Nobilio L, Ugolini C. Selective Referrals In A 'Hub And Spoke' Institutional Setting: The Case Of Coronary Angioplasty Procedures. Health Policy 2003:62(1)95-107.

Rienhoff O. Medical Research Networks--An International Comparison. Stud Health Technol Inform 2003:96; 163-7.

TASK FORCE REPORT

How Can Data Management Be Handled Most Efficiently?

Michael Fehlings, M.D.

Professor of Neurosurgery

Krembil Chair in Neural Repair and Regeneration

University of Toronto

Toronto

Roger Lewis, M.D., Ph.D.

Professor

UCLA Medical Center - Harbor

Department of Emergency Medicine

Torrence

Chelsea Kidwell, M.D.

Associate Professor

UCLA Medical Center

NIH Stroke Program

Sidney Starkman, M.D.

Professor

University of California, Los Angeles

Department of Emergency Medicine and Neurology

Los Angeles

Introduction

In order to facilitate multi-center clinical trials in emergency neurological conditions, it is critical that data be collected in an efficient, cost effective, and readily accessible manner that facilitates merging data from a variety of sources and takes into consideration current HIPAA compliance issues. The full report of this task force addresses critical issues related to 1) the type of data entry forms to use; 2) the challenges of correctly enrolling and tracking patients; 3) the need to merge data from a variety of sources; 4) the need for centralized data safety monitoring; 5) the protocol compliance; 6) the need to generate regular (e.g. monthly) data reports; 7) the need to be compliant with current HIPAA regulations. Based on current advances in web-based technologies, electronic data forms, availability and access to the internet, it is apparent that the management of the ENCTN would be greatly facilitated by an efficient web-based data management system.[1-6].

Key Issues to Handle:

1. Type of Data Entry Forms to Use

Many potential options for data entry are currently available each with their own advantages and disadvantages These are summarized in Table 1.

The group discussed many of the advantages and disadvantages of the various approaches outlined in Table 1. There was considerable enthusiasm for a hybrid model combining the use of paper records by a study coordinator that would subsequently be entered into the electronic database.

Table 1

Options for Data Entry	Advantages	Disadvantages
Paper records which are faxed to a centralized office.	Ease of data entry at study sites	Volume of transmitted paper. Need for optical recognition software to facilitate electronic registry.
Paper forms with subsequent entry into an electronic database with regular transmission of electronic records.	This approach potentially combines the advantages of direct paper entry with entry into an electronic record.	Data retrieval and monitoring of the study cannot be done instantaneously and requires a central collation of electronic records.
Use of electronic forms which are subsequently entered into an electronic database.	This avoids the use of paper records.	One does not have the backup of paper records.
Direct web-based data entry.	Allows for rapid online monitoring of the trial.	A labour intensive approach which may not be practical for many emergency neurology conditions.
A hybrid model between the use of paper records which are subsequently entered into an electronic database which is merged to internet-based approaches.		HIPAA compliance issues are critical with the use of internet-based approaches. Web site security is a critical issue.

The use of web-based data management systems has many advantages and some disadvantages which are summarized below in Table 2. However, the efficiencies of web-based data management systems are considerable and outweigh the disadvantages.

 

Table 2: Web-based Data Management Systems

Advantages

· Provides efficient system for maintaining up-to-date protocols and amendments

· Option for automatic queries for missing data

· Substantial decrease in cost (paper, mailing, personnel)

· Option for web-based randomization

· Worldwide access (ideal method for trial of large number of centers)

· Data captured in real time

· Edits can be logged and reports of edits available for review

· Option for automatic generation of reports

Disadvantages

· Need for high-speed internet access at sites

· Investigator comfort level

· Computer programming / IT expertise (up-front work and ongoing maintenance of database)

2. Challenges of Collecting Data-Advantages of a Web-Based Approach

It is critical that all patients are tracked in prospective clinical trials related to emergency neurological conditions. It is recognized that most patients may not meet all of the entry criteria. However, these data will still need to be tracked. The use of a web site to identify the entry and exclusion criteria was recognized as an extremely helpful approach, which could facilitate this. To maximize the efficiency of the ENCTN and the quality of collected data, a centralized electronic data collection strategy will be required. Desirable characteristics of this strategy include:

• The use of web-based data collection forms, allowing any institution with Internet access to directly enter patient data into the database. The web form should include real-time validation of entered data, using both value limits and internal consistency checks. The web form should interface with each study's database, so that each new patient can be immediately checked against the list of prior patients and multiple enrollments on the part of individual patients can be immediately detected.

• The web-based data collection strategy must conform to patient confidentiality and HIPAA guidelines through the use of industry-standard encryption and authentication strategies, but should not require the installation of specific client software at each institution.

• A similar web-based data collection strategy should be implemented for the reporting of adverse events (AEs), unexpected adverse events (UAEs), and significant adverse events (SAEs). The system should automatically link initial and follow-up SAE reports, and generate letters and/or electronic mail items to inform appropriate investigators, sponsors, and regulatory agencies in conformity with applicable adverse event requirements for the individual study.

1. Merging Data From a Variety of Sources

Imaging data including CT, magnetic resonance studies and plain X-rays are likely to be of critical importance in emergency neurology trials including stroke, traumatic brain injury, and spinal cord injury. It is important that electronic databases or internet-based data management tools are able to integrate these sources of data. It is essential that imaging protocols between various study centers be standardized as much as possible and that agreement be achieved in terms of the appropriate image formats to be used.

An existing shortcoming of hospital-based investigations of treatments and outcomes of neurologic emergencies is the general lack of integration of data from the prehospital setting and the integration of outcome data from other sources (e.g., other health care institutions, long-term care facilities, and coroner's records). In order to facilitate the probabilistic linking of study records to records in ancillary data sources, the information collected from each patient must include sufficient personally identifying information to facilitate record linkage. While the collection, transmission, and storage of personally identifying information makes the security of those data collection efforts even more critical, the potential benefit from being able to link information from disparate sources in aiding the understanding of the full range of health care system utilization and patient outcomes outweighs the difficulties likely to be encountered.

