NIH EMERGENCY MEDICINE RESEARCH ROUNDTABLES/SPECIAL CONTRIBUTION
NIH Roundtable on Emergency Trauma ResearchCharles B. Cairns, MD, Ronald V. Maier, MD, Opeolu Adeoye, MD, Darryl Baptiste, PhD, William G. Barsan, MD,
Lorne Blackbourne, MD, Randall Burd, MD, PhD, Christopher Carpenter, MD, MSc, David Chang, PhD, William Cioffi, MD,Edward Cornwell, MD, J. Michael Dean, MD, MBA, Carmel Dyer, MD, David Jaffe, MD, Geoff Manley, MD, PhD,
William J. Meurer, MD, MS, Robert Neumar, MD, PhD, Robert Silbergleit, MD, Molly Stevens, MD, MSCE, Michael Wang, MD,Debra Weiner, MD, PhD, David Wright, MD
Roundtable External Participants and Roundtable Steering Committee and Federal Participants*
Study objective: The National Institutes of Health (NIH) formed an NIH Task Force on Research in EmergencyMedicine to enhance NIH support for emergency care research. The NIH Trauma Research Roundtable wasconvened on June 22 to 23, 2009. The objectives of the roundtable are to identify key research questionsessential to advancing the scientific underpinnings of emergency trauma care and to discuss the barriers andbest means to advance research by exploring the role of trauma research networks and collaboration betweenNIH and the emergency trauma care community.
Methods: Before the roundtable, the emergency care domains to be discussed were selected and experts in each ofthe fields were invited to participate in the roundtable. Domain experts were asked to identify research priorities andchallenges and separate them into mechanistic, translational, and clinical categories. During and after theconference, the lists were circulated among the participants and revised to reach a consensus.
Results: Emergency trauma care research is characterized by focus on the timing, sequence, and timesensitivity of disease processes and treatment effects. Rapidly identifying the phenotype of patients on the timespectrum of acuity and severity after injury and the mechanistic reasons for heterogeneity in outcome areimportant challenges in emergency trauma research. Other research priorities include the need to elucidate thetiming, sequence, and duration of causal molecular and cellular events involved in time-critical injuries, and thedevelopment of treatments capable of halting or reversing them; the need for novel experimental models ofacute injury; the need to assess the effect of development and aging on the postinjury response; and the needto understand why there are regional differences in outcomes after injury. Important barriers to emergency careresearch include a limited number of trained investigators and experienced mentors, limited researchinfrastructure and support, and regulatory hurdles.
Conclusion: The science of emergency trauma care may be advanced by facilitating the following: (1) development ofan acute injury template for clinical research; (2) developing emergency trauma clinical research networks; (3)integrating emergency trauma research into Clinical and Translational Science Awards; (4) developing emergencycare–specific initiatives within the existing structure of NIH institutes and centers; (5) involving acute trauma andemergency specialists in grant review and research advisory processes; (6) supporting learn-phase or small, clinicaltrials; (7) performing research to address ethical and regulatory issues; and (8) training emergency care investigatorswith research training programs. [Ann Emerg Med. 2010;56:538-550.]
SEE RELATED ARTICLES AND EDITORIALS,P. 522, 551, 565, 568.
INTRODUCTIONThe National Institutes of Health (NIH) Roundtable on
Emergency Medicine Research on Trauma, cosponsored by theNational Institute of General Medical Sciences and the National
Institute for Neurological and Stroke Disorders (NINDS), was
538 Annals of Emergency Medicine
held on June 22 and 23, 2009. The overall goal of theroundtable was to engage basic, translational, and clinicalresearchers in a discussion to assist the NIH in prioritizingsupport for research in acute injury conditions, with anemphasis on time-sensitive diseases requiring rapid evaluationand treatment. Emergency trauma research includes discoveryand characterization of the mechanisms and response to injury,
translational research, and applied clinical research.1-3
Volume , . : November
Cairns et al NIH Roundtable on Emergency Trauma Research
The primary challenge of emergency trauma care involvesefficient management of undifferentiated patients with acuteinjury that have the potential for significant morbidity andmortality, and in whom the effectiveness of therapy is timecritical in the range of minutes to hours.3 Mechanistic researchthat delineates the timing, sequence, and duration of causalmolecular and physiologic events in acute injury is essential todeveloping optimized diagnostic and therapeutic strategies.
Clinical research that improves the accuracy, speed, safety,and cost-effectiveness of diagnosis and treatment in theemergency care setting has the potential to both improveoutcomes and reduce resource use.1,3,4 Real-time bedsidestrategies that guide diagnostic and therapeutic approaches andmonitor response to therapy will facilitate personalizedoptimization of care interventions. Finally, research intodiagnostic and therapeutic strategies that take into account thecharacteristics of the emergency care environment (eg, patientpopulations defined by broad clinical presentations and out-of-hospital settings) is most likely to have lasting and broadinfluence.
Unique aspects of emergency trauma clinical researchbecause of the time sensitivity of disease and critical conditionof patients include consent and enrollment issues, conduct ofresearch in the out-of-hospital setting, a wide range of patientages, and the effect of environmental factors on the hostresponse to injury.1-4 The effect of emergency trauma researchon society could be profound because injury remains the leadingcause of productive years lost in the United States.1,5
MATERIALS AND METHODSGoals of the Roundtable
In preparation for the roundtable, the NIH released a requestfor information in June 2008 to solicit input on needs andchallenges in the field of emergency medicine research.6
Respondents to the request for information individuallyidentified with a large group of professional organizationsspanning the adult and pediatric fields of emergency medicine,trauma, and critical care, along with nursing and emergencymedical services (EMS) providers. The largest group of disease-oriented responses to the request for information was classifiedunder the topics of general trauma and neurologic injuries,including traumatic brain injury and spinal cord injury. Otherinjury areas identified included burn injury, out-of-hospitalenvironments, and special populations such as children andgeriatric patients. Within these injury domains, the non-disease-related responses were largely related to diagnosis/monitoring,therapeutic interventions, out-of-hospital care, andepidemiology.
