Discovery Activities

Programs for Emergency Preparedness (PREP)

J. Perren Cobb, BA, MD, FCCM 

The appropriate treatment of critically ill or injured patients can vary from minute to minute. Thus, timely access to reliable data is one of the foundations of contemporary intensive care. It follows that optimal responses during public health emergencies, for both clinicians and decision-makers, would benefit from comprehensive, real-time event reporting. This should include physiologic patient data that are needed to provide immediate insight into the impact of the event on critical healthcare resources and to identify groups at high risk for morbidity and mortality. The goal of the Program for Emergency Preparedness (PREP) is to significantly enhance the national capability to rapidly glean crucial information regarding the clinical course of acute illness and injury and guide clinical resource requirements during emergent events through the following 6 aims: 

• Develop a national network of acute and critical care research organizations of academic and community hospitals for adults and children across the care continuum 
• Develop a rapid communication network with quarterly queries to assess national health system stress 
• Develop infrastructure for prospective trials of national public health emergencies, such as influenza and anthrax 
• Develop a national data coordinating center 
• Conduct human subject research review with local and national institutional review boards, such as the Public Health Emergency Research Review Board 
• Coordinate with international organizations and clinical trials groups 

Working with the Office of the Assistant Secretary for Preparedness and Response, the Food and Drug Administration, the Biomedical Advanced Research and Development Authority, the Centers for Disease Control and Prevention, the National Institutes of Health, leading professional organizations, and the Homeland Security Information Network, PREP is developing mechanisms for rapid data collection, analysis, and dissemination of findings during public health emergencies. Pre-event work on protocols, data collection processes, rapid analysis techniques, and means to quickly disseminate findings to stakeholders are all crucial for making clinical science networks effective at enhancing the response. PREP will leverage existing infrastructure to both strengthen pre-event operational science capabilities and provide timely data and situational awareness across the emergency care continuum during public health emergencies. Critical illness and injury professional organizations will use this rapid dissemination plan to inform their membership, in aggregate representing over 150,000 frontline clinicians, thereby saving lives and minimizing suffering based on the timely accurate guidance gleaned from operational science. 

 

Prevention of Organ Failure (PROOF)

Michelle Ng Gong, MD 

The Preventions of Organ Failure (PROOF) program is composed of critical care researchers, represented by a multidisciplinary group of critical care specialists from anesthesia, emergency medicine, internal medicine, pulmonary, surgery, and trauma, who are interested in the prevention of injuries and diseases and their progression in the critically ill population. 

Projects conducted under PROOF: 

• Lung Injury Prediction Study 
• Lung Injury Prevention Study - Aspirin (LIPS-A) 
• Accurate Prediction of Prolonged Ventilation (APPROVE) and Prevention of Organ Failure Checklist (PROOFCHECK) 
• iCERTAIN 
• HEMAIR 

Critical Illness Outcomes Study (CIOS)

Jonathan E. Sevransky, MD, FCCM 

Variations in both structure and process are known to affect clinical outcomes in intensive care units (ICUs). With both increasing demand and increasing costs of adult critical care, it is important to understand how best to reduce variations in care. The Critical Illness Outcomes Study (CIOS) was designed to characterize the organizational structure, processes of care, use of protocols, and outcomes of ICUs and to determine which of these structural and process-of-care factors might be associated with outcomes such as inpatient mortality. 

Of the 94 U.S. ICUs we approached, 69 are participating in the study; 25 (36%) are medical, 24 (35%) surgical, and 20 (29%) mixed. We surveyed the 69 ICUs about their organization, size, volume, staffing, processes of care, and use of protocols, and investigated the relationship of structure and process to ICU mortality. 

We collected patient demographic and treatment information on 1 day each week until at least 100 patients were enrolled in each participating ICU. We have completed enrollment, with more than 6400 patients, and are currently validating the data collected. CIOS is planning a second study to better determine which factors might be associated with high-performing ICUs. 