 

3.1 Neuroimaging

Web-based transfer of neuroimaging data offers a number of advantages as well as challenges in the clinical trial setting. Internet electronic transfer of data eliminates the need to copy and mail hardcopies of images. This reduces the potential for images to be unreadable or uninterpretable due to poor copy quality. HIPAA compliance and patient confidentiality are important challenges since electronic imaging data is likely to contain the patient's name and hospital ID number. Typically, a program is used to strip the patient identifiers from the header and replacing with a unique study ID number prior to image analysis.

3.2 Resources Needed for Electronic Neuroimaging Data Transfer and Storage

• Database programmer / manager with expertise in various imaging formats
• Standardization of imaging sequences and protocols across sites and scanners
• Protocols for image file format and data transmission
• Ongoing technical support
• Web site for secure/safe data transfer (e.g. SFTP, VPN)
• Sufficient disk space for data storage and back-up system

Program for de-identifying imaging data prior

1. Need for Centralized Data Safety Monitoring

Although it is acknowledged that data initially entered by study sites, even with the use of internal and external validity checking, will not result in data of a quality that allows conclusive and final analysis, the data will likely be of sufficient quality to support ongoing Data Safety Monitoring Board (DSMB) activities. Members of the DSMB, or at least the Chair, should have the capability of performing limited queries of each study and adverse event (AE) database, to support real-time data safety monitoring activities. In addition, the system should incorporate the possibility of setting specific "alarm" thresholds regarding types, numbers, and/or patterns of adverse events or study outcomes, resulting in the alerting of the DSMB Chair.

1. Protocol Compliance and Data Quality Issues

It is critical that data management tools be used which can be efficient in tracking outliers. This could include the detection of sites that are enrolling at much higher or at much lower than expected levels. This could also include detection of serious imbalances or variations from protocol compliance. The use of web-based data management tools would again allow this to be done in an efficient manner.

The strategies developed to ensure data quality should explicitly consider and address the possibility of scientific misconduct, including data fabrication. For each study, the use of "filters" should be considered to identify institutions, investigators, or research staff whose data submissions appear to differ systematically from those of other institutions, investigators, or staff. Paradoxically, these filters may identify for audit data that is collected surprisingly rapidly or is surprisingly complete. An ongoing strategy for ensuring data quality, which includes audits of submitted data during the conduct of the study, focusing on those institutions whose data appear atypical or those institutions enrolling the largest numbers of patients should be conducted, rather than simply submitting queries and verifying data at the end of the trial.

1. Generation of Regular (e.g. Monthly) Data Reports

The data management systems used should facilitate the generation of regular data reports that could allow the monitoring of the trial. Potentially any of the tools used in Table 1 could facilitate this type of an approach.

1. HIPAA Compliance Issues

It is critical that any data management tools be HIPAA compliant. It is important

that any patient identifiers be eliminated from the database records or from any of the imaging or laboratory reports included. There is potential sensitivity around the use of web-based data management systems with regard to HIPAA compliance issues. However, by maintaining tight web site security and by eliminating patient identifiers from the web-based transfer of information, the issue of HIPAA compliance should be manageable.

1. Summary

To maximize the efficiency of the Emergency Neurology research network and the quality of collected data, a centralized electronic data collection strategy will be required. As illustrated in Figure 1, the use of web-based approaches has several advantages. Capability to incorporate complex data from a variety of sources will need to be built in. This approach could facilitate the activities of a Data Safety Monitoring Board. HIPAA compliance issues are important to consider when designing the centralized data collection strategy. Several examples of successful application of web-based strategies to run clinical trials have been reported in the literature and attest to the utility and practicality of this approach.^1-6

Figure 1 Web-based Integration/Management of a Clinical Trials Network

Source: Marks RG et al Paradigm shifts in clinical trials enabled by information technology. Statistics in Medicine 2001; 20: 2683-2696

References

1. Boukhors, Y., et al., The use of information technology for the management of intensive insulin therapy in type 1 diabetes mellitus. Diabetes Metab, 2003. 29(6): p. 619-27.

2. Marks, R., et al., Enhancing clinical trials on the internet: lessons from INVEST. Clin Cardiol, 2001. 24(11 Suppl): p. V17-23.

3. Marks, R.G., M. Conlon, and S.J. Ruberg, Paradigm shifts in clinical trials enabled by information technology. Stat Med, 2001. 20(17-18): p. 2683-96.

4. Meadows, B.J., Eliciting remote data entry system requirements for the collection of cancer clinical trial data. Comput Inform Nurs, 2003. 21(5): p. 234-40.

5. Rangel, S.J., et al., Development of an internet-based protocol to facilitate randomized clinical trials in pediatric surgery. J Pediatr Surg, 2002. 37(7): p. 990-4; discussion 990-4.

6. Stahl, D.C., et al., Web Services-based Access to Local Clinical Trial Databases: A Standards Initiative of the Association of American Cancer Institutes. Proc AMIA Symp, 2003: p. 624-8.

TASK FORCE REPORT

What Are The Human Subjects Research Issues?

How Will Minority Recruitment Be Handled?

Michelle Biros, M.D., M.S.

Professor, Emergency Medicine

Hennepin Faculty - Hennepin County Medical Center

Minneapolis

Peter Panagos, M.D.

Assistant Professor of Emergency Medicine

Brown Medical School

Providence

James Quinn, M.D., M.S.

Research Director

Stanford University

Department of Emergency Medicine

Palo Alto

Jeff Saver, M.D.

Professor

UCLA Stroke Center

Department of Neurology

Los Angeles

Introduction

While there are many ethical aspects that apply to all human subjects research, the ENCTN must specifically safeguard rights of human subjects in research made vulnerable by a devastating disease or injury. These include, but are not limited to, the issues of meaningful consent for research participation, privacy of the medical record, and the adequate recruitment of minority patients into clinical trials. There are also unique aspects of clinical trials research and out of hospital research that pose special challenges for emergency neurology researchers.

 

Issues of consent [DLG2]

In terms of consent for research participation, patients with neurological emergencies fall into three categories; those with unquestionable capacity, those with questionable capacity, and those who clearly are not competent to give consent.