There have been few rigorous scientific evaluations ofimportant questions in these areas. Of even more importance isthe fundamental lack of a process to perform programmaticresearch in out-of-hospital (EMS) environments and to translateinjury-related research from adults to children (and potentially
vice versa).
Volume , . : November
Other fundamental issues identified included challenges inconsent, the time-sensitive interactive pathobiology amongmultiple organ systems, and use of undifferentiated signs andsymptoms involved in emergency trauma research. This researchinherently crosses the missions of multiple individual NIHinstitutes and centers. The structure and orientation of the NIHand its institutes and centers as a cluster of silos was perceived asa barrier to emergency medicine research.
The overall goal of this roundtable was to have an opendiscussion about emergency trauma research and to highlightthe gaps that should be of high priority for research support.The primary intended audience for this report will be the officeof the Director of the NIH, along with leadership of the NIHinstitutes and centers, the recently developed Emergency CareCoordination Center, the US Department of Health andHuman Services, and researchers in acute illness and injury.
RESEARCH AGENDA: PROPOSALS ANDCHALLENGESClinical Trials in Emergency Trauma
Injury remains the leading cause of productive years of lifelost in the United States.1 Furthermore, rapid diagnosis andearly treatment in patients with acute injury have been shown toimprove patient outcomes, prevent long-term disability, anddecrease health resource use.3
The intersection of the timing and severity of injury and itsimportance to defining the “acute injury phenotype” is an areain which greater discovery will be important to the design offuture clinical research. Understanding the timing and sequenceof events on the patient level leading to the progression of injuryis fundamental to the conduct of successful studies andgeneration of high-quality data.
The emergency department (ED) and out-of-hospitalsettings represent unique potential laboratories for injuryresearch because of their all inclusive populations, includinggroups frequently underrepresented in medical research(minorities, economically disadvantaged, lower level ofeducation, lacking in health insurance, all ages).3,4 In addition,these settings allow for collaboration and integration betweenmultiple specialties and disciplines.1,2 These settings are alsounique because of the scope and severity of clinically importantconditions.7 For example, multiple organ system trauma isimportant and common in clinical practice, yet it isuncommonly incorporated into experimental research models.This discordance results in an important void of prospectivehuman clinical trials in favor of retrospective studies based onobservational data sets, often with limited data. The clinicalcomplexity of trauma will require collaboration between basic,translational, and clinical scientists for experimental modeling ofbench-to-bedside research addressing mechanisms, diagnosis,treatment, and prevention of traumatic injury.
Emergency trauma research will require a betterunderstanding of events and interventions early in thepostinjury period. Little is known about traumatic events in the
out-of-hospital setting.8 Thus, research into these events will
Annals of Emergency Medicine 539
NIH Roundtable on Emergency Trauma Research Cairns et al
require research across multiple settings, including the out-of-hospital (EMS), ED, and inhospital locations.
The response to injury across the developmental spectrumfrom infancy through geriatric age needs further definition. Afundamental question in postinjury deterioration in geriatricpatients is whether it is the result of differences caused by agingor the result of preexisting comorbid disease. Conversely, ininfants and children, an understanding of the effect of earlydevelopment will be required to further characterize the injuryresponse.
Incorporation of an acute care phenotype may requiredifferent research approaches and alternative trial designs. Theseapproaches could include a therapy individually optimized toeach individual in the trial (similar to early goal-directed therapyin sepsis). This would tailor the phenotyping of the acute injurystate to the treatment goals and study outcomes. Time-dependent surrogate outcomes for injury interventions are alsoneeded. This need for individualization and time-dependentsurrogates has not been addressed well by contemporary clinicaltrial designs.
There are challenges in the consent and enrollment processthat are unique to emergency research.1-4 The time-sensitivenature and severity of the disease frequently preclude thepossibility of standard consent processes. Although there areprovisions to allow for studies with waiver or exception frominformed consent for emergency research, these consent optionsare difficult to implement and uncommonly used.3 Challengesto the use of expedited or waived consent provisions includevarying comfort of local institutional review boards in theinterpretation of “common rule” provisions, challenges andcosts of community consultation, and lack of standardization inpublic disclosure.3 Because of these challenges, there is thepotential for bias against acute care research because of a lack ofstandardized approaches or lack of content expertise in acuteconditions. Finally, adoption of the “common rule” is not evenuniversal throughout the federal system.
Because of these issues, a centralized process for acute illnessand trauma research would be useful. There is the precedent ofthe National Cancer Institute, which developed a centralinstitutional review board to standardize study processes,develop consent approaches, and provide support for localinstitutional review boards. A centralized institutional reviewboard, potentially at the NIH, with domain expertise in acuteillness and injury and the difficulties inherent in emergencyresearch could be very beneficial. The development of acommon consent template for an institutional review boardreview that explained issues in the context of acute researchwould be valuable. These consent templates could discussapproaches that are sensitive to the needs of acute care research.These consent approaches include verbal, short form, waived,and exception approaches.4
Many current grant review panels and NIH study sectionslack content expertise on consent and enrollment issues in acute
illness and trauma research. For example, there is no guidance to
540 Annals of Emergency Medicine
study sections on how to review human subjects protection inacute care research. With the potential for a lack of contentexpertise in acute care research, there is the potential forsystematic bias against acute care on study sections. A solutioncould be the development of a transinstitute review body foracute care research, which could provide guidance to studysections on the suitability of consent and enrollment plans forproposed clinical research.
