NIH EMERGENCY MEDICINE RESEARCH ROUNDTABLES/SPECIAL CONTRIBUTION
NIH Roundtable on Opportunities to Advance Research onNeurologic and Psychiatric Emergencies
Gail D’Onofrio, MD, MS, Edward Jauch, MD, MS, Andrew Jagoda, MD, Michael H. Allen, MD,Deirdre Anglin, MD, MPH, William G. Barsan, MD, Rachel P. Berger, MD, MPH, Bentley J. Bobrow, MD,
Edwin D. Boudreaux, PhD, Cheryl Bushnell, MD, MHS, Yu-Feng Chan, MD, Glenn Currier, MD, Susan Eggly, PhD,Rebecca Ichord, MD, Gregory L. Larkin, MD, MS, MSPH, MA, Daniel Laskowitz, MD, MHS, Robert W. Neumar, MD, PhD,
David E. Newman-Toker, MD, PhD, James Quinn, MD, MS,Katherine Shear, MD, Knox H. Todd, MD, MPH, Douglas Zatzick, MD
Roundtable External Participants and Roundtable Steering Committee and Federal Participants*
Study objective: The Institute of Medicine Committee on the Future of Emergency Care in the United StatesHealth System (2003) identified a need to enhance the research base for emergency care. As a result, aNational Institutes of Health (NIH) Task Force on Research in Emergency Medicine was formed to enhanceNIH support for emergency care research. Members of the NIH Task Force and academic leaders inemergency care participated in 3 Roundtable discussions to prioritize current opportunities for enhancingand conducting emergency care research. We identify key research questions essential to advancing thescience of emergency care and discuss the barriers and strategies to advance research by exploring thecollaboration between NIH and the emergency care community.
Methods: Experts from emergency medicine, neurology, psychiatry, and public health assembled to reviewcritical areas in need of investigation, current gaps in knowledge, barriers, and opportunities. Neurologicemergencies included cerebral resuscitation, pain, stroke, syncope, traumatic brain injury, and pregnancy.Mental health topics included suicide, agitation and delirium, substances, posttraumatic stress, violence,and bereavement.
Results: Presentations and group discussion firmly established the need for translational research to bringbasic science concepts into the clinical arena. A coordinated continuum of the health care system thatensures rapid identification and stabilization and extends through discharge is necessary to maximizeoverall patient outcomes. There is a paucity of well-designed, focused research on diagnostic testing,clinical decisionmaking, and treatments in the emergency setting. Barriers include the limited number ofexperienced researchers in emergency medicine, limited dedicated research funding, and difficulties ofconducting research in chaotic emergency environments stressed by crowding and limited resources.Several themes emerged during the course of the roundtable discussion, including the need fordevelopment of (1) a research infrastructure for the rapid identification, consent, and tracking of researchsubjects that incorporates innovative informatics technologies, essential for future research; (2) diagnosticstrategies and tools necessary to understand key populations and the process of medical decisionmaking,including the investigation of the pathobiology of symptoms and symptom-oriented therapies; (3)collaborative research networks to provide unique opportunities to form partnerships, leverage patientcohorts and clinical and financial resources, and share data; (4) formal research training programs integralfor creating new knowledge and advancing the science and practice of emergency medicine; and (5)recognition that emergency care is part of an integrated system from emergency medical services dispatchto discharge. The NIH Roundtable “Opportunities to Advance Research on Neurological and PsychiatricEmergencies” created a framework to guide future emergency medicine–based research initiatives.
Conclusion: Emergency departments provide the portal of access to the health care system for most patientswith acute neurologic and psychiatric illness. Emergency physicians and colleagues are primed to investigateneurologic and psychiatric emergencies that will directly improve the delivery of care and patient outcomes. [AnnEmerg Med. 2010;56:551-564.]
Volume , . : November Annals of Emergency Medicine 551
NIH Roundtable on Neurologic and Psychiatric Emergencies D’Onofrio
SEE RELATED ARTICLES AND EDITORIALS,P. 522, 538, 565, 568.
INTRODUCTIONThe underlying hypothesis of emergency care research is
that the development of rapid diagnostic tools and earlytherapeutic interventions in acute illness will improvepatient outcomes. Time is of the essence for most neurologic andbehavioral emergencies, many of which are characterized by rapidonset and progression of cerebrovascular injury. Basic science andclinical research have shown that the opportunity for mitigatingpermanent cellular damage is on the order of minutes to hours.Thus, a focus on developing therapies and creating the knowledgeto improve emergency care is crucial.
There are many challenges to developing and implementinga robust research agenda. Foremost, enlargement of anemergency care research workforce is needed. Researchfellowships should be promoted; the emergency medicinecommunity needs to make better use of the National Institutesof Health (NIH) career development “K” mechanism. There arealso very few if any training “T” awards within departments ofemergency medicine. Financial constraints and fragmentedsystems are also challenging for emergency research. In addition,informed consent and privacy considerations are issues for time-sensitive research. Multicenter research networks have provenuseful in circumventing some of these challenges.
Emergency care offers unique access to at-risk populationsand has been shown to be an opportune site for early diagnosisand implementation of prevention strategies (eg, HIV,unhealthy alcohol and drug use, and suicide). The emergencydepartment (ED) offers an opportunity to advance diagnosticsand to develop risk-stratification paradigms, specific acutetherapeutic interventions, prevention strategies, and publicpolicy initiatives.
Goals of This InvestigationOn December 3 to 4, 2008, the NIH hosted a roundtable
entitled “Opportunities to Advance Research on Neurologicaland Psychiatric Emergencies,” designed to create a frameworkfor research by exploring opportunities, gaps, and challenges forresearch in the emergency care setting.
MATERIALS AND METHODSThe planning group for the roundtables convened with
members from NIH staff from the National Institute ofNeurological Disorders and Stroke, the National Institute onMental Health, the Eunice Kennedy Shriver National Instituteon Child Health and Human Development, and the NationalCenter on Research Resources, and extramural emergency careresearch advisors. The planning group developed the agendaand selected participants and presenters according to responsesto an NIH request for information. The planning groupconsidered the request for information responses, the state of
the science, and the burden of each disease and narrowed down
552 Annals of Emergency Medicine
topics through a consensus voting process for focused discussionin the final agenda. The goal of the meeting was to identifyconcrete questions in emergency care research that would fillimportant gaps in the missions and portfolios of NIH institutesthat support neuroscience and mental health.