Category 1: Patients able to provide consent

The goal in any consent process is to allow patients to make autonomous informed decisions by being able to fully comprehend the nature and risks of participation in a particular research project. For a neuroemergency clinical trial, a patient who is able to give consent would be one with a minimal CNS insult that would not impair comprehension and understanding. Practically speaking the person obtaining the consent and the physician in charge of the care for that patient would be responsible for ascertaining the patient's ability to understand and comprehend. Such neurological conditions that may fit this criteria include minor closed head injuries (GCS >14), transient ischemic attacks (TIA), small vessel strokes, resolved seizures, isolated peripheral neuropathies, and spinal cord lesions.

Informed consent is an educational process, and the signature of the patient on the informed consent document testifies that the process has occurred and is understood. In reality, expedient patient care and the short therapeutic window of most investigational agents puts a time pressure on the consent process and precludes an extensive give and take interchange between the patient and the investigator. Much essential information is therefore relayed within the document itself. Standard consent forms are mandated to incorporate specific elements, such as specific language to satisfy HIPAA requirements,¹ and therefore have become complex legal documents that require educational levels that may preclude understanding by a large majority of patients. In a study done in 1997, it was shown that the average educational level required to read a consent document was 12^th grade ², and with increasing complexity and additional mandated language, even higher read skill is likely required today. Complex consent documents lose their ability to educate and inform patients and appear to discriminate against various socioeconomic and racial groups^3-5. The effort and time necessary to carefully read these documents, even for patients who can understand the language used, may bias patients to refuse participation in a study. For example, one study has demonstrated that among patients who participated in a clinical trial, less than 50% actually understood the requirements and risks of the trial they participated in⁶. Another study showed that with verbal explanation and consent, patients would agree to participate in a clinical trial at a rate of 87% while if actually required to read and sign a consent form, only 35% would participate ^7. In these situations, trial enrollment becomes only a fraction of what it could and should be and patients who participate may not be representative of the patients affected by the pathology and who may most benefit from the treatment being considered. In the end, this leads to costly and inefficient trials that, if completed, produce results with questionable validity and generalizabilty.

Solution

The Network will have to address these consent issues and ensure that consent is informed, efficient, allows opportunities for all minorities, women and children to participate and does not discriminate against those with limited formal education and poor socioeconomic backgrounds. Those seeking consent will be responsible for improving the patient's understanding of the trial and its risks while improving the likelihood of participation of patients.^{8, 9}This will involve a multi-modal approach: a uniform training of all site investigators on the process of informed consent, addressing local IRB concerns, recognition of the levels of comprehension of the general patient population and how patients can best be educated about each specific study protocol.

Category 2: Patients with questionable capacity

Broadly conceived, informed consent requires that an individual be informed about the nature and purpose of the research, that the decision to participate be voluntary, and that the person be competent to rationally understand the treatment risks and benefits. ¹⁰ Evaluation of a patient's capacity to provide research consent is a challenging aspect of research in neurological emergencies. Rapid assessment of the patient with questionable capacity is frequently required, as the neuroemergent patient is prone to multiple conditions that impair decision-making competence. The patient is generally in a setting of distress, and is asked to make a decision regarding a condition that did not exist minutes or hours earlier, about which they have no long-settled perspective.¹¹ The patient is often desperately ill, a setting in which immediate judgments of trust and hope may be weighed more heavily than pure rational calculation.¹² The patient is frequently an elder, with age-related impairments in cognitive function.¹³

Most important, acute brain injury frequently directly impairs one or more of the numerous cognitive capacities required for competent decision-making. Competent decision-making requires highly interactive functioning of numerous cognitive domains, including attention, memory, language, perception, calculation, executive function, and emotional/affective processing.^14-18 Simple bedside conversation with a patient may provide a misleading impression of decisional capacity. The patient with a profound neglect syndrome may be able to converse in a friendly and sophisticated manner, but be utterly unaware of the deficit that has brought them to the hospital and for which research intervention is being considered. Conversely, the patient with severe dysarthria may be unable to express their wishes verbally, but actually possess full cognitive competence. The "competence" construct must also be recognized as a continuum, rather than a dichotomy. The complexities of different research protocols and legal standards for competence vary.^16,19 Patients may have adequate capacity to appreciate and consent to simpler studies, but be incompetent to evaluate more complex studies.

Formal assessment of mental status and rendering of etiologic diagnosis do not definitively determine competency. Brief assessment of the key cognitive domains of orientation, attention, language, and memory are part of the routine clinical evaluation of patients with neuroemergencies and will help to identify individuals with profound impairments who are clearly not competent to make an informed decision. However, simple bedside mental-status screening techniques are insufficient to fully discriminate competent from non-competent patients,²⁰ and detailed neuropsychologic testing is not practical for conditions in which additional brain injury occurs minute-by-minute.

Solution

A functional ability-based evaluation is a promising approach to the rapid evaluation of competency in patients with neuroemergencies. This approach is based on an influential model of competency that recognizes four higher-level functional abilities that together comprise capacity: 1) the ability to communicate a decision, 2) the ability to understand factual information about the nature of the disorder and the risks and benefits of study participation, 3) the ability to appreciate the consequences of a participation decision, and 4) the ability to provide rational reasons for a participation decision. ^19,21 The assessment of these abilities can be carried out through a brief clinical interview or through the use of a standardized assessment tool.¹⁰ If all functional abilities central to the consent capacity are intact, the patient is determined competent.