Given the variation in the timing and care settings of acutecare research, larger study populations may be required toadequately test interventions. Many currently used researchprotocols greatly limit the number of eligible patients, can becomplicated to implement, and are relatively insensitive to thetiming of interventions. One practical solution would be theimplementation of large simple trials. These pragmatic trialscould capture a large patient population and be designed forquickly implemented interventions using streamlined protocolsand broader inclusion criteria.
Larger sample sizes will be required because of diagnosticuncertainty as interventions are targeted earlier after injury.Understanding of the longitudinal course of injury progression(from seconds to minutes to hours) is extremely limited foracute injury. Thus, earlier patient phenotyping based on themechanism of disease as the basis for the intervention will berequired. In addition to traditional approaches of mechanism ofinjury, organ system level, and physiologic characterizations, thephenotyping approach will probably require molecularcharacterization of genomic, proteomic, and metabolomicelements. Statistical approaches that address misclassificationmay also be important because trials that enroll or observepatients earlier and earlier will have more uncertainty about theaccuracy of the diagnosis. These challenges exist particularly inout-of-hospital settings, especially in rural areas. Interventionalstudies in these settings and other austere environments may beparticularly relevant to combat and remote locations.
Most current studies, including analyses of administrativedatabases, prospective observational studies, and clinical trials,do not have a good mechanism to link to what happens topatients either before or after they leave the hospital. Integrationof out-of-hospital data to ED and hospital data is needed tobetter address the full time spectrum of emergency and traumaresearch. Federal privacy rules seriously impede the ability tolink EMS and hospital data. Therefore, the ability to followindividual patients through the key units of service during anepisode of care postinjury is limited.3 A potential solutionwould be the development of a universal unique patientidentifier to link information from the National HighwayTraffic Safety Administration and other (National EmergencyMedical Services Information System, National EmergencyDepartment Sample, National Inpatient Sample) databases.
Template for Acute Injury Clinical ResearchA standardized approach would be valuable to the
appropriate design, review, and conduct of acute injury research.
A proposed template is presented in Table 1. This template
Volume , . : November
Cairns et al NIH Roundtable on Emergency Trauma Research
would address key issues in patient identification, consent andenrollment, disease characterization, specialized care settings,special populations, comorbidities, and study methodology,including the following:● Acute injury patients need to have a firm characterization of
the timing and severity of the injury. This will require rapiddiagnostic strategies. In addition, this characterization needsto be monitored over time, with bedside monitoring ofdisease and effect of therapies.
● Enrollment strategies must be rapid; thus, consentapproaches need to be time sensitive and apply to broadpopulations.
● Study protocols need to incorporate all appropriate caresettings, including EMS, ED, operating room, ICU, andrehabilitation facilities, thus requiring interdisciplinary and
Table 1. Template for clinical research in acute trauma.
Component Aspects
Patient identification Characterization of injury acuity over timeCharacterization of severity of injury over time
Consent andenrollment
Rapid strategies for enrollmentTime-sensitive consent approachesAll-inclusive populations
Acute injuryphenotype
Rapid, definitive diagnostic strategiesTime-dependent acuity of injuryBedside monitoring of disease and effect of
Interdisciplinary and specialty collaborationSpecial populations All-inclusive populations, including underserve
underrepresented groupsPediatrics and geriatric populations
Comorbidities General system injuryCentral nervous system injuryCombined general system and central nervous
system injuryPreexisting and associated conditions
Study designconsiderations
Inclusion population broader than targetedpopulation
Adaptive trial designsHeightened need for preliminary studiesAppropriate follow-up evaluations
EFIC, Exception from informed consent; CSR, Center for Scientific Review.
specialty collaboration.
Volume , . : November
● Study protocols need to incorporate the needs of and injuryimpact on special patient populations, including pediatricand geriatric patients.
● Study protocols need to account for comorbidities, includingconcomitant general system trauma, central nervous systeminjury, intoxication, substance abuse, and preexistingconditions.
● Studies will need to incorporate optimal study designs,including enhanced inclusion strategies, adaptive trialdesigns, individualized treatment regimens, and appropriatefollow-up and outcome evaluations.Coordination of acute injury clinical research could facilitate
the review of research proposals, as well as integrate effortsacross multiple institutes and medical disciplines. A specificcentralized source listing currently funded projects in acute
Solutions
Development of rapid diagnosticsCharacterization and definition of acute injury phenotypeNIH help on HIPAA consent and databasesCentral institutional review board for NIH emergency and trauma
research (transinstitute)Transfederal agency central institutional review boardFederal support of institutional review board reciprocityClinical Translational Science Institutes support of emergency
trauma consent and enrollment initiativesDevelopment of EFIC, waiver, verbal, short-form consent
approachesNIH specialized qualified reviewers to CSR and instituteNeed to rapidly phenotype injuryDiagnostic approaches consistent with injury mechanismsTiming and severity markers of injury and host responseBiomarkers (genomic, proteomic, and metabalomic)Develop meaningful surrogates for therapeutic outcomesTheragnostics
, EMS system integration and standardizationMust assure specialty integration
Specialized geriatric settingsRecognize aging vs disease vs trauma effectsPediatric settingsAccount for ontogeny and aging effectsRecognize complexities in predictive modelingRecognize premorbid function
Standardization of careProtocolized careHarmonizationOutcomes database linkageRequest for information better acute injury trial designNIH requires acute injury study templateAcute researcher input into review for directed programsClinical Translational Science Institutes wiki for social
networking
, ICU
d and
injury, investigators with acute care expertise, and guidance on
Annals of Emergency Medicine 541
NIH Roundtable on Emergency Trauma Research Cairns et al
the conduct of acute injury clinical trials could serve as avaluable resource for investigators, reviewers, andadministrators. This NIH source could complement the effortsof the Emergency Care Coordination Center to develop a searchengine for emergency-related grant opportunities acrossmultiple agencies. As mentioned, many current grant reviewpanels and NIH study sections lack content expertise on issuesof acute illness and injury research. The further development ofspecial emphasis panels would ensure that sufficient acuteillness/injury content expertise is brought to the review of acuteinjury research.