Cerebral resuscitation is a broad area of investigation withinthe scope of emergency care and includes cerebral resuscitationafter cardiac arrest, ischemic and hemorrhagic stroke, traumaticbrain injury, status epilepticus, central nervous system infection,and noninfectious encephalopathy. This section focuses oncerebral resuscitation after cardiac arrest.
Resuscitation is attempted in approximately 175,000 patientswith out-of-hospital cardiac arrest each year in the UnitedStates, and the overall mortality is 92%.1 In patients who regainspontaneous circulation after out-of-hospital cardiac arrest,mortality is approximately 70%, with two thirds of deathscaused by neurologic injury and a significant number ofsurvivors never regaining normal function.2 The spectrum ofpost–cardiac arrest brain injury ranges from amnestic syndrometo brain death. Therapeutic hypothermia is the onlypost–cardiac arrest intervention demonstrated to improveoutcomes in prospective randomized clinical trials.3-5
Challenges for clinical cardiac arrest research are similar tothose in research for other acute conditions: interventions areunscheduled and occur at multiple locations of care; diverseteams of care providers are involved; therapeutic interventionsare extremely time sensitive, which limits the feasibility ofinformed consent; and there is a need for medical staffinterested and trained in this field of research. NIH hasaddressed some of these problems through establishing networkssuch as the Resuscitation Outcomes Consortium and theNational Institute of Neurological Disorders (NINDS) andStroke–sponsored Neurologic Emergencies Treatment Trial(NETT) Network.
In terms of bench research, a major challenge is that cardiacarrest and post–cardiac arrest syndrome are diseases in whichthe pathophysiology and therapeutic strategies are integrallyrelated among multiple organ systems. Investigation in the fieldcould be optimized by basic science and translational researchthat more globally examine the pathophysiologic mechanisms oftotal body ischemia and reperfusion, including the interactionsamong injured organ systems. However, the scopes of suchprojects make them challenging to review within the focusedexpertise of most existing NIH study sections. In addition, thescope crosses over multiple NIH institutes, making it difficultfor investigators to identify the appropriate institute forfunding. Specific opportunities for research in cerebralresuscitation are highlighted in Figure 1.
Pain is the most common reason for seeking health care and,as a presenting complaint, accounts for up to 78% of visits tothe ED.6,7 Yet pain is a subjective symptom for which noobjective measure exists. Adequate analgesia in EDs is animportant goal of treatment; however, the underuse of
analgesics, termed oligoanalgesia, occurs in a large proportion of
Volume , . : November
bral
D’Onofrio NIH Roundtable on Neurologic and Psychiatric Emergencies
ED patients. Risk factors for oligoanalgesia include the extremesof age, certain clinical conditions (eg, sickle cell disease), andminority patient ethnicity.8 Analgesic treatment decisions areoften influenced by concerns about prescription drug abuse andnot by evidence-based clinical trials.
An increasing amount of research focuses on acute paintreatment in the ED, with an emphasis on randomizedcontrolled trials comparing various treatment strategies for acuteconditions, including renal colic, abdominal pain,musculoskeletal injuries, and migraine. Yet very little researchaddresses differences among patients in terms of pain-relatedgenetic markers, experimental pain thresholds and tolerance,precise descriptions of underlying pain, predicting individualresponses to therapy, and longitudinal follow-up to determineimportant outcomes of treatment. Despite the knownpsychosocial influences on pain and its treatment, only limitedresearch addresses depression, health beliefs, and substanceabuse in relation to pain in the ED. Care providers are knownto be idiosyncratic in their pain assessment and treatmentdecisions, yet little research helps us understand the role ofprovider variability in determining pain outcomes. Specificopportunities for research in pain are highlighted in Figure 2.
Clinical stroke research was one of the first areas to embracecollaboration with emergency care investigators in NIH-fundedstudies. Prior to tissue plasminogen activator (tPA), clinicalstroke research was conducted by neurologists and the time totreatment for interventions did not reflect data from preclinicaltrials. Treatment was given in 24 to 48 hours. The NIH-fundedtPA pilot study testing onset to treatment as early as 90 minutes
Research Opportunities/Gaps
Sequence and time course of molecular events that mediate ne
Assess variability in these mechanisms according to injury seve
Global mechanisms of total body ischemia and reperfusion, in
Mechanisms by which hypothermia induces neuroprotection.
Current options for optimizing hypothermia, including optimindividual dosing, or combination therapies.
Natural brain mechanisms of repair, remodeling, and endogen
Procedures to optimize controlled reperfusion, including contcirculation, mitochondrial inhibition.
The role of neuroprotective pharmacology, including proteasemodulators, inflammation modulators.
The role of brain monitoring (noninvasive techniques) to ind
Coma reversal—diagnosis of reversible coma and developmen
Prognostication of futility—especially accuracy with effective
Figure 1. Cere
required adapting methods for identification and early
Volume , . : November
assessment of patients in the ED and utilizing the expertise ofemergency physician investigators.
Today, the National Institute of Neurological Disorders andStroke funds the Neurologic Emergencies Treatment TrialNetwork, a clinical research network for phase III trials testinginterventions for neurologic emergencies in the out-of-hospitalor ED setting.9,10,11 A similar network exists through NationalHeart, Lung, and Blood Institute’s (NHLBI) ResuscitationOutcomes Consortium. Both the Neurologic EmergenciesTreatment Trial Network and Resuscitation OutcomesConsortium focus on conditions requiring emergencytreatment, even though these conditions often cross specialtyboundaries. These networks could become models forintegrating disciplines across the spectrum of emergency care (asopposed to disease/discipline centered networks). Some of thetrials that Neurologic Emergencies Treatment Trial Network iscurrently undertaking or proposing include evaluations ofalbumin in acute stroke, anticonvulsant medications before EDarrival, progesterone for traumatic brain injury, acuteclopidogrel after transient ischemic attack, and hypothermia inspinal cord injury.