Who should make the determination of capacity to consent is an additional issue of concern in neuroemergency research. In non-emergent situations, the National Bioethics Advisory Commission recommended: "an independent, qualified professional" not associated with the research team assess the potential subject's capacity to consent, to avoid the potential conflict of interest that occurs when research investigators themselves make a consent assessment. ²² However, the use of a disinterested third party to assess consent capacity is generally not practicable in neuroemergency research, where assessments must occur rapidly and unexpectedly. In this setting, formal procedures to guide investigators in assessing competence may be employed to guard against biased judgments. ^12,23

 

Category 3: Patients unable to provide informed consent

In certain situations, informed consent cannot be obtained from the potential study subject because the disease process has made the patient non-competent. The FDA's Final Rule for exemption from informed consent has been developed to provide specific patient safeguards in circumstances where consent cannot be obtained from the patient and the legally authorized representatives are not available to speak on behalf of he patient ²⁴. The Final Rule was designed to allow local IRBs and investigators some authority for determining how its regulations could be met, but include little specific guidance on how to fulfill its requirements. The issues that face investigators and IRBs in appropriate implementation of the Final Rule have been described in detail elsewhere ²⁵: these challenges will also face the Emergency Neurology Clinical Trials Network.

Lack of uniformity in the interpretation of the Final Rule exists within the FDA itself, and this has added to the challenge of successful and widespread application. For example, one of the 5 criteria required to be satisfied for a study to proceed under an exception from informed consent is that "research participation offers the possibility of direct therapeutic benefit to the subjects." The different branches of the FDA have interpreted this criterion variably. The drug evaluation branch of FDA (Center for Drug Evaluation and Research) has at times taken this criterion to require that only large, definitive phase 3 trials designed to fully evaluate efficacy may employ exception from informed consent procedures. Yet, it is unclear how these emergency treatments can safely advance to phase 3 trials if they cannot first be tested in phase 2 trials under similar research circumstances. In contrast, the device evaluation branch of FDA has interpreted this criterion less strictly, and permitted use of the emergency exception regulations in early safety and technical efficacy trials.

 

Solution

The FDA requirements are the currently the only existing regulatory mechanisms to allow research to go forward when prospective informed consent is not possible. The Emergency Neurology Clinical Trials Network should develop a cadre of experts who can assist local investigators and local IRBs in the ethical and practical fulfillment of the current requirements of the Final Rule. The Network should develop close liaisons within the FDA to assist in implementation of protocols that require the use of the exception from informed consent. The Network should also conduct concurrent research on the process of informed consent and exception from informed consent, so that a consensus, data based proposal can be advanced to the FDA to consider refinement or clarification of the existing regulations, where needed, in order to allow important studies to go forward in an ethical fashion when consent is not obtainable.

 

 

Confidentiality, Patient Tracking and HIPAA

In order to assure that study results can be generalized to the broader population, the Emergency Neurology Network must be able to offer study participation to all eligible patients, informing them of the risks in a way which does not discriminate against their race, gender, education or socioeconomic class. At each individual study site, all patients eligible for enrollment into an ongoing trial must be easily and appropriately identified. Eligible patients or their surrogates must be provided with enough time and information to allow them to understand the risks and benefits associated with research participation and make an informed decision. Unless processes that guarantee this are in place for the Network, the ability of the neurological network to recruit and offer participation to all eligible patients will be compromised, and trial results will be limited.

To promptly and efficiently identify all patients eligible for neurological emergency trials, a real time tracking and notification system will be necessary. Such systems, if used by people not involved in the care of the patients and only for research purposes, would require IRB authorization and a HIPAA waiver to track potential patients. While de-linking patient identifiers and complaints is possible, at some point this information will need to be combined to allow investigators to identify and approach potential patients for follow-up. This could be interpreted by some as directly in conflict with current HIPPA regulations. Limiting access to patient identifying information to only local site investigators involved in the enrollment and care of potential patients could potentially overcome HIPAA issues. However, in any research Network, information sharing must be possible; interpretation of HIPAA guidelines are currently institution and agency dependent, making this problematic for multi-centered research¹.

Another ethical issue regarding patient confidentiality arises when data or laboratory specimens (i.e. blood) are collected at the time of patient enrollment and saved for analysis at a later date. For example, once the patient identifiers are removed from the samples, what happens if later testing reveals a worrisome result? Does consent for enrollment into the study include banking of blood/tissue samples for future testing that had not yet been developed at the time of consent?

Beyond their ethical and regulatory importance, consent and privacy issues play important roles in the eventual success of clinical trials and the validity and generalizabilty of the results they produce. Any successful trial or network will have to successfully deal with these issues.

Solution

The Emergency Neurology Network should develop a panel of experts who can assist investigators, institutions and IRBs in ethical methods to comply with HIPPA while still sharing patient information among study investigators. Close liaison with regulators will be needed for assistance in consistent interpretation of the HIPPA requirements. Such a liaison will also provide an opportunity for continuous education for federal regulators and hospital administrators of the importance of neuroemergency research and the uniqueness of methods required to answer our study questions.

Equitable Enrollment of Study Subjects

A variety of novel approaches to achieve patient enrollment in neuroemergency clinical trials have evolved over the past decade. These accomplishments provide a foundation for enhanced enrollment in future studies. Some effective strategies include the following:

The "Commando Model": Having a core group of central study investigators personally go to several different hospitals in a region to enroll patients in acute trials is an effective way to bring acute trial expertise to the community hospital setting. The Greater Cincinnati Northern Kentucky Stroke Team is a paradigmatic example of this approach.²⁶

The Trauma System Model: Diverting patients to designated critical care facilities, such as Trauma Centers or Stroke Centers, where research is one aspect of institutional expertise, ensures greater access to research for patients in a region.

Dedicated Research Staff in ED: Several studies have demonstrated that full time staff placed in the ED solely to identify patients for clinical trials and perform screening evaluations substantially increases the successful recruitment of study subjects.^27-29 Use of undergraduate students or paramedics in this function and providing them academic credit or a small amount of monetary remuneration is cost-effective, and ensures continuous surveillance for patients when ED staff must attend to pressing clinical events.