The Clinical Translational Science Institutes (awardedthrough the Clinical Translational Science Institutes program)may be another mechanism to facilitate the conduct of acuteillness and injury research.9 Traditional NIH-fundedinstitutional clinical research programs such as General ClinicalResearch Programs have focused on protocols in relatively stablepatients in a set research location within the hospital. Researchin acute injury and illness will require specialized protocols indifferent hospital settings, including the ED, ICU, andoperating rooms. Thus, acute phase research has beenhistorically underrepresented at the General Clinical ResearchCenter and even Clinical Translational Science Institutes level.Further guidance and support from NIH will be required toenable acute injury research across the Clinical TranslationalScience Institutes.9
Specific Example Conditions: Post-Injury HemorrhagicShock and Central Nervous System Injury
General and central nervous system trauma is a leading causeof morbidity and mortality, yet successful clinical interventionaltrials remain elusive.10 Yet there is evidence that therapeuticinterventions are beneficial. For example, the wide variations in
Table 2. Integrating clinical research approaches: example of p
Approach Immediate Opportunity
Mechanistic What is the time course and sequence of events thatacute and delayed organ failure during and afterhemorrhagic shock?
How can we characterize the level of severity of the inresponse over time?
What is the effect of sex and developmental status oinjury response?
What causal mechanisms are best therapeutic targetsTranslational What are the optimal methods for maintaining tissue
delivery during life-threatening hemorrhage?What bedside diagnostic and monitoring tools can be
developed or used to guide better goal-directedresuscitation?
Clinical What novel, time-sensitive therapies or combination otherapies improves both acute and long-term outcopatients with life-threatening hemorrhage?
What novel, bedside diagnostic and monitoring technican guide the selection, dosing, and duration of theto improve outcome in life-threatening hemorrhage?
acute central nervous system injury outcomes (mortality from
542 Annals of Emergency Medicine
traumatic brain injury varies from 18% to 65% by center)suggest that there are therapeutic approaches that are beneficialand, conversely, those that are potentially detrimental.
There remain fundamental challenges to the design andconduct of clinical trials in acute injury. There is a lack ofadequate characterization of premorbid (preinjury) healthstatus. Thus, the outcomes of injury trials should be designed ina way that is individualized. As an example, a head-injured 30-year-old may need to return to work to achieve an excellentoutcome; however, a head-injured 90-year-old may need to beable feed himself to achieve an excellent outcome.
In addition, further characterization of the severity of injurymay be required. For example, frontal contusions represent aspecific disease within central nervous system injury in whichthe current research does not provide adequate information tophenotype. Additional measures such as cognitive testing maybe needed to better define both disease severity and outcome ofinterventions.
It is also important to tailor acute phenotypes to mechanismsof the injury response, which will require the integration ofmechanistic translational and clinical approaches for bothgeneral trauma (Table 2) and central nervous system injury(Table 3). This will require more attention to clinical correlatesin experimental studies. For example, animal models thatinvolve pretreatment with an agent inherently do not translatewell to human trials with a 12- to 24-hour treatment window.
Use of an integrated generalized template for trial design(Table 1) in both general trauma and central nervous systeminjury could help in the evaluation and design of trials. Thisintegrated approach could assess for clarity in entry criteria andalso emphasize need for a mechanistic approach to phenotypingand the response to interventions. Specific requests for proposalsfor better acute injury/illness trial designs would facilitate the
jury hemorrhagic shock.
Long-Term Objective
to Create animal models of hemorrhagic shock that incorporatethese time and severity relationships to better reflect humanclinical phenotypes, including very young, older, and thosewith comorbid conditions.
Develop molecular patterns of the injury response specific tothe severity and duration of the injury.
n Move the tools that achieve immediate opportunities to thebedside—must be usable in initial clinical care, minutes tohours deployable.
in
s
Create new national standards and guidelines to ease consentand enrollment for those participating in injury research,recognizing that current waiver/exemption criteria do notfully address the demands and needs.
Train the future science leaders for these efforts.
ostin
lead
jury
n the
?oxyge
fmes
quesrapie
evolution and incorporation of this template. Such proposals
Volume , . : November
Cairns et al NIH Roundtable on Emergency Trauma Research
could explore new clinical trial simulation testing and theincorporation of Bayesian methods in which a learning phase isfollowed by a confirmatory phase. Furthermore, the NIHshould advocate for a team approach to acute central nervoussystem injury clinical research in which emergency medicine,trauma surgery, neurosurgery, critical care, and neurologicrehabilitation are all involved.
Observational Patient-Oriented ResearchBecause clinical trials generally have strict inclusion
criteria and consent requirements, they do not yield unbiasedpopulations of patients with diseases of interest. Thus,important advances of knowledge in the field have comefrom observational studies of broad databases. For example,the Traumatic Coma Databank, which has data that aremore than 30 years old, has provided most of the human-level data about the degree to which hypoxia andhypotension cause secondary injury and degrade clinical
Table 3. Integrating clinical research approaches: example of c
Approach Immediate Opportunity
Mechanistic What are the time course and sequence of molecularevents that cause neuronal dysfunction and death acentral nervous system trauma?