All trials of neuroprotectants to date have been unsuccessful,although there may be design issues at fault rather than futilityof the intervention. Similarly, for subarachnoid hemorrhage andintracerebral hemorrhage, there are no proven strategies foremergency management and no proven treatment, and theusefulness of surgery or procoagulants is unclear. Specificopportunities for research in emergency stroke are highlighted
injury during global brain ischemia and reperfusion.
reperfusion conditions, age, and sex.
ing interaction of injured organ systems after cardiac arrest.
NIH Roundtable on Neurologic and Psychiatric Emergencies D’Onofrio
Approximately 5 million patients visit US EDs each year,complaining of syncope, weakness, dizziness, or vertigo. Patientswith these symptom complexes present a diagnostic challengebecause the possible causes are diverse. Roughly 5% to 15% ofthese patients will have a serious underlying disorder, with therest experiencing benign conditions.11,12 Critical misdiagnosesremain frequent, particularly among patients with isolateddizziness and vertigo. For example, diagnoses for patients withposterior circulation strokes presenting with dizziness are missed35% of the time because the patients never receive anappropriate diagnostic evaluation. Determining the mostaccurate clinical predictors and providing physicians real-timeclinical decision support may improve the identification and
Research Opportunities/Gaps
Investigations into the transition of acute to chronic pain. Idebe screened in acute care settings (eg, genetic markers, depress
Development of biomarkers or neuroimaging modalities that
Validation of assessment instruments to measure prescriptionDevelopment of brief interventions to treat prescription drug
Development of best pain treatments (eg, multimodal therapibehaviors that can decrease pain in the ED or lead to improvephysician empathy).
Study of painless presentations of typically painful conditionstolerance.
Interventions to address persistent ethnic disparities in ED pa
Figure 2. Acute
Research Opportunities/Gaps
Improved study design (eg, adaptive trials, use of Bayesian appcombinations) and strategies (eg, earlier enrollment, consent e
Studies into the optimization of physiologic parameters (eg, odelivery, temperature) in all forms of acute central nervous sysinsufficiently studied.
Need for improved techniques or technologies to allow goal-dbe evaluated in clinical trials.
Mechanistic studies of secondary brain injury have identifiedrarely led to any successful clinical therapies.
The modes of failure in translational trials of neuroprotectant
Better clinical trials are needed to deliver on the promise of thof neuroprotection when proven successful (eg, dose optimiza
Effect of development on the response to neurologic must be
Figure 3. Em
prompt treatment of patients with serious conditions while
554 Annals of Emergency Medicine
eliminating unnecessary testing and admission for many ofthose with benign causes.
Numerous studies have identified risk factors to predict high-risk syncope patients who would benefit from hospitaladmission and more extensive evaluation. Research should focuson improving physician decisionmaking and improvedefficiency in the use of tests and hospital admission based onrisk stratification. Although recent work has identified bedsidepredictors of stroke in a subset of high-risk patients withvertigo,13 no well-validated predictors are available for theaverage patient presenting with dizziness. Ample evidencesuggests resources are not being used efficiently to differentiatethose with benign inner ear disorders from those with dangerous
ation of predictors of chronic pain transformation that can
nform about pain and addiction.
abuse risk among patients presenting to the ED with pain.e in this context.
trategies (eg, regional anesthesia), or physician-relatedtcomes (eg, decreased risk of delirium, adverse effects,
silent ischemia) measuring experimental pain thresholds and
atment.
chronic pain.
hes, endpoints based on individuals, evaluation of treatmentptions, controlled postintervention care).38
um values for blood pressure, blood glucose, oxygenpathology are associated with clinical outcomes but remain
ed individual physiologic parameter manipulation that must
erous forms of putative neuroprotection, but these have
rain injury must be identified and addressed.
clinical investment and to optimize the clinical applicationin hypothermia).
r understood, especially in the pediatric population.
ncy stroke.
ntificion).
can i
drugabus
es), sd ou
(eg,
in tre
roacxem
ptimtem
irect
num
s in b
e pretion
bette
strokes. There has been little funding for improving diagnostic
Volume , . : November
dizz
D’Onofrio NIH Roundtable on Neurologic and Psychiatric Emergencies
accuracy and efficiency of clinical decisionmaking in patientswith undifferentiated symptoms despite the potential to makesignificant gains in cost-effectiveness and improved patientoutcomes. For their part, researchers and clinicians need toclearly define vague symptoms to better phenotype patients toensure similarity in study protocols. Finally, technical andlogistic barriers to implementing electronic tracking, alerting,and clinical decision support systems through electronic healthrecords need to be overcome.14 Specific opportunities forresearch in syncope, dizziness, and vertigo are highlighted inFigure 4.
Numerous acute neurologic emergencies could benefit fromthe development of disease-specific biomarkers. Initial diagnosisand management of stroke could be assisted by a biomarker-based diagnostic test if it were widely available, applicable in theout-of-hospital setting, and had a rapid turnaround time. Abiomarker-based diagnostic test to identify patients with mildtraumatic brain injury who are at high risk for intracranialpathology could aid in early triage and management decisionsabout the need for observation, admission, or imaging studiesand to identify patients at risk for neurobehavioral sequelae.15
The ideal biomarker would be released from the target organ(eg, neurons, glia), would readily cross the blood-brain barrier,would be quickly detectable in blood, would have releasepatterns indicative of disease progression, and would be specificto the type of central nervous system injury. Yet identifyingclinically useful biomarkers in acute neurologic disease remainsa formidable challenge because of the complex cell distributions,numerous cell types, and problems overcoming the blood-brainbarrier.16,17
Although biomarker-based diagnostic approaches havedemonstrated a great deal of promise in the early evaluation of
Research Opportunities/Gaps
Invest in research infrastructure to electronically screen healthelectronic screens for orders and results would be useful to aleor enrollment in a clinical trial.
Develop diagnostic decision support systems to improve the eidentifying patients for treatment or enrollment in trials.
Develop “smart” automated patient interviewing systems to cadeveloping clinical prediction rules.
Develop national or regional mechanisms for data sharing acrto facilitate improved diagnosis of neurologic emergencies.
Pair therapeutic trials (eg, for cerebellar stroke) with diagnostidizziness or vertigo) to improve research efficiency.
Fund specific studies to develop, validate, and disseminate clinimaging.