Computerized Monitoring for Patients: Automated pager notification of the research team when information entered into the hospital's computerized registration system identifies a potential trial patient is an additional method that can ensure identification of potential study subjects when the attention of ED staff is directed to more pressing acute patients, precluding them from performing notification.³⁰

In addition to these tested approaches, several novel strategies of patient recruitment may be worthy of study. These include:

Telemedicine Consent for Research: Telemedicine is a powerful technique for emergent neurologic patients because the audiovisual examination of the patient is a key element of the neurologic evaluation. Telemedicine has been successfully employed by several groups to evaluate stroke patients, with the physician-expert at a central hospital and the patient at a remote hospital or in a moving ambulance with broadband transmission capability.^31-33 Employing telemedicine to evaluate patients for clinical trials and to remotely elicit informed consent for research is a natural extension of these current practices, and has proved successful in pilot application. ³⁴

Brief Initial Consent: Research consent forms are subject to ever-increasing length as efforts to protect the rights of research subjects through federal and state regulations continuously introduce new topics to be covered. Acute clinical trial consent forms now typically are 5-15 pages in length. Increased length of consent forms reduces readability and comprehension of core material, even in leisurely outpatient settings,³⁵ likely impairing a patient's ability to identify and focus on the most critical elements of a study description in the emergency setting. A two-stage consent form process may be envisioned for neuroemergency clinical trials as an alternative to the current, single, encyclopedic consent form strategy. In the two-stage process, the consent provider (patient or proxy) would first read an abbreviated consent from of no more than one page that covered only the most critical aspects of study design and subject rights. An initial decision to participate would be based on this consent form, permitting study therapies to be initiated before too much additional nervous system injury transpired. The consent provider would then read a comprehensive consent form that provided all study details. At this point, a final decision would be made regarding study entry and participation. Implementation of this two-stage consent process might require revision of current consent regulations and the refinement of existing statistical techniques to handle staged patient trial entry, but would likely facilitate both greater informed consent and earlier enrollment in neuroemergency trials.

Solutions

While it may not be possible to control for consistent recruitment and absolute identification of all patients eligible for a particular study at a particular site, one important variable that the Network can control is supporting local sites to help recruit, educate and help enroll patients. Studies have shown that enrollment varies tremendously depending on who seeks consent ^3,8,9. A successful network should provide education, training and support for those seeking consent at the individual sites. To make this feasible there should be a critical number of trials going on at each site so that at least one patient/day is enrolled from each ED site into a trial. This would allow a critical mass of patients to maintain the system infrastructure for each site and maintain the skill sets at the site for those performing the trial (recruitment, enrollment and consent, data collection). It would also allow an infrastructure to do trials and recruit patients with rare neurological conditions. Finally sharing experiences and performance at each site will allow sites to maximize skills towards recruitment and enrollment and maintain momentum for the trials network ³⁶. Institution-specific techniques should be supported by the Network to enhance study patient recruitment. Continuing medical education and study updates should be provided to maintain interest and ensure the retention of experienced research assistants.

 

Special issues related to Research in Neuroemergencies in Minority Populations:

Acute Ischemic Stroke

Acute stroke is one of the highest prevalence neurological emergencies encountered in the emergency department setting. Stroke mortality rates may differ substantially by sex and race. There are also substantial geographic variations in stroke mortality rates, with residents of the traditional 'stroke belt' having a mortality ratio approximately 40% higher than the rest of the US. Several population-based studies, such as Rochester, Greater Cincinnati/Northern Kentucky Stroke Study (GCNK) and the Northern Manhattan Stroke Study (NOMASS) have demonstrated a significant racial disparity in all-type stroke incidence rates and risk factors among different racial-ethnic groups. ^37-40 For patients seeking treatment for an ischemic stroke, intravenous (IV) tPA is less frequently used in African American patients presenting within 3 hours of symptom onset. ⁴¹ Gender also plays a role in management and outcome. In a large sample of European stroke patients, female sex was associated with a higher short-term mortality and was a major discriminating factor for the use of diagnostic resources or therapeutic interventions.⁴² As medical providers, it is important to be aware of these disparities in care.

Using stroke as example, there are several possible explanations for the disparity in ethnicity and gender for these patients. Patients often delay presenting to the hospital for treatment. This may be due to language and cultural barriers, and perhaps a degree of distrust in the medical system and lack of confidence in treatment. Real ethnic variation in risk factors and stroke subtypes also exist. Primary care and stroke experts may not be aware of these differences. We practice in a medical system that is biased against ethnicity and socioeconomic status, language, with multiple educational and cultural barriers in place. Therefore, it is easier to maneuver through if English speaking, insured and educated. Additionally, many hospitals serving larger minority populations lack the research and clinical infrastructure needed to perform clinical research due to the reimbursement structure. Researchers are less likely to look towards these institutions if the emotional and financial cost of performing research is high.

We must understand that ethnicity is multi-factorial, composed not just of skin color or language, but a mixture of cultural, religious, linguistic, dietary, political and other variables that identifies individuals as belonging to a group or population. Additionally, our population-based research demonstrates that ethnicity and gender is a risk factor or marker for a collection of risk factors with complex relationships, both simple (e.g. income) and complex (e.g. racism/sexism). These risk factors led to serious physiologic impact on individual health.

 

Solutions

Once recognized, several solutions to these issues exist. First, the medical community should be aware of individual risk factors and inherent treatment disparities. Once recognized, these differences in stroke etiology, severity and prognosis can be addressed. In many communities, an ethnic discordance between medical team and patient exists. Successful research teams must be sensitive to language and cultural diversity within their communities while seeking out education and community involvement. They must promote teams containing members sharing similar lineage as the populations being served. Finally, education on specific disease prevalence, recognition, treatment and prevention must begin in the schools. Children are ultimately their parents' and grandparents' keepers and will bring this information home with them.