Practices Surrounding the Identification, Prevention, and Treatment of Delirium in the ICU

Amy L. Dzierba, PharmD, BCPS, FCCM 

This study aims to describe the practices of detection, prevention, and treatment of delirium in adult intensive care unit (ICU) patients across institutions and to compare the perceived and actual activities surrounding detection, prevention, and treatment activities in a snapshot. This study will provide ICU clinicians, hospital administrators, and researchers with information on discrepancies between actual patient care and recently published evidence-based guideline recommendations. 

A Multicenter Study to Evaluate Predictive Factors for Multidrug-Resistant Healthcare-Associated Pneumonia in Critically Ill Patients (DEFINE)

Ishaq Lat, PharmD, FCCM 

Pneumonia is a leading cause of death in the United States and is associated with significant costs to the healthcare system. Increasing rates of multidrug-resistant (MDR) pathogens challenge critical care clinicians to provide effective antimicrobial therapy while preserving the armamentarium of effective therapies. Literature describing the incidence and epidemiology of MDR pneumonia in the United States is limited. We conducted this study across 35 U.S. sites to elucidate the incidence of MDR pneumonia in the critical care setting. 

Checklist for Early Recognition and Treatment of Acute Illness and Injury (CERTAIN)

Rahul Kashyap, MBBS 

Checklist for Early Recognition and Treatment of Acute Illness and injury (CERTAIN) is designed to standardize the approach to the evaluation and treatment of acutely decompensating patients. The design and content were informed by a survey of clinicians from diverse international settings. Available in electronic (laptop and mobile) and paper formats, CERTAIN provides evidence-based diagnostic checklists, clinical decision support, educational modules on performing critical procedures, and the ability to time and document real-time interventions. 

http://www.icertain.org 

Vitamin C, Corticosteroids, and Thiamine in Sepsis Study (VICTAS) 

Jonathan Sevransky, MD, MHS 

Sepsis is a clinical syndrome characterized by life-threatening organ dysfunction caused by a dysregulated host response to infection. The treatment of sepsis currently consists of expedient supportive care and control of infection. Recently, a combination of inexpensive medications has been asserted to effectively combat the dysregulated metabolism that accompanies sepsis. These medications include vitamin C, thiamine (vitamin B1), and hydrocortisone. The assertion is based on the before-and-after data reported by a single hospital. The trial received widespread public interest. Response has been mixed, with some clinicians suggesting that the results are too good to be true and others asking whether the combination of medications should be given to all sepsis patients. The most general opinion reflects clinical equipoise: genuine uncertainty among the expert medical community over whether the medication combination treatment will be beneficial, with reciprocal uncertainty as to whether the medication combination will be wasteful or even harmful. 

We propose to evaluate this potential treatment in a randomized placebo-controlled, multicenter trial. This trial will require 40-60 sites and will have as primary outcome measures the number of days of vasopressor- and ventilator-free days. 

Severe ARDS: Generating Evidence (SAGE)

Pauline K. Park, MD, FCCM 

Approximately one-quarter of patients with acute respiratory distress syndrome (ARDS) develop severe hypoxemia (Pao2/Fio2 < 100). In large series, severe hypoxemia has been associated with high observed mortality rates, approximating 40%-50%. The severity of hypoxia in the majority of these patients is established at initial presentation, suggesting an opportunity for early intervention. The development of strategies to reduce mortality is hampered by the difficulty of conducting randomized trials in this population. 

While a number of interventions for ARDS have been shown in randomized, controlled trials to lead to improved outcomes, studies to date indicate that the use of these evidence-based practices is highly variable and inconsistent. At the same time, unproven treatment modalities are commonly used in the management of ARDS. The SAGE study was conducted to evaluate current U.S. practices in the management of severe ARDS, both to inform practicing clinicians and to form the basis for future research. 

Objective: Assessment of early management of severe ARDS, including ventilator management and use of rescue therapy. 