Figure 4. Syncope,
neurologic emergencies, a number of challenges remain,
Volume , . : November
including the absence of clinical consensus on the necessary testcharacteristics for the clinical setting (eg, higher sensitivity overspecificity), the proper design of studies in the absence of acriterion standard, and acceptance of a unitless diagnostic valuerepresenting an integrated biomarker model. Specificopportunities for research in brain biomarkers are highlighted inFigure 5.
Pregnant women are understudied in clinical research, andtheir special health and clinical considerations, especiallyneurologic, in the emergency setting are poorly understood.Pregnancy-related conditions that are likely to be observed in anED involve women who are either less than 20 weeks pregnantor postpartum. All other gestational stages are likely to beobserved in obstetrics triage.
There are few prospective studies of headaches in pregnantwomen, and often the cohorts are too small to studyassociations between headaches and neurologic emergencies. Inaddition, in clinical practice physicians are reluctant to evaluatepregnant women with headaches even with a computedtomographic scan of the brain for fear of fetal radiationexposure.18 Migraine is particularly important because of its linkto preeclampsia, as well as stroke, in young women.19 Yet therelation between migraines and stroke or other conditions thatmay increase the risk for stroke, such as preeclampsia, is still notunderstood. Specific opportunities for research in neurologicemergencies during pregnancy are highlighted in Figure 6.
Increasing rates of ED visits for mental health problems andsuicide attempts are reported in the United States and abroad,20
rendering the ED the de facto option for initial identificationand treatment of suicidal patients.21 Of individuals who attemptsuicide, 1.8% die and 16% repeat the attempt within a year.Thus, the ED is a key site to capture through surveillance and
l-7 (HL-7) data for presenting symptoms; in addition,estigators of patients potentially eligible for an intervention
ncy of health care delivery, at the same time rapidly
e detailed symptoms data directly from patients for use in
Ds on critical diagnoses (eg, stroke, aneurysm, meningitis)
ls (eg, for diagnosis of stroke in patients presenting with
decision rules related to dizzy patients and those requiring
iness, and vertigo.
levert inv
fficie
ptur
oss E
c tria
ical
case-finding the population most at risk of suicide, including
Annals of Emergency Medicine 555
emic
erge
NIH Roundtable on Neurologic and Psychiatric Emergencies D’Onofrio
individuals who are injured, intoxicated, and lack access toroutine care.
New approaches to ED screening for mental disorders areneeded.22 ED suicide registries, based on similar modelstracking cancer or cardiovascular risk, could provide data forsuicide studies, promote standardization of definitions, facilitateefforts to engage ED staff in screening and referring suicidalpatients, identify individuals likely to be most at risk aftertrauma or disasters, detect clusters of suicide attempts, andmonitor the emergence of new overdoses.
Beyond screening for suicide risk, better interventions forED mental health patients are very much needed. For example,computer-elicited feedback to physicians about the suicideideation status of a patient has been shown to increasepsychiatric referrals. Other possibilities include rapid-onsetantidepressants and tailored therapies based on an individualpatient’s genome, transcriptome, and proteome. EDneuroimaging (functional magnetic resonance imaging, positronemission tomography imaging) could be used to identify thestructural and biochemical bases of mental illness.
Current barriers to ED mental health research include lack ofstandardized definitions of suicidal behavior and validated
Research Opportunities/Gaps
Define the role of biomarkers in the out-of-hospital setting fo
Identify whether there is a role for biomarkers in selecting stro
Risk stratification of patients with transient ischemic attack to
Differentiate ischemic from hemorrhagic stroke (although comdeciding whether to administer reperfusion agent).
Establish the infrastructure to enable and facilitate risk stratifi
Use of serial markers to predict inhospital deterioration after s
Identify biomarkers for traumatic brain injury assessment, eva
Develop and use biomarkers as surrogate endpoints for clinica
Figure 5. Biomarkers in isch
Research Opportunities/Gaps
Epidemiology of acute neurologic and psychological events/co
Develop decisionmaking tools to determine when to image prthose with migraine history, cardiovascular risk factors, and in
Biomarkers for migraine and cerebrovascular disease, when va
Noninvasive imaging for pregnant women with neuroultrason
Multicenter registries to clarify association between peripartum
Figure 6. Neurologic em
screening tools, ED staff attitudes toward the mentally ill,
556 Annals of Emergency Medicine
poorly designed outcome measures, concerns about informedconsent and exclusion of highly suicidal patients from trials, fewdiagnostic profiles to inform risk stratification, and dearth offunding for ED-based suicide research. Vulnerable populations,including the seriously mentally ill, immigrants, the homeless,adolescents, and the elderly, are overrepresented among EDsuicidal patients. Novel, ED-based interventions that includethese high-risk populations and are responsive to their needsand preferences for acute care are needed.
Specific opportunities for research in ED suicide screeningand surveillance are highlighted in Figure 7.
Agitation and delirium are common symptoms or syndromesof patients who enter the ED with “altered mental status.”Aggressive medical and security procedures for behavioralproblems have evolved into accepted practices in clinical culturewithout the benefit of science.
Agitation, a change in mental state associated with increasedmotor activity and accompanied by distress or arousal, appearsin many psychiatric and medical conditions, including delirium.Delirium, both hyperactive and hypoactive forms, is a medicalemergency but remains poorly understood. Approximately 10%of elderly ED patients are delirious, and up to 80% of cases are
ging patients at risk for stroke or traumatic brain injury.
atients who are candidates for reperfusion therapy.
tify which patients merit admission.
ed tomographic scans will probably always be necessary for
n of suspected stroke.
e or traumatic brain injury.
on, intervention, and prognosis.
ls.
and traumatic brain injury.
ions in pregnant women.
nt women who present with headaches or migraines, such assed blood pressures.
ed, should also be studied in pregnant women.
phy.
graine and other diseases.
ncies during pregnancy.
r tria
ke p
iden
put
catio
trok
luati
l tria
ndit
egnacrea
lidat
ogra
mi
undetected.23 Identifying the cause of agitation and
Volume , . : November
cree
of a
D’Onofrio NIH Roundtable on Neurologic and Psychiatric Emergencies
differentiating medical from psychiatric causes are key tosuccessful management.