The unequal burden of several diseases among minority populations and the undeserved is well described. Minority groups with diverse features related to socioeconomic status, age, and sex, often receive disparate medical care. It is further evident that these disparities lead to a reduction of minorities in clinical trials. Once the barriers of clinical-trial enrollment of minority populations are recognized, these populations can begin to benefit from trials that investigate quality of life, increased survival and improved access to medical care

Circumstances Unique to Out- of- -Hospital Research in Neuroemergencies

Recent studies have demonstrated the important role of rapid transport by emergency medical services (EMS) to the hospital in the care of the patient sustaining an acute neurological insult. ^43,44 EMS personnel often are the first medical contact for patients sustaining a stroke or traumatic brain injury. This allows EMS personnel the rare opportunity to provide early assessment and treatment. Today, there are barriers to timely and effective care in the out-of-hospital setting. Patients with brain injury resulting from a traumatic brain injury (TBI), acute stroke or cardiac arrest have speech disorders, depressed level of consciousness and are often confused. As a consequence, these individuals are vulnerable, have a diminished autonomy and require explicit protection. ^44-46

While current therapies available to physicians for the treatment of many neurological emergencies are less than perfect, reasons for optimism exist. Research in brain function and dysfunction is beginning to elucidate the underlying physiology of cause of damage and possible effective therapies, and any of these new therapies will need to be tested in well-designed clinical trials. Unique to the practice of emergency medicine, often the physician or EMS agency (practicing under physician license) starts care that is deemed appropriate without an informed consent, particularly if the patient is unable to communicate their actual desire. But in emergency neurological research, the issue is less clear and the only realistic way to conduct clinical research is to enroll patients without their consent.

To date, research in the out-of -hospital setting for the impaired patient has been difficult to conduct. Aside from the obvious issues inherent to research outside the safe walls of the hospital, stricter guidelines for obtaining consent were enacted by the US government in 1993, essentially shutting down much resuscitation research. The 1996 FDA/DHHS Final Rules for granting exemption from informed consent are available, but limited guidance exists for applying them, especially in the out-of-hospital setting. Other barriers still exist that make brain injury research outside the hospital less likely to succeed and include:

1. Inconsistent and variable local community EMS training and funding

2. Lack of resuscitation studies with proper scientific and therapeutic benefit

3. Difficulty obtaining consent in a time-sensitive manner from the patient or an appropriate proxy

4. Significant time delays from disease onset until EMS activation.

Solutions

Several possible solutions to these issues could improve this situation. There

are now numerous quality research projects with indisputable scientific equipoise being

proposed. The federal government could offer strong public support for the initial exception from consent involving research on vulnerable patient populations. This would offer local IRB committees some assurance of safety in granting consent exceptions for impaired patients. Led by individuals interested in advancing the study of neurological emergencies, an improved EMS curriculum centered on neurological injury identification, assessment and management, including the consent process, could be designed (e.g. within BLS, ACLS and EMS curriculums). In addition, several groups have proposed the utilization of technology such as cellular phones and televideo to assist in out - of- -hospital care management.⁴⁷ This could allow decisions to be made up to 30-40 minutes earlier in a disease where minute delays really do count.^48,49 Finally, studies show that the majority of patients don't arrive in the ED early enough to diagnosis and treat acute stroke with thrombolytics or other therapies with a short time window. Public knowledge of stroke signs and symptoms, combined with the patients impaired ability to communicate leads to delay in presentation for care. ^50-54 Focused behavioral interventions directed at the family and coworkers who frequently contribute to delays in EMS activation and hospital arrival should be designed.

Summary

Most issues associated with informed consent and privacy issues will be common for all centers participating in the network. The goals of the Emergency Neurology Clinical Trials Network will be to inform and educate patients and to protect patient privacy. It thus makes sense that an infrastructure is put in place to help individual sites deal with these issues. Site investigators will need to be aware of local and state regulations that may be unique to their circumstance, however they should be supported when local site interpretation and requirements become problematic and effect recruitment and enrollment. Furthermore individual site access to hospital and patient data for real-time tracking is another issue and may become problematic as hospitals start to limit access to tighten perceived security and privacy issues, as well as limit time and demands on their IT resources. Advisement from an expert panel, developed from within and outside of the Network, may help investigators address these problems in an ethical and practical manner. Such a panel could also provide surveillance for additional unanticipated ethical dilemmas arising as research moves forward, and address other concerns that could impact patient or surrogate acceptance and enrollment into research studies at critical, stressful times.

References

1. Rodriguez TA. HIPAA privacy regulations clarified: let calm prevail. J Med Pract Manage 2003; 19(2):61-6.

2. Mader TJ, Playe SJ. Emergency medicine research consent form readability assessment. Ann Emerg Med 1997; 29(4):534-9.

3. Gorkin L, Schron EB, Handshaw K et al. Clinical trial enrollers vs. nonenrollers: the Cardiac Arrhythmia Suppression Trial (CAST) Recruitment and Enrollment Assessment in Clinical Trials (REACT) project. Control Clin Trials 1996; 17(1):46-59.

4. Crowley R, Casarett D. Patients' willingness to participate in symptom-related and disease-modifying research: results of a research screening initiative in a palliative care clinic. Cancer 2003; 97(9):2327-33.

5. Wolf AM, Nasser JF, Wolf AM, Schorling JB. The impact of informed consent on

patient interest in prostate-specific antigen screening. Arch Intern Med 1996;

156(12):1333-6.

6. Criscione LG, Sugarman J, Sanders L, Pisetsky DS, St Clair EW. Informed consent in a clinical trial of a novel treatment for rheumatoid arthritis. Arthritis Rheum 2003; 49(3):361-7.

7. Brod MS, Feinbloom RI. Feasibility and efficacy of verbal consents. Res Aging 1990; 12(3):364-72.

8. Isaacman DJ, Reynolds EA. Effect of a research nurse on patient enrollment in a clinical study. Pediatr Emerg Care 1996; 12(5):340-2.

9. Tercyak KP Jr, Johnson SB, Kirkpatrick KA, Silverstein JH. Offering a randomized trial of intensive therapy for IDDM to adolescents. Reasons for refusal, patient characteristics, and recruiter effects. Diabetes Care 1998; 21(2):213-5.

10. Marson D, Dymek M, Geyer J. Informed consent, competency, and the neurologist.

Neurologist 2001;7(6):317-326.

11. Agard A, Hermeren G, Herlitz J. Patients' experiences of intervention trials on the

treatment of myocardial infarction: is it time to adjust the informed consent

procedure to the patient's capacity? Heart 2001;86(6):632-637.

12. Bosk CL. Obtaining voluntary consent for research in desperately ill patients. Med

Care 2002;40(9 Suppl):V64-68.

13. Cherniack EP. Informed consent for medical research by the elderly. Exp Aging Res

2002;28(2):183-198.