Specific aims: 

• Describe U.S. management practices and variation in the use of ventilator strategies and rescue modalities in patients with severe ARDS 
• Determine prospectively the factors in early severe ARDS associated with survival or need for adjuvant therapy 
• Evaluate characteristics, management, and survival in patients admitted to SAGE centers compared with patients transferred to SAGE centers from other sites 
• Determine the variability in the use of tidal volume and positive end-expiratory pressure and its association with subsequent mortality among patients with severe ARDS on extracorporeal membrane oxygenation 

Design: multicenter, prospective, observational cohort study conducted through participating sites between October 2016 and April 2017 

 Inclusion criteria: 

• Age above 18 years 
• Patients with acute respiratory failure in the intensive care unit requiring invasive mechanical ventilation 
• Presence of severe ARDS 

Structure, Process, and Utilization of Intermediate Care in the United States

David N. Hager, MD, PhD 

Intensive care resources are limited, while the number of patients needing intensive care is increasing. It has previously been recognized that, among patients admitted to intensive care units (ICUs), many do not require intensive care but are admitted for close monitoring. Intermediate care units (IMCUs), also known as high-dependency units, step-down units, or progressive care units, were created to accommodate patients who do not require intensive care but whose needs surpass the care and monitoring feasible on a general ward. Patients may be transitioned to an IMCU after being stabilized in an ICU or having worsened on a general ward or may be directly admitted from the emergency department or post-anesthesia care unit. 

Over the past 20 years, billing for intermediate care and the prevalence of IMCUs have increased. However, the optimal staffing structure, physical layout, and admission guidelines for these units are not well defined. This is complicated by differences in regional needs, institutional missions, clinical expertise, and physical resources. This marked heterogeneity of IMCUs and the characteristics of the patients they serve has resulted in limited generalizability of IMCU patient outcomes and cost-effectiveness studies to date. A better understanding of IMCU organizational structure paired with patient outcomes would greatly inform the use of this level of care in the future. 

This will be a study of variability in the structure and use of IMCUs in different regions and centers. We will survey centers to characterize the current structure of intermediate care in the United States. 

Early After TBI Study: Opposing Sympathetic Drive Early After Traumatic Brain Injury - Phase 3 (OSD-3)

Bellal Joseph, MD, FACS 

Severe traumatic brain injury (TBI) is defined by a Glasgow Coma Scale (GCS) score of less than or equal to 8. This sets a host-adaptive neuroendocrine, immune-metabolic, and inflammatory response that is integrated by an increased sympathetic drive and exaggerated catecholamine surge. This unopposed sympathetic drive triggers a secondary brain insult that occurs hours to days after the primary TBI and manifests as systemic and intracranial hypertension, abnormal heart rate variability, agitation, cerebral edema, and cerebral hypoperfusion to collectively cause the poor neuropsychological outcome. Opposing the sympathetic drive through neutralization of the catecholamine actions in TBI is therefore a viable option for neuroprotection against a secondary brain insult. However, there is no clinical evidence from a prospective randomized trial demonstrating the safety and effectiveness of opposing the sympathetic drive after TBI. 

Methods/design: OSD-3 is a 2-arm, single-blinded, block-randomized, controlled phase III multicenter trial of the efficacy of early opposing of the sympathetic drive as a neuroprotection strategy in TBI patients with a GCS score of less than or equal to 8. One-half of the qualified patients will be enrolled via exception from informed consent or proxy consent and block-randomized to the opposed sympathetic drive experimental arm. This arm will receive IV propranolol every 6 hours at an adjustable dose to achieve a targeted heart rate of less than 100 beats/min. Propranolol will be held for hypotension (systolic blood pressure < 100 mm Hg) or bradycardia (heart rate < 60 beats/min). The maximum daily dose for the treatment of hypertension of 640 mg will not be exceeded in this study. Patients randomized to the unopposed sympathetic drive control arm will receive the standard-of-care treatment and will not receive propranolol or other β-adrenoceptor antagonist or α-2-agonist. If a patient randomized to the unopposed sympathetic drive arm develops hypertension and increased heart rate, this patient will be treated according to the standard of care by the patient’s trauma team. End-point measurements include plasma catecholamine and metabolites levels, serum levels of neuronal injury-specific biomarkers, heart rate variability, arrhythmia occurrence, infection rate, medication use, agitation measures, coma- and ventilator-free days, intensive care unit (ICU) and hospital lengths of stay, and mortality. Neuropsychological outcomes in the domains of reasoning, concentration, problem-solving, and memory, together with the Glasgow Outcome Scale - Extended (GOS-E) will be performed at hospital discharge and at 3 and 12 months after the injury. 