Research in the ED on agitation and delirium is difficult.There are few robust clinical programs appropriately staffed toperform emergency mental health assessments. Consent is aproblem in many emergency conditions but particularly so inthis context, given that many potential subjects are subjectsinvoluntarily and outcomes can be difficult to ascertain.Consequently, studies of agitation and delirium are oftenperformed in other settings and may not be ecologically valid inemergency settings. Challenges in methodology further impederesearch on agitation and delirium: definitions have not beenstandardized, detection is poor, the prevalence and course arenot well described, causes are diverse and underlyingmechanisms are not understood, and measurement scales arenot standardized.
There is no consensus concerning screening or measurementof agitation or even the value of making the diagnosis in theemergency setting.24 Although there is a consensus thattreatment of the underlying condition is most desirable,25 thebiology of agitation deserves further study. Other research needs
Research Opportunities/Gaps
Appropriate ED suicide and mental health screening tools nee
Randomized controlled trials of innovative ED-initiated treatminterventions and strategies that most benefit specific subpopu
Brief, low-cost psychosocial interventions should be further de
Enhance mechanistic understanding of acute mental health prneuropsychiatry and related fields.
Real-world effectiveness trials to examine the utility of ED-bainterventions to reduce suicidal behavior and repeated suicide
Knowledge translation research to convert findings into real-wensure the broad uptake of evidence and best practices for EDother mental health problems.
Figure 7. ED suicide s
Research Opportunities/Gaps
Better assessment of both delirium and agitation to support aptreatment.
Better definition of outcomes, including calming and sedation
Development of nonpharmacologic interventions for agitation
Studies of the safety and efficacy of chemical and mechanical
New treatments for delirium.
Figure 8. Management
include clarity on the roles of mechanical restraints and other
Volume , . : November
cocontributors to agitation and assessments of care approaches,including “de-escalation.” Although delirium rating scales havebeen well validated, the effect of these tools on physicianbehavior and patient disposition needs to be assessed. Specificopportunities for research in agitation and deliriummanagement are highlighted in Figure 8.
The estimated costs of alcohol and other drug misuse to theUnited States is greater than $414 billion per year. From 1992to 2000, alcohol caused an estimated 68.8 million ED visits(7.9%), with 17.6 million adults reported to have metDiagnostic and Statistical Manual, Fourth Edition criteria forabuse/dependence. In 2005, an estimated 1.45 million ED visitswere related to drug use. An estimated 20.4 million Americansolder than 11 years use illicit drugs every month. Patients withunhealthy alcohol and drug use are more likely to use the ED astheir source of care, and timely screening, brief intervention,and referral to treatment (SBIRT) has the potential to close thegap between treatment need and services.26 Studies of tobaccouse in ED patients found prevalence rates as high as 48% inurban, medically underserved areas. Furthermore, 4.9% of alladult visits, 6.8% of admissions, and 10% of all hospital
be developed and validated.
s targeting acute suicide risk are required to determinens of suicidal patients.
ped and tested on high-risk populations.
ms through transdisciplinary research with investigators in
uicide screening and referral processes and of ED-initiatedpts.
policies and sustainable program enhancements and toagement of suicidal patients, as well as those experiencing
ning and surveillance.
riate treatment; in particular, thresholds for involuntary
ints and other interventions in emergency settings.
gitation and delirium.
d to
entlatio
velo
oble
sed sattem
orldman
prop
.
.
restra
financial charges were smoking-attributable.27
Annals of Emergency Medicine 557
acco
NIH Roundtable on Neurologic and Psychiatric Emergencies D’Onofrio
Although screening and brief interventions for alcoholdependence have been reported to be effective in the primarycare setting, the results from randomized controlled trials in EDsettings have been mixed. A recent meta-analysis revealed that,although interventions did not significantly reduce alcoholconsumption, they were associated with approximately half theodds of experiencing an alcohol-related injury.28 However, thescience for alcohol SBIRT is well advanced compared with thatfor tobacco and other drug use in the ED. Currently, there areno adequately validated screening tools for drug use and nopublished randomized controlled trials investigating the efficacyof SBIRT for drug misuse in the ED setting.27 Few rigorousrandomized controlled trials have been conducted to test ED-based SBIRT for tobacco control.
There are many unanswered questions in the ED setting tobe investigated about efficacy and effectiveness of screening,intervention, treatment in the ED and referral for alcohol,tobacco, and other drug misuse. Continued support from NIHfor investigator-initiated grants in this area, as well as multi-institute–supported request for applications and support fornetworks and collaborations, will be needed to move the fieldforward. Specific opportunities for research in substance abusein the ED are highlighted in Figure 9.
Posttraumatic stress disorder is a psychiatric symptombrought on by exposure to a traumatic life-threatening eventand has symptoms of intrusive memories (“reliving” the event),avoidance (of potential triggers), and arousal (outbursts,irritability, and overreaction to stress). Millions of traumaticinjuries with the potential to develop into posttraumatic stressdisorder are treated in the ED every year. Posttraumatic stressdisorder is estimated to occur in 10% to 25% of US adult injuryadmissions.29 Conditions and factors including severe orintentional injury, early psychological distress, previous trauma,
Research Opportunities/Gaps
Develop and validate feasible ED-based screening instruments
Investigate efficacy of brief interventions, including key compbooster, and who is best to provide the intervention.
Test new technologies and SBIRT delivery methods in the em
Determine what subset of patient is most likely to benefit.
Refine methodological issues: Limit assessment reactivity, credshow small effects and differences among patient populations,
Test effectiveness of SBIRT among special populations.
Include cost-effectiveness analysis studies.
Test treatments that can be initiated in the ED (buprenorphin
Investigate methods to improve adoption of research findings
Figure 9. Alcohol, tob
and a history of psychiatric illness, as well as being female,
558 Annals of Emergency Medicine
nonwhite, and of low socioeconomic status, all exacerbateposttraumatic stress disorder.
Cognitive behavioral therapy, exposure therapies, and otherbehavioral techniques are effective in reducing posttraumaticstress disorder symptoms and may also address comorbiddepression. The challenge comes in linking acute-settingpatients to these efficacious outpatient treatments.Pharmacologic methods are a little less well studied and at thispoint are inconclusive. To develop interventions that areultimately feasibly delivered in ED/trauma center settings, it isimportant to establish methods that include representativesampling procedures for acute care patients, providers, andpractice settings. The ED/trauma center can provide anexcellent model for using clinical epidemiology to developinterventions for the emergency settings. Trauma registriesprovide clinical and demographic information on all patientstreated within the acute care inpatient setting; thus, thecharacteristics of an individual patient or subgroup of patientsscreened for or included in an investigation can be comparedwith that of the entire clinical population of injured patientspresenting for care.30,31 Specific opportunities for research inposttraumatic stress are highlighted in Figure 10.