14. Alexander MP. Clinical determination of mental competence. A theory and a

retrospective study. Arch Neurol 1988;45(1):23-26.

15. Kim SY, Karlawish JH, Caine ED. Current state of research on decision-making

competence of cognitively impaired elderly persons. Am J Geriatr Psychiatry

2002;10(2):151-165.

16. Marson DC, Chatterjee A, Ingram KK, Harrell LE. Toward a neurologic model of

competency: Cognitive predictors of capacity to consent in Alzheimer's disease

using three different legal standards. Neurology 1996;46(3):666-672.

17. Holzer JC, Gansler DA, Moczynski NP, Folstein MF. Cognitive functions in the

informed consent evaluation process: a pilot study. J Am Acad Psychiatry Law

1997;25(4):531-540.

18. Moser DJ, Schultz SK, Arndt S, Benjamin ML, et al. Capacity to provide informed

consent for participation in schizophrenia and HIV research. Am J Psychiatry

2002;159(7):1201-1207.

19. Appelbaum PS, Roth LH. Competency to consent to research: a psychiatric

overview. Arch Gen Psychiatry 1982;39(8):951-958.

20. Kim SY, Caine ED. Utility and limits of the mini mental state examination in

evaluating consent capacity in Alzheimer's disease. Psychiatr Serv

2002;53(10):1322-1324.

21. Grisso T. Evaluating Competencies: Forensic Assessments and Instruments. New

York: Plenum Press, 1986.

22. Commission NB. Research Involving Persons with Mental Disorders that May Affect

Decisionmaking Capacity. Rockville, MD: National Bioethics Advisory Commission,

1998

23. Miller F, Rosenstein D. Independent capacity assessment: a critique. Bio Law

1999;11:S432-S439.

24. Department of Health and Human Services. Protection of human subjects ;informed

consent and waiver of consent requirements in certain emergency research.

(codified at 21CFR 50, et.al., 45CFR 46) Federal Register, ( Oct 2, 1996),61(

192):51500-33.

25. Biros MH. Research without consent; current status , 2003. Ann Emerg Med 2003;

452 ( 4) : 550-564.

26. Persse D, Hinton RC, Acker JE. Templates for organizing stroke triage. In:

Improving the Chain of Recovery for Acute Stroke in Your Community. Bethesda,

MD: National Institutes of Health, 2003

27. Hollander JE, Singer AJ. An innovative strategy for conducting clinical research: the

academic associate program. Acad Emerg Med 2002;9(2):134-137

28. Cobaugh DJ, Spillane LL, Schneider SM. Research subject enroller program: a key

to successful emergency medicine research. Acad Emerg Med 1997;4(3):231-233.

29. Bradley K, Osborn HH, Tang M. College research associates: a program to increase

emergency medicine clinical research productivity. Ann Emerg Med 1996;28(3):328-

333.

30. Weiner DL, Butte AJ, Hibberd PL, Fleisher GR. Computerized recruiting for clinical

trials in real time. Ann Emerg Med 2003;41(2):242-246.

31. Wang S, Lee SB, Pardue C, Ramsingh D, Waller J, Gross H, et al. Remote

evaluation of acute ischemic stroke: reliability of National Institutes of Health Stroke

Scale via telestroke. Stroke 2003;34(10):e188-191.

32. Levine SR, Gorman M. "Telestroke" : the application of telemedicine for stroke.

Stroke 1999;30(2):464-469.

33. LaMonte MP, Cullen JF, Gagliano DM, Gunawardane R, Hu P, Mackenzie C, et al.

TeleBAT: Mobile telemedicine for the Brain Attack Team. J Stroke Cerebrovasc Dis

2000;9:128-135.

34. Saver JL, Kidwell C, Eckstein M, Starkman S, for the FAST-MAG Pilot Trial

Investigators. Prehospital neuroprotective therapy for acute stroke: results of the

Field Administration of Stroke Therapy - Magnesium (FAST-MAG) Pilot Trial. Stroke

2004;e106-e108.

35. Baker MT, Taub HA. Readability of informed consent forms for research in a

Veterans Administration medical center. JAMA 1983;250(19):2646-2648.

36. Grant CH 3rd, Cissna KN, Rosenfeld LB. Patients' perceptions of physicians

communication and outcomes of the accrual to trial process. Health Commun

2000;12(1):23-39.

37. Brown RD, Whisnant JP, Sicks JD, et al. Stroke incidence, prevalence and survival:

secular trends in Rochester, Minnesota through 1989. Stroke;27:373-380.

38. Broderick J, Brott T, Kothari R, Miller R, et al. The Greater Cincinnati/Northern

Kentucky Stroke Study: preliminary first-ever total incidence rates of stroke among

blacks. Stroke 1998;29:415-421.

39. Sacco RL, Boden-Albala B, Gan R, et al. Stroke incidence among white, black and

Hispanic residents of an urban community: the Northern Manhattan Stroke Study.

Am J Epidemol 1998;147:259-268.

40. Kissela B, Schneider A, Kleindorfer D, et al. The excess burden of stroke among

blacks. Stroke 2004;35:426-431.

41. Johnston SC, Fung LH, Gillum LA, et al. Utilization of intravenous tissue-type

plasminogen activator for ischemic stroke at academic medical centers: The

influence of ethnicity. Stroke 2001;32:1061-1068

42. DiCarlo A, Lamassa M, Baldereschi M, et al. Sex differences in the clinical

presentation, resource use, and 3-month outcome of acute stroke in Europe: data

from a multicenter multinational hospital-based registry. Stroke 2003;34:1114-1119.

43. Barsan WG, Brott TG, Broderick JP, et al. Time of hospital presentation in patients

with acute stroke. Arch Intern Med 1993;153:2558-2561.

44. Kothari R, Sauerbeck L, Jauch E, et al. Patients' awareness of stroke signs,

symptoms, and risk factors. Stroke 1997:28:1871-1875.

45. Bircher NG. Resuscitation research and consent: Ethical and practical issues. Crit

Care Med 2003;31 (5 Supplement):S379-S384.