How do these change according to injury severity,developmental status, sex, comorbid conditions, anpolytrauma?
What causal mechanisms are best therapeutic targets
Translational What therapies improve neurologic outcome after traumbrain and spinal cord injury, and what are the optimtiming, dose, and duration of these therapies, eitheapplied individually or in combination?
What bedside diagnostic and monitoring tools can bedeveloped or used to guide better guide therapy?
Clinical What novel, time-sensitive therapies or combinations otherapies improve outcome in patients with traumatbrain or spinal cord injury?
What novel, bedside diagnostic and monitoring techniqcan guide the selection, dosing, and duration oftherapies to improve outcome in central nervous sysinjury?
SEP, Special Emphasis Panel.
outcomes in traumatic brain injury.11 In addition, guideline
Volume , . : November
adherence and implementation research are other areas thatbenefit from observational research approaches. Furthermore,there is a need to update the epidemiology and progressionof acute injury conditions. This observational approach toacute injury phenotyping could also complement efforts forprospective acute injury trials.
Outcome measures for acute injury are needed that capitalizeon common data elements available in observational databases.Registries targeted to critical research questions in acute injurycould be helpful. In addition, federally funded clinical trials ininjury should be encouraged to collect common data elementsin usable form. Incorporation of standardized common dataelements would allow for pseudoexperimental designs andprospective cohort studies to address important questions notanswerable by clinical trials. Such studies may be able to providemeaningful information based on the differences in outcomesacross care centers.
There is need for the development of a formal traumatic
l nervous system injury.
Solutions
Create methods and technology (ie, bioassay, biomarker, imaging)to serially measure the time course and sequence of causalmolecular events in animal models of central nervous systeminjury.
Compare the time course and sequence of causal molecularevents in different age groups and sexes, and in models withcomorbid conditions, including polytrauma.
Recruit grant reviewers with understanding the timing andsequence of emergency and acute injury.
Develop SEP to focus on time course, sequence, andmeasurement of mechanisms in acute injury
Requests for applications to address need for biomarker for timecourse and sequence of injury mechanisms
Move the tools that achieve immediate opportunities to thebedside—usable in initial clinical care and deployable withinminutes to hours.
Marker for injury mechanism, timing, and severity.Marker for interventional effect (response to therapy)Develop “-omic” approaches to phenotyping over time.Facilitate collaborations with other fields (engineering, analytic
sciences)Collaborative Activities to Promote Translational Research (NINDS)
supplemental grants for collaborationsDevelopment of a preclinical network consortiumNeed to identify mechanism of injury (or surrogate).Need to monitor mechanistic proxy of intervention.Create new national standards and guidelines to ease consent
and enrollment for those participating in injury research,recognizing that current waiver/exemption criteria do not fullyaddress the demands and needs.
Encourage T2 researchCommunity partner engagement in Clinical Translational Science
InstitutesMultidisciplinary teamwork approach to clinical researchEffective clinical research training and mentor programs
entra
fter
d
?
aticalr
fic
ues
tem
research infrastructure. This infrastructure could be similar to
Annals of Emergency Medicine 543
NIH Roundtable on Emergency Trauma Research Cairns et al
the one built through National Cancer Institute for the “war oncancer.” This would allow for a “war on trauma,” withattendant emphasis on advancing discovery and improvinghealth. A template for this infrastructure could be theResuscitation Outcomes Consortium epistry, which requiredsubstantial training and standardization across study sites.Analysis of standardized data enabled the discovery of 100%magnitude differences in resuscitation survival amongResuscitation Outcomes Consortium study sites.11 Althoughthe epistry project has been a highly useful aspect of theResuscitation Outcomes Consortium, it was not initially fundedthrough conventional NIH means.
Linkage of patient-oriented outcomes to mechanisms of injuryprogression would be integral to better understanding the timingand sequences of the injury response. National Cancer Instituteexpertise in database development would be beneficial to thedevelopment of an acute injury data set. In addition, it may beuseful to try to match common data elements across NIH andother federal agencies. It would also be important for EMS tostandardize and integrate data elements into this larger data set.
The development of a standardized acute injury outcomesdatabase could lead to better population-based perspective oninjury. Embedding more detailed data on injury conditions ofinterest (head trauma) within the design of a nationallyrepresentative survey such as the National Hospital AmbulatoryMedical Care Survey would have substantial appeal. Also it may behelpful to intersect with national efforts to standardize electronicmedical records and clinical billing. Finally, data sharing and theproduction of publicly available data sets is currently just a checkbox in scientific review of clinical trials, rather than a condition foraward. The NIH is in a unique position to advance the cause ofcollecting standardized, time-stamped information in clinical injuryresearch.
The goals for the development, design, and analysis forobservational studies in acute injury are summarized inFigure 1.
Basic and Translational Research in Emergency TraumaFundamental questions about the timing and sequence of events
after acute injury research remain, such as, When does a neurondie? The limitations of past human clinical trials in central nervoussystem injury and ischemia are rooted not only in the methods thathave been used in the trials themselves but also in the basic scienceresearch questions that contributed to their design. Developing atimeline of events after injury at the organ, cellular, genomic, andproteomic levels would lead to more targeted treatments toimprove human outcomes. Even current animal data are limitedwith respect to these questions because most measurements arecross-sectional (ie, animal is killed at a given time and histologicstudies are then performed). This approach fundamentally limitsthe ability to understand the timing and sequence of events afterinjury and thus translate understanding to the dynamic processesthat occur in human injury.