Violence in its various forms is often interconnected andoccurs across the life span in individuals, families, andcommunities.32 Because of the magnitude of the problem andthe focus of this group, this discussion is confined to intimatepartner violence,33 in particular neurologic and mental healthconditions associated with intimate partner violence. Patientsexperiencing intimate partner violence frequently present to theED for care of acute or chronic medical conditions related tointimate partner violence. The ED is an ideal place to identifyand provide interventions for these patients exposed to intimate
alcohol, tobacco, and other drug use.
ts of intervention, minimum effective dose, necessity of a
ncy setting.
control groups, the need for large sample populations touring and analyzing variability in drinking patterns.
ng-acting naltrexone, etc).
practice.
, and other drug use.
for
onen
erge
iblecapt
e, lo
into
partner violence.
Volume , . : November
em
D’Onofrio NIH Roundtable on Neurologic and Psychiatric Emergencies
About 25% of women in the United States experienceintimate partner violence during their lifetime. In 85% of cases,intimate partner violence involves women abused by malepartners, but adolescents, the elderly, and men, as well as thosein same-sex relationships, are also abused. In addition, 3 to 10million children witness intimate partner violence annually.Intimate partner violence has been estimated to have an annualcost of $8.3 billion. Some short- and long-term physical andmental health sequelae of intimate partner violence includeposttraumatic stress disorder (61%), depression (50%), andsuicidality (20%). Approximately 74% of intimate partnerviolence patients have sustained traumatic brain injury, with51% sustaining recurrent traumatic brain injury and 25%seeking medical care.
Many ED studies of intimate partner violence have focusedon the epidemiology among women with intimate partnerviolence, tools for use in identifying intimate partner violenceamong ED patients, the effect of acute interventions, andresearch on intimate partner violence outcomes. Amonginteresting research outside of the ED are studies of factors thatinfluence resilience or positive adaptation in the context ofsignificant adversity and the effects of witnessing intimatepartner violence in children and adolescents.34
Large, longitudinal outcome studies are necessary to assessthese issues in intimate partner violence, but funding forintimate partner violence research often falls between the cracks.Cross-disciplinary, multi-institute collaborations will benecessary to conduct this research. Further, when conductingintimate partner violence research, concerns for researchparticipants’ safety often make it difficult to follow up intimatepartner violence research participants. To improve surveillanceand compare across studies, the field must use standarddefinitions that have been developed.33 Specific opportunitiesfor research in intimate partner violence in the ED arehighlighted in Figure 11.
Most pediatric deaths occur in hospitals, either in the ICU orED; in approximately 5% of cases, children are declared deadon arrival. Physicians in these hurried settings are not only facedwith the difficulty of trying to save the life of the child but also
Research Opportunities/Gaps
Clinical epidemiologic investigation that develop trauma centthe generalizability of acute care posttraumatic stress disorder
Research on drugs hypothesized to prevent posttraumatic streglucocorticoids, adrenergic/�-blockers, opioid/analgesics, and
Prospective cohort and clinical trial investigations that integra
Stepped care/collaborative care intervention approaches that cand psychopharmacologic interventions targeting posttraumat
Figure 10. Posttraumatic stress in
with complex communication tasks, including discussing the
Volume , . : November
death with families, dealing with police and other legal issues,and discussing organ donation and autopsies.
An estimated 10% to 20% of bereaved individuals developsymptoms of complicated grief. Complicated grief is a patternof extreme symptoms—persistent disbelief, anger and bitterness,recurrent painful emotions, preoccupation with thoughts aboutthe deceased—that are prolonged (�6 months) and associatedwith mental and physical health outcomes.35 Emerging researchsuggests a much greater prevalence of complicated grief amongbereaved parents than among other bereaved individuals.Bereaved parents are at an increased risk for physical and mentalillness and mortality from natural and unnatural causes.
Recommended strategies to promote healthy parentalbereavement at the site of a child’s death include effectivecommunication,36 parental presence during procedures, physiciantraining, availability of bereavement services, and postmortemfollow-up, including telephone calls, cards, and parent-physicianmeetings. There is a need to assess the effectiveness of parent-physician postmortem meetings on promoting healthy parentalbereavement and reducing physician burnout.37
Detection, prevention, and treatment of bereavement at the siteof a child’s death lag behind that in other settings (with the possibleexception of pediatric palliative care). This lag may be due to theproportionately small number of deaths (relative to those of adults)and the focus on cure. There are few diagnostic criteria andtreatment recommendations for complicated grief. Studies areneeded to better determine incidence and prevalence ofconsequences of bereavement among bereaved parents; studies areparticularly needed that take into account the site and trajectory ofdeath and the population. Studies are also needed to determinestrategies to promote healthy bereavement and prevent mental andphysical consequences of the death of a child. Specific opportunitiesfor research in parental bereavement and complicated grief in theED are highlighted in Figure 12.
The NIH roundtable “Opportunities to Advance Researchon Neurological and Psychiatric Emergencies” created aframework to guide future emergency medicine–based researchinitiatives. Experts from emergency medicine, neurology,psychiatry, and public health assembled to establish the critical
d ED population-based registries to further assessments ofvention trials.
order (hypothalamic pituitary adrenal [HPA]/tive serotonin reuptake inhibitors [SSRIs]).
tcome assessments of patients with traumatic brain injuries.
ine care management with evidence-based psychotherapeuticress disorder and related disturbances such as depression.
ergency and acute care settings.
er aninter
ss disselec
te ou
ombic st
areas in need of investigation. Presentations and group
Annals of Emergency Medicine 559
ate p
eme
NIH Roundtable on Neurologic and Psychiatric Emergencies D’Onofrio
discussion firmly established the need for translational researchto bring basic science concepts into the clinical arena.Discussions also established the importance of a coordinatedcontinuum of care that ensures rapid identification andstabilization on the front end of the health care system to
Research Opportunities/Gaps
Surveillance, prevention, and intervention strategies.