The specific aims are to demonstrate that early administration of propranolol after severe TBI: 

• Decreases the plasma levels of catecholamine (epinephrine, norepinephrine, and dopamine) and their breakdown products (vanillylmandelic acid, metanephrine, and normetanephrine) measured at day 7 
• Decreases serum levels of neuronal injury-specific biomarkers (S100-B, neuron-specific enolase, myelin basic protein, ubiquitin C-terminal hydrolase, tau protein, and glial fibrillary acidic protein) measured at day 7 
• Decreases heart rate variability, agitation, arrhythmia, and infection as assessed daily 
• Decreases ventilator days, ICU and hospital lengths of stay, and in-hospital mortality as assessed daily 
• Improves neuropsychological outcome in the domains of reasoning, concentration, problem-solving, and memory as measured at hospital discharge and at 3 and 12 months after injury 
• Improves neurologic outcome as measured by GOS-E at 3 and 12 months after injury 

Monitoring EEG in Comatose ICU Patients: A Point Prevalence Observational Study of Intensive Care Unit Experiences (MECIP)

Manuel M. Buitrago Blanco, MD, PhD 

Nonconvulsive seizures detected by continuous electroencephalography (cEEG) occur in about 30% of patients in coma. cEEG is also indicated for aiding in management and prognosis of critically ill patients with brain injury. Guidelines recommend the use of cEEG monitoring in this population to better aid in their management, yet implementation remains challenging. 

This research proposal consists of an observational study of the use of cEEG for the evaluation and treatment of patients in coma in intensive care units (ICUs) across the United States. The goal is to assess for current practice, quality, barriers to implementation and opportunities for improvement in the use of cEEG for evaluation and treatment of patients in coma. 

This is a cross-sectional point prevalence observational detailed study of users, equipment, readers, patient categories, and EEG duration in both specialty and general ICUs. One hundred institutions will be recruited through an outreach effort via the Neurocritical Care Research Network (NCRN) and the Critical Care Research Network (CCR-Net). With the endorsement of the NCRN, participation of its 47 institutional members is anticipated. We expect to enroll additional institutions, both academic and nonacademic, via the CCR-Net. 

Through MECIP, we expect to identify areas of top priority to improve the delivery of standard of care in coma patients admitted to critical care units by identifying the main barriers to implementation of cEEG. The direct impact of this study will translate into better delivery of care and improved ability to execute collaborative inter-institutional research endeavors in the future. 

Specific aims are to: 

• Determine clinical practices for patients in coma in U.S. ICUs. We hypothesize that cEEG monitoring is underutilized in coma in most centers, independent of unit type. 
• Identify effects on predefined outcomes such as mortality, length of stay, and adverse effects of cEEG monitoring of coma patients. We hypothesize that utilization of goal-directed cEEG monitoring reduces mortality. 
• Define specific barriers to practical implementation of cEEG in coma patients once the indication for testing has been recognized by intensive care specialists. We hypothesize that the 2 main barriers are equipment unavailability and lack of neurology EEG readers. 

Titration of Inspired Oxygen During Mechanical Ventilation Using Electronic Alerts via Electronic Health Records: A Multicenter Study

Sonal R. Pannu, MD, MS 

Hyperoxia, defined as fraction of inspired oxygen (Fio2) of greater than 0.5, can be injurious, augments ventilator-associated lung injury, and is associated with higher mortality. Liberal oxygenation practices are also associated with increased mortality in subsets of critically ill patients with post-cardiac arrest, stroke, and traumatic brain injury. Fio2 is titrated via oxygen saturations (Spo2); however, there is often a delay in reducing Fio2 despite adequate Spo2. Processes of ventilator weaning and liberation may be delayed with inadequate titration. Hyperoxia prevails in most intensive care unit (ICU) settings due to poor awareness of the adverse effects of even mild hyperoxia and the fear that even mild or short duration of hypoxia could be life-threatening. Therefore, there is a critical need to institute measures to improve the practice of Fio2 titration in a conservative range to maintain optimal oxygen saturation. 