Translational research should be bidirectional between
animal and clinical studies. Both successful and failed clinical
544 Annals of Emergency Medicine
trials should lead to more experimental studies. Conversely,experimental studies should pay more attention to optimaldosing, duration of treatment, and the opportunity window fortreatments with convincing signals of efficacy. As an example,thrombolytic trials in stroke were effective in animal models butfailed in early studies in humans until the correct time windowand dosing were more clearly defined.
Thus, the identification of appropriate surrogate markers fortrauma outcomes should be pursued. Ideally, a treatment mustbe effective in modifying a specific surrogate marker before adefinitive clinical trial is performed. Adaptive trial designsshould be used to identify the optimal timing and dosing ofinterventions after acute injury.
It will be important to better understand the timing andsequence of events in neuronal ischemia, reperfusion, and death.New biomarkers including laboratory tests, imaging, and thelike need to be developed; ideally, these would measure not onlycellular death but also cells at risk of or in the process of dying.These newly identified biomarkers could be used to define acuteinjury phenotype over time and the response to interventions.
A dedicated NIH review system, potentially study section orspecial emphasis panel, for experimental studies on the timingand sequence of events after injury trauma could supportrequests for applications for studies to better delineate thelongitudinal mechanism of cellular injury during acute injurystates in both experimental studies and clinical settings.
Translational research that “results to everyday clinicalpractice and health decisionmaking” is of great importance toemergency medicine and acute injury care.12,13 As an analogy toinjury, the field of sepsis has advanced by using genomic,proteomic, and metabolomic methods in the acute phase.Examining temporal changes in these parameters, with multiple
Figure 1. Observational research priorities in emergencytrauma.
changes over time, can be an analytical challenge. Bringing
Volume , . : November
Cairns et al NIH Roundtable on Emergency Trauma Research
together scientists from different disciplines could helpaccelerate research in the field. The NIH has formal programs,such as the NINDS Collaborative Activities to PromoteTranslational Research program or the National Institute ofGeneral Medical Sciences Large Scale Collaborative ResearchConsortia (Glue Grants), that encourage collaboration betweenscientists from the preclinical and clinical realms. The extensionof such translational research programs with an emphasis onbuilding collaborations with emergency medicine, trauma, andacute injury researchers could create more importantconnections within and between institutions to acceleratediscovery in acute injury.
To facilitate these translational research endeavors, methods forreal time, noninvasive cellular injury monitoring (both in humansand animal models) are needed to better investigate the sequence,timing, and mechanism of injury and to evaluate whethertreatments are leading to a cessation or reversal of cellular injury.These prioritized goals for translational research in emergency
Figure 2. Strategies for better translational research inemergency trauma.
trauma are summarized in Figure 2.
Volume , . : November
TRAINING TO ADDRESS CRITICAL NEEDS INEMERGENCY TRAUMA RESEARCH
Training was identified as a key but surmountable obstacle toadequate emergency care research. The roundtable participantsbelieved the NIH has an opportunity to advance emergency careresearch through targeted training programs for evolvingemergency researchers, support of mentors, and systematicdevelopment of a pipeline of future investigators.
The advancement of emergency care research would befacilitated by having an NIH clearinghouse of trainingopportunities and by expansion of stage-specific trainingmodels. The priorities for training initiatives for emergencytrauma research are summarized in Figure 3.
Given the structure of the NIH and the varied interests ofthe different federal funding agencies, emergency care andtrauma researchers may straddle different institutes, centers, oragencies. The various agencies that participate in conduct andtraining of emergency care research may not be known to theindividual investigator, which leads to investigators expendingenergy and resources with little yield because of “shoppingaround” for appropriate funding sources. This could serve as adeterrent to the pursuit of emergency care research.
Currently, there is no single source from which interestedemergency care investigators may gain information abouttraining and research funding, making it difficult to identifyopportunities for emergency care research. A central federal
Clearinghouse of Training Opportunities
Central repository of training opportunities in acute injury
Centralized access to acute injury resources at theEmergency Care Coordination Center, NIH, NationalCenter for Research Resources
Expansion of Stage-Specific Training Models
Debt forgiveness programs
K-12 national (transinstitute)
R25 (NINDS model)
Training Supplements
Clinical Translational Science Institutes supplementaltraining support for acute injury investigators
Emergency and acute injury network support for training
Investigator Development
Mentorship development (K24 model)
Clinical trialist (boot camp)
Figure 3. Training priorities for emergency traumaresearch.
clearinghouse could provide information on training options
Annals of Emergency Medicine 545
NIH Roundtable on Emergency Trauma Research Cairns et al
and funding programs available to emergency care researchersacross different agencies. This clearinghouse may provide much-needed information about available training paths and fundingmechanisms from different agencies, strengths and weaknessesof these paths, and access to national experts and mentors;ultimately, the clearinghouse may facilitate the improvement ofemergency care services in the United States through research.In the meantime, the NIH may also coordinate intra- andinterinstitute training of emergency care investigators.Specifically, the NIH may:● improve transparency of training and funding opportunities
for emergency care research across the various institutes andcenters at the NIH; and
● provide an incentive for pursuit of emergency care researchby expanding loan repayment programs and providingtransparency in the process of loan repayment awards.There is no single dedicated source of information for NIH
training and funding opportunities in emergency care research.Emergency care research traverses various institutes and centersat the NIH. This is a necessity of the multidisciplinary nature ofinjury and emergency care research. Thus, continuedtransinstitute funding and participation in emergency research iswarranted. However, the individual investigator may havedifficulties in finding the appropriate institute and mechanismfor funding. Further, investigators may not be aware of fundedand ongoing projects, given the current limitations of existingdatabases. Improved coordination across NIH institutes andcenters would allow better access to information on training andfunding opportunities at the NIH. Potentially, such an officecould be set up in the Emergency Care Coordination Center orthe National Center for Research Resources.