Effect of acute ED-initiated interventions on mental health ou
Prevalence and development of interventions for both victimgroups.
Applicability of early treatments for acute stress disorder and pstress disorder in patients experiencing ongoing intimate partn
Prevalence and effects of intimate partner violence on neurolo
Diagnosis and effects of recurrent traumatic brain injury.
The relationship between postconcussive syndrome/traumatic
Prevalence of strangulation and resultant anoxic brain damage
Development of neuropsychiatric and neurobehavioral sequel
What populations should be evaluated for sequelae of traumatconducted (biomarkers, neuroimaging)?
Risk-factor prevalence and the effect of interventions on intim
Comparative effectiveness studies of intimate partner violence
Studies on resilience, factors that protect against posttraumativiolence patients’ abilities to access safety and determine optim
The relative importance of internal competencies versus externexposed to violence.
Vicarious traumatization of ED health care providers.
Figure 11. Intim
Research Opportunities/Gaps
Determine prevalence and predictors of bereavement at the sit
Test site-of-death interventions to promote healthy bereaveme
Conduct descriptive and epidemiologic studies on:● Differences in quality, frequency, and severity of parenta● Current hospital practices and attitudes and experiences
Develop and evaluate the feasibility and effectiveness of currenon-site tools to predict and prevent complicated grief, postmo
Assess how children’s death affects physicians and other clinic
Encourage collaborations with behavioral scientists.
Figure 12. Parental bereav
maximize overall patient outcomes.
560 Annals of Emergency Medicine
There were several themes that emerged during the course ofthe roundtable discussion:● Both the challenges and opportunities for research in
emergency medicine are driven by the unique attributes ofemergency settings; emergency services are gateways, the
es.
atterer depression and differences among racial/ethnic
raumatic stress disorder to prevent chronic posttraumaticiolence.
onditions.
n injury and posttraumatic stress disorder.
kes, and transient ischemic attacks.
strangulation and traumatic brain injury.
ain injury and anoxia? What types of evaluations should be
artner violence (alcohol interventions).
rventions, including health care–based crisis intervention.
ss disorder, and factors that influence intimate partnerpproaches to treatment.
ntext and cultural influences in the resilience of children
artner violence.
death.
nd reduce the incidence of complicated grief.
eavement by site of death, death trajectory, and populationspital staff
sed interventions (parental presence during procedures,parent-physician meetings, physician care and training).
nt and complicated grief.
tcom
and b
ostter v
gic c
brai
, stro
ae of
ic br
ate p
inte
c streal a
al co
e of
nt a
l berof ho
tly urtem
ians.
patients are often unknown, their problems are urgent, and
Volume , . : November
D’Onofrio NIH Roundtable on Neurologic and Psychiatric Emergencies
contact with the emergency service is brief or episodic. Thisposes unique problems for identification, consent, andtracking of research subjects. Therefore, the development ofa research infrastructure that incorporates innovativeinformatics technologies will be essential for future research.
● Patients often present with serious, nonspecific symptomsthat can impair communication in the absence of aknown history or diagnosis. The development ofdiagnostic strategies and tools is necessary to understandkey populations and the process of medicaldecisionmaking. Better methods for investigating thepathobiology of symptoms and symptom-orientedtherapies are also needed.
● Emergency medicine is an integrated system of care,including emergency medical services dispatch, out-of-hospital care, and ED management to admission oroutpatient transfer. This allows for multiple portals ofpatient entry and exit, with opportunities for enrollment ofresearch subjects into trials, coordination of care, andtracking late outcomes.
● Collaborative research networks can provide uniqueopportunities to form partnerships, leverage patient cohortsand clinical and financial resources, and share data. Theseexisting networks should be expanded and additionalnetworks formed that focus on emergency research.
● The development of formal research training programs isintegral for creating new knowledge and advancing thescience and practice of emergency medicine.In conclusion, EDs provide the portal of access to the
health care system for most patients with acute neurologicand psychiatric illness. There is a great need to promoteresearch in these settings. Emergency physicians andcolleagues are primed to investigate neurologic andpsychiatric emergencies that will directly improve thedelivery of care and patient outcomes.
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. The authorshave stated that no such relationships exist. See theManuscript Submission Agreement in this issue for examplesof specific conflicts covered by this statement.
Publication dates: Received for publication February 19, 2010.Revision received June 7, 2010. Accepted for publication June16, 2010.
Address for correspondence: Gail D’Onofrio, MD, MS,Department of Emergency Medicine, 464 Congress Ave, Ste260, New Haven, CT 06519; 203-785-7590, fax 203-785-
REFERENCES1. Nichol G, Thomas E, Callaway CW, et al. Regional variation in out-
of-hospital cardiac arrest incidence and outcome. JAMA. 2008;300:1423-1431.
2. Neumar RW, Nolan JP, Adrie C, et al. Post-cardiac arrestsyndrome: epidemiology, pathophysiology, treatment, andprognostication. A consensus statement from the InternationalLiaison Committee on Resuscitation (American Heart Association,Australian and New Zealand Council on Resuscitation, EuropeanResuscitation Council, Heart and Stroke Foundation of Canada,InterAmerican Heart Foundation, Resuscitation Council of Asia,and the Resuscitation Council of Southern Africa); the AmericanHeart Association Emergency Cardiovascular Care Committee; theCouncil on Cardiovascular Surgery and Anesthesia; the Council onCardiopulmonary, Perioperative, and Critical Care; the Council onClinical Cardiology; and the Stroke Council. Circulation. 2008;118:2452-2483.
3. Bernard SA, Gray TW, Buist MD, et al. Treatment of comatosesurvivors of out-of-hospital cardiac arrest with inducedhypothermia. N Engl J Med. 2002;346:557-563.
4. Hypothermia After Cardiac Arrest Study Group. Mild therapeutichypothermia to improve the neurologic outcome after cardiacarrest. N Engl J Med. 2002;346:550-556.
5. Majersik JJ, Silbergleit R, Meurer WJ, et al. Public health impactof full implementation of therapeutic hypothermia after cardiacarrest. Resuscitation. 2008;77:189-194.
6. Todd KH, Ducharme J, Choiniere M, et al. Pain in the emergencydepartment: results of the Pain and Emergency Medicine Initiative(PEMI) multicenter study. J Pain. 2007;8:460-466.