46. Kim DT, Spivey WH. A retrospective analysis of institutional review board and

informed consent practices in EMS research. Ann Emerg Med 1994;23:70-74.

47. LaMonte M, Bahout M, Hu P, et al. Telemedicine for acute stroke: triumphs and

pitfalls. Stroke 2003;34:725-728.

48. Crocco T, Gullett T, Davis SM, et al. Feasibility of neuroprotective agent

administration by prehospital personnel in an urban setting. Stroke 2003;34:1918-

1922.

49. Kidwell CS, Starkman S, Eckstein M, et al. Identifying stroke in the field:

prospective validation of the Los Angles stroke screen (LAPSS). Stroke

2000;31:71-76.

50. Pancioli AM, Broderick J, Kothari R, et al. Public perception of stroke warning signs

and knowledge of potential risk factors. JAMA 1998;279:1288-1292.

51. Morganstern LB, Staub L, Chan W, et al. Improving delivery of acute stroke

therapy: The TLL Temple Foundation Stroke Project. Stroke 2002;22:160-166.

52. Morris DL, Rosamond W, Madden K, et al. Prehospital and emergency department

delays after acute stroke: the Genentech stroke presentation survey. Stroke

1996;31:2585-2590.

53. Rosamond WD, Gorton RA, Hinn AR, et al. Rapid response to stroke symptoms:

the Delay in Accessing Stroke Healthcare (DASH) Study. Acad Emerg Med.

1998;5:45-51.

54. Schroeder ED, Rosamond WD, Morris DL, et al. Determinants of use of emergency

medical services in a population with stroke symptoms: the Second Delay in

Accessing Stroke Healthcare (DASH II) Study. Stroke 2000;31:2591-2596.

STEERING COMMITTEE AND TASK FORCE MEMBERS

Co-Chairs

 

Arthur Pancioli, MD William Barsan, M.D.

Associate Professor and Vice-Chairman Professor and Chair

Department of Emergency Medicine University of Michigan

University of Cincinnati Department of Emergency Medicine

Cincinnati, OH Ann Arbor, MI

NINDS Program Director

 

Robin Conwit, M.D.

Program Director, Clinical Trials

National Institute of Neurological Disorders and Stroke

National Institutes of Health

Bethesda, MD

 

Task Force Members

Michelle Biros, M.D., M.S. Professor, Emergency Medicine Hennepin County Medical Center HCMC and University of Minnesota Minneapolis, MN Guy Clifton, M.D. Professor and Chairman The University of Texas Medical School Department of Neuosurgery Houston, TX John Duldner, M.D., M.S. Research Director and Assistant Professor Akron General Medical Center Department of Emergency Medicine Strongsville, OH Michael Fehlings, M.D., Ph.D., FRCSC Professor of Surgery Director, Neural and Sensory Science Program Toronto Western Hospital Toronto, Ontario, Canada Judd Hollander, M.D. Professor of Emergency Medicine University of Pennsylvania Department of Emergency Medicine Philadelphia, PA Stephan Mayer, M.D. Associate Professor Columbia University Department of Neurology New York, NY Lewis Morgenstern, M.D. Director of the Stroke Program University of Michigan Department of Neurology and Emergency Medicine Ann Arbor, MI Peter Panagos, M.D. Assistant Professor of Emergency Medicine Brown Medical School Department of Emergency Medicine Providence, RI James Quinn, M.D., M.S. Research Director Stanford University Department of Emergency Medicine Palo Alto, CA Claudia Robertson, M.D. Professor of Neurology Baylor College of Medicine Department of Neurology Houston, TX Jeff Saver, M.D. Professor of Neurology UCLA Stroke Center Department of Neurology Los Angeles, CA Sidney Starkman, M.D. Professor Department of Emergency Medicine and Neurology Los Angeles, CA Ian Stiell, M.D., MSC, FRCPC Professor Ottawa Health Research Institute Department of Emergency Medicine Ottawa, Ontario, Canada Melinda Kelley, Ph.D. Science Advisor Office of Science Policy and Planning National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, MD Naomi Kleitman, Ph.D. Program Director Repair and Plasticity National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, MD John Marler, M.D. Associate Director for Clinical Trials National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, MD Mary Ellen Michel, Ph.D. Program Director National Institute of Neurological Disorders and Stroke National Institutes of Health Rockville, MD Audrey Penn, M.D. Deputy Director National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, MD Bernard Ravina National Institute of Neurological Disorders and Stroke National Institutes of Health Rockville, MD Frances Yee, Ph.D. Scientific Program Analyst Clinical Trials National Institute of Neurological Disorders and Stroke National Institutes of Health Rockville, MD

Edward Jauch, M.D. Assistant Professor University of Cincinnati Department of Emergency Medicine Cincinnati, OH Chelsea Kidwell, M.D. Associate Professor UCLA Medical Center NIH Stroke Program Bethesda, MD Roger Lewis, M.D., Ph.D. Professor of Emergency Medicine UCLA Medical Center - Harbor Department of Emergency Medicine Torrance, CA Dan Lowenstein, M.D. Professor of Neurology University of California - San Francisco Department of Neurology San Francisco, CA David Matchar, M.D., F.A.C.P. Director and Professor of Medicine Duke University Department of Medicine Durham, NC Gretchen Tietjen, M.D. Professor and Chair Medical College of Ohio Department of Neurology Toledo, OH David Wright, M.D. Assistant Professor and Assistant Director Emergency Medicine Research Center Emory University Atlanta, GA NINDS STAFF Constance Atwell, Ph.D. Director Division of Extramural Research National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, MD Marian Emr Director Office of Communications and Public Liaison National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, MD Brandy Fureman, Ph.D. Office of Science Policy and Planning National Institute of Neurological Disorders and Stroke National Institutes of Health Bethesda, MD Ronnie Horner, Ph.D. Program Director Office of Minority Health and Research National Institute of Neurological Disorders and Stroke National Institutes of Health Rockville, MD Margaret Jacobs Program Director National Institute of Neurological Disorders and Stroke National Institutes of Health Rockville, MD