According to the American Medical Association, the averageeducational debt of medical students graduating in 2007 wasabout $140,000. Medical education debt and medical schooltuitions are outpacing inflation rates. Roundtable participantssuggested that current debt levels at graduation may serve as adisincentive for pursuit of an academic career. Although theNIH does currently have a loan repayment program, its currentstructure could be improved in a manner that rewards younginvestigators who are committed to a research career. Forexample, no feedback is currently provided by the NIH after acandidate is not selected for a loan repayment program award.Although candidates are encouraged to reapply, the current lackof transparency likely discourages some applicants. Further, itwas suggested that the period of eligibility for the loanrepayment program could be extended such that a candidatewho may not have appeared committed to a career in injuryresearch but subsequently becomes a successful researcher couldstill seek loan repayment program funding.
A shortage of emergency care researchers at every level ofexpertise was acknowledged at the roundtable. In contrast, asizeable number of medical students pursue fields related toacute injuries and emergency care such as emergency medicine,
trauma surgery, and critical care. A large proportion of these
546 Annals of Emergency Medicine
students and residents (�25%) are interested in academiccareers.1 Recruitment of these potential academicians toresearch would address the relative shortage of emergency careresearchers. To this end, it was recommended that the NIHtarget training models to the different levels of acute injuryresearch expertise through the following realms:● support for a national, transinstitute K12● program support for mentorship development● network development/supplemental support
Many impediments to successful attainment of a careerdevelopment grant (K award) from the NIH were discussed.The 2- to 3-year period immediately after residency andfellowship training was identified as a critical period duringwhich burgeoning investigators may become disillusioned withthe process of sorting out where and how to receive funding foremergency care research from the NIH. There are successfulexamples of programs targeted to investigators during thiscrucial phase. The American Academy of Neurology offersannual career development awards that protects juniorinvestigators for up to 2 years. The reported success rate inachieving NIH K awards from these investigators is up to90%.14 Thus, if junior investigators with an interest inemergency care research are provided with protected time earlyin their career, this could translate into significant advances inthe field of emergency care. Institutions with well-developedresearch programs are often able to offer the K12 grant to juniorinvestigators.
The K12 is an Institutional Research and Academic CareerDevelopment Award that is awarded to educational institutionsor professional organizations to support career developmentexperiences for clinicians leading to research independence.K12s are often funded for 2 to 3 years, with the expectation thatawardees will seek career development awards from the NIH atthe end of the K12 period. Junior researchers at non-K12institutions do not have an opportunity for this level of support.
The National Institute of Child Health and HumanDevelopment offers a national K12 program for pediatriccritical care investigators. In the more than 4 years since itsinception, this program has funded 16 pediatric critical carescholars throughout the United States. Funding is provided for2 to 3 years, and scholars are expected to obtain NIH fundingor continued support from their department in subsequentyears. A national advisory committee of senior investigatorsparticipates in the selection and development of scholars.Although not a national K12 program, the National Heart,Lung, and Blood Institute is developing a program of K12programs across the country, with a requirement formultidisciplinary proposals, which is often necessary inemergency care research. The Emergency Medicine Foundationissues annual career development awards but is only able tosupport 1 scholar per year. In 2009, 26 applications for theEmergency Medicine Foundation awards were received,15 whichsuggests that there is ample interest in academic careers among
junior emergency care investigators. Thus, the NIH may foster
Volume , . : November
Cairns et al NIH Roundtable on Emergency Trauma Research
this interest through provision of a dedicated national K12award program and transparency of information about fundingopportunities for emergency care researchers.
The limited number of emergency care researchers who arequalified to serve as mentors for junior investigators wasrecognized by roundtable participants. In addition todeveloping a pipeline of future investigators, providingintermediate-stage investigators with the skills to becomesuccessful mentors was endorsed. The NIH may support thisgoal by targeting existing funding mechanisms for mentorshipdevelopment toward emergency care researchers. For example,the K24 Mid-Career Investigator Award in Patient-OrientedResearch is designed to provide support for clinicians and allowthem protected time to devote to patient-oriented research andto act as mentors for beginning clinical investigators. The R25grant award mechanism is another instrument through whichthe NIH may support the training of new emergency careresearchers. The Research Education Grant (R25) is awarded toinstitutions/organizations that propose to support thedevelopment or implementation of programs that relate toeducation and training of clinical residents and fellows to fostercareers as clinician-scientists. Eligible principal investigatorswould have a track record of successful mentoring and NIHresearch funding. Support for the principal investigator andother faculty participating in the research education project isnot allowable under the R25, although there is no limit to thenumber of potential participants. The T32 and other mentoredtraining grants also do not provide salary support for mentors.This puts potential mentors in a position whereby they areencouraged to assume much responsibility but are offered littlein the way of tangible rewards in return. Given the financialconstraints of academic departments and the pressure to balancepatient care and academic productivity, it was recommendedthat the NIH consider rewarding mentors through salarysupport for the time they commit to mentoring. Finally, not allNIH institutes or centers currently offer the K24 and R25. Inparticular, no institute offers dedicated programs for emergencycare researchers. These mentorship development awards are assuch subject to the issues relating to transparency and lack of acentral source of information described earlier.