7. Todd K, Cowan P, Kelly N. Chronic or recurrent pain in theemergency department: a national telephone survey of patientexperience. Ann Emerg Med. 2007;50:S37.
8. Pletcher MJ, Kertesz SG, Kohn MA, et al. Trends in opioidprescribing by race/ethnicity for patients seeking care in USemergency departments. JAMA. 2008;299:70-78.
9. ER: the gateway for neurologic emergency trials? Ann Neurol.2006;60:A12-14.
10. Barsan WG, Pancioli AM, Conwit RA. Executive summary of theNational Institute of Neurological Disorders and Strokeconference on Emergency Neurologic Clinical Trials Network. AnnEmerg Med. 2004;44:407-412.
11. Huff JS, Decker WW, Quinn JV, et al. Clinical policy: critical issuesin the evaluation and management of adult patients presenting tothe emergency department with syncope. Ann Emerg Med. 2007;49:431-444.
12. Newman-Toker DE, Hsieh YH, Camargo CA Jr, et al. Spectrum ofdizziness visits to US emergency departments: cross-sectionalanalysis from a nationally representative sample. Mayo Clin Proc.2008;83:765-775.
13. Kattah JC, Talkad AV, Wang DZ, et al. HINTS to diagnose strokein the acute vestibular syndrome: three-step bedside oculomotorexamination more sensitive than early MRI diffusion-weightedimaging. Stroke. 2009;40:3504-3510.
14. Quinn J, Durski K. A real-time tracking, notification, and web-based enrollment system for emergency department research.Acad Emerg Med. 2004;11:1245-1248.
15. Unden J, Romner B. A new objective method for CT triage afterminor head injury—serum S100B. Scand J Clin Lab Invest. 2009;69:13-17.
16. Lynch JR, Blessing R, White WD, et al. Novel diagnostic test foracute stroke. Stroke. 2004;35:57-63.
17. Laskowitz DT, Kasner SE, Saver J, et al. Clinical usefulness of abiomarker-based diagnostic test for acute stroke: the BiomarkerRapid Assessment in Ischemic Injury (BRAIN) study. Stroke.
NIH Roundtable on Neurologic and Psychiatric Emergencies D’Onofrio
18. Ramchandren S, Cross BJ, Liebeskind DS. Emergent headachesduring pregnancy: correlation between neurologic examination andneuroimaging. AJNR Am J Neuroradiol. 2007;28:1085-1087.
19. Bushnell CD, Jamison M, James AH. Migraines during pregnancylinked to stroke and vascular diseases: US population basedcase-control study. BMJ. 2009;338:b664.
20. Larkin G, Beautrais A. Mental health and emergency medicine: aresearch agenda. Acad Emerg Med. 2009;16:1110-1119.
21. Larkin GL, Smith RP, Beautrais AL. Trends in US emergencydepartment visits for suicide attempts, 1992-2001. Crisis. 2008;29:73-80.
22. Claassen CA, Larkin GL. Occult suicidality in an emergencydepartment population. Br J Psychiatry. 2005;186:352-353.
23. Han JH, Zimmerman EE, Cutler N, et al. Delirium in olderemergency department patients: recognition, risk factors, andpsychomotor subtypes. Acad Emerg Med. 2009;16:193-200.
24. Lukens TW, Wolf SJ, Edlow JA, et al. Clinical policy: critical issuesin the diagnosis and management of the adult psychiatric patientin the emergency department. Ann Emerg Med. 2006;47:79-99.
25. Allen MH, Currier GW, Carpenter D, et al. The expert consensusguideline series. Treatment of behavioral emergencies 2005.J Psychiatr Pract. 2005;11(suppl 1):5-108; quiz 110-112.
26. D’Onofrio G, Degutis LC. Screening and brief intervention in theemergency department. Alcohol Res Health. 2004;28:63-72.
27. Cunningham R, Bernstein S, Walton M, et al. Alcohol, tobacco, andother drugs: future directions for screening and intervention in theemergency department. Acad Emerg Med. 2009;16:1078-1088.
28. Havard A, Shakeshaft A, Sanson-Fisher R. Systematic review andmeta-analyses of strategies targeting alcohol problems inemergency departments: interventions reduce alcohol-relatedinjuries. Addiction. 2008;103:368-376; discussion 377-378.
29. Zatzick D, Jurkovich GJ, Rivara FP, et al. A national US study ofposttraumatic stress disorder, depression, and work andfunctional outcomes after hospitalization for traumatic injury. AnnSurg. 2008;248:429-437.
30. Zatzick D, Koepsell T, Rivara F. Using target populationspecification, effect size, and reach to estimate and compare thepopulation impact of two PTSD preventive interventions.Psychiatry. 2009;72:346-359.
31. Zatzick D, Roy-Byrne P, Russo J, et al. A randomizedeffectiveness trial of stepped collaborative care for acutely injuredtrauma survivors. Arch Gen Psychiatry. 2004;61:498-506.
32. Krug E, Dahlberg L, Mercy J, et al, eds. World Report on Violenceand Health. Geneva, Switzerland: World Health Organization; 2002.
33. Saltzman L, Fanslow J, McMahon P, et al. Intimate partnerviolence surveillance: uniform definitions and recommended dataelements. In: Prevention NCfI, ed. 1999 version 1.0. Atlanta, GA:National Center for Injury Prevention Control, Center for DiseaseControl and Prevention; 1999.
34. Mitchell C, Anglin D, eds. Intimate Partner Violence—A Health-based Perspective. New York, NY: Oxford University Press; 2009.
35. Simon NM, Shear KM, Thompson EH, et al. The prevalence andcorrelates of psychiatric comorbidity in individuals withcomplicated grief. Compr Psychiatry. 2007;48:395-399.
36. Levetown M. Communicating with children and families: fromeveryday interactions to skill in conveying distressing information.Pediatrics. 2008;121:e1441-1460.
37. Meert KL, Eggly S, Pollack M, et al. Parents’ perspectives regardinga physician-parent conference after their child’s death in thepediatric intensive care unit. J Pediatr. 2007;151:50-55, e51-52.
38. Grieve AP, Krams M. ASTIN: a bayesian adaptive dose-responsetrial in acute stroke. Clin Trials. 2005;2:340-351; discussion