The development of a transinstitute, multidisciplinarynetwork dedicated to emergency care research was proposed asan additional means by which the NIH may facilitate thetraining of emergency researchers. The Resuscitation OutcomesConsortium is an example of such a network funded bymultiple federal agencies in the United States and Canada.16
The Resuscitation Outcomes Consortium is designed to provideinfrastructure and project support for clinical trials and otheroutcome-oriented research in the areas of cardiopulmonaryarrest and severe traumatic injury. Yet the ResuscitationOutcomes Consortium investigations are particularly gearedtoward EMS research in cardiac and trauma populations and assuch do not fully capture the spectrum of emergency injury
research. Supplemental support for the Resuscitation Outcomes
Volume , . : November
Consortium is provided by the funding agencies for juniorinvestigators to receive training in clinical or fundamentalresearch and to conduct clinical or bench research inresuscitation. Other evolving federally funded emergency careresearch networks that could be leveraged for the developmentof junior investigators in emergency trauma include thePediatric Emergency Care Applied Research Network,17 theNeurological Emergency Treatment Trials program,18 and theUS Critical Illness and Injury Trials Group.19
CONCLUSIONSEnhancing the research base supporting the care of trauma
emergencies will require progress in specific mechanistic,translational, and clinical domains; effective collaboration ofacademic investigators across traditional clinical and scientificboundaries; federal support of research in high-priority areas;and overcoming limitations in available infrastructure, researchtraining, and access to patient populations.
Supervising editors: Donald M. Yealy, MD; Michael L.Callaham, MD
Funding and support: By Annals policy, all authors are requiredto disclose any and all commercial, financial, and otherrelationships in any way related to the subject of this articlethat might create any potential conflict of interest. See theManuscript Submission Agreement in this issue for examplesof specific conflicts covered by this statement. Funding for theRoundtable on Emergency Trauma Research was provided bythe NIH.
Publication dates: Received for publication February 17, 2010.Revision received May 8, 2010. Accepted for publication May11, 2010.
Reprints not available from the authors.
Address for correspondence: Charles B. Cairns, MD,Department of Emergency Medicine, University of NorthCarolina, 170 Manning Dr, Chapel Hill, NC 27599; 919-843-3045, fax 919-966-3049; E-mail [email protected].
REFERENCES1. Institute of Medicine of the National Academies. Hospital-Based
Emergency Care: At the Breaking Point. Washington, DC: NationalAcademies Press; 2007.
2. Nathens AB, Rivara FP, Jurkovich GJ, et al. Management of theinjured patient: identification of research topics for systematicreview using the Delphi technique. J Trauma. 2003;54:595-601.
3. Courtney DM, Neumar RW, Venkatesh AK, et al. Uniquecharacteristics of emergency care research: scope, populations,and infrastructure. Acad Emerg Med. 2009;16:990-994.
4. Glickman SW, Anstrom K, Li L, et al. Challenges in enrollment ofminority, pediatric and geriatric patients in emergency and acutecare clinical research. Ann Emerg Med. 2008;51:775-780.
5. Sun LH, Glod M. DC Metro Red Line crash. Washington Post.June 23, 2009. Available at: http://www.washingtonpost.com/wp-dyn/content/linkset/2009/06/23/LI2009062301806.html?sid�ST2009062603770. Accessed October 25, 2009.
6. NIH Task Force on Research in the Emergency Setting: request
for information: soliciting input on current needs in emergency
NIH Roundtable on Emergency Trauma Research Cairns et al
medicine research. Available at: http://www.grants.nih.gov/grants/guide/notice-files/NOT-NS-08-021.html. Accessed October 25,2009. NOT-NS-08-021.
7. Pitts SR, Niska RW, Xu J, et al. National Hospital AmbulatoryMedical Care Survey: 2006 emergency department summary.August 6, 2008. No. 7. Available at: http://www.cdc.gov/nchs/data/nhsr/nhsr007.pdf. Accessed November 10, 2008.
8. Institute of Medicine of the National Academy of Sciences.Emergency Medical Services at the Crossroads. Washington, DC:National Academies Press; 2007.
9. Hollander JE, Gaulton GN, Courtney DM, et al. Facilitatingemergency care research networks: integration into the CTSAinfrastructure. Acad Emerg Med. 2009;16:1005-1009.
11. Guidelines for the management of severe traumatic brain injury.J Neurotrauma. 2007;24(suppl 1):S1-106.
12. Nichol G, Thomas E, Callaway CW, et al; Resuscitation OutcomesConsortium Investigators. Regional variation in out-of-hospital cardiacarrest incidence and outcome. JAMA. 2008;300:1423-1431.
13. Woolf SH. The meaning of translational research and why itmatters. JAMA. 2008;299:211-213.
14. American Academy of Neurology. Clinician Scientist DevelopmentAwards. American Academy of Neurology Web site. Available at:http://www.aan.com/go/foundation/research/development.Accessed October 25, 2009.
15. Emergency Medicine Foundation. Career Development GrantAwards. Available at: http://www.emfoundation.org/WorkArea/linkit.aspx?LinkIdentifier�id&ItemID�414. Accessed October 25,2009.
16. Morley P. Steady as a ROC: the Resuscitation OutcomesConsortium. Resuscitation. 2008;78:105-106.
17. Pediatric Emergency Care Applied Research Network. ThePediatric Emergency Care Applied Research Network (PECARN):rationale, development, and first steps. Acad Emerg Med. 2003;10:661-668.
18. Neurological Emergencies Treatment Trials (NETT) Network.Neurological Emergencies Treatment Trials Web site. Available at:http://www.nett.umich.edu/nett/welcome. Accessed November10, 2008.
19. Cobb JP, Cairns CB, Bulger E, et al. The United States CriticalIllness and Injury Trials Group: an introduction. J Trauma. 2009;67(2 suppl):S159-160.