The plan is to conduct a step-wedge, clustered, randomized implementation by sequential adoption every 3 months in participating ICUs with concurrent controls available until all sites adopt the protocol. 

The aims of this study are to: 

• Reduce the duration of hyperoxia in mechanically ventilated critically ill patients 
• Demonstrate improved ICU outcomes of increased ventilator-free days and shorter ICU length of stay 
• Study clinicians for burden of electronic alerts and learning effect of the alerting process 

 

VOLUME-CHASERS: Observation of Variation in Fluids Administered and Characterization of Vasopressor Requirements in Shock

Jen-Ting Chen, MD 

While fluid resuscitation is a mainstay of treatment for most patients with shock, excessive volume resuscitation is associated with worse clinical outcomes. Many studies have shown that dynamic hemodynamic measurements can predict fluid responsiveness, but little is known about their association with important clinical outcomes. The overall goal of VOLUME-CHASERS is to conduct a multicenter, observational cohort study across a broad range of hospitals, including patients in the emergency department, intensive care unit (ICU), and non-ICU areas, to determine the variability in shock resuscitation and modalities used to determine the amounts of fluid and doses of vasopressor administered. We will explore the possible outcome differences associated with this variability in practice. 

Inhaled Versus Early Systemic Steroids for Treatment of Pneumonia (INVESST Pneumonia)

Emir Festic, MD, MS, FCCM 

Inhaled Versus Early Systemic Steroids for Treatment of Pneumonia (INVESST Pneumonia) will be a multicenter, double-blind, placebo-controlled, 3-arm randomized trial to compare the efficacy of early treatment with an inhaled corticosteroid combined with a beta-agonist versus systemic corticosteroids versus usual care for prevention of acute respiratory failure (ARF) requiring mechanical ventilation and to reduce hospital length of stay in patients with severe pneumonia. 

The 3 aims are to: 

• Test the efficacy of early treatment with an inhaled corticosteroid (budesonide, 0.5 mg) combined with a beta-agonist (formoterol, 20 µg) or systemic steroid (IV methylprednisolone, 0.5 mg/kg) versus usual care for the prevention of ARF 
• Compare the efficacy and side effect profiles of inhaled delivery of a corticosteroid combined with a beta-agonist versus systemic steroids in their effect on hospital length of stay, duration of need for supplemental oxygen, hyperglycemia, and arrhythmias 
• Identify biologic pathways associated with progression to ARF in patients with pneumonia and explore the effect of systemic versus inhaled delivery of corticosteroids on peripheral markers of inflammation (interleukin-6 and C-reactive protein), inflammasome activation (interleukin-18), and markers of endothelial (angiopoietin-2) and epithelial (receptor for advanced glycation end-products [RAGE]) lung injury. 

High-Flow Oxygen Versus Positive Pressure Ventilation in the Emergency Department Program (USCIIT-HOPE)

Jarrod M. Mosier, MD, FCCM 

The management of acute hypoxemic respiratory failure (AHRF) in the emergency department (ED) is difficult because of resource and logistical challenges. This is particularly true for patients with, or at risk for, acute lung injury. Many clinicians prefer treating these patients with noninvasive positive pressure ventilation (NIPPV) with the goals of preventing the need for deep sedation and invasive mechanical ventilation. More recently, a newer and potentially more efficacious therapy has been introduced: high-flow nasal cannula (HFNC) with a heated, humidified circuit and adjustable fraction of inspired oxygen with flows between 40 and 60 L/min. Recent studies have shown that HFNC might be superior for patients with AHRF in the intensive care unit. However, the utility of starting this therapy earlier in the disease course—in the ED—is unknown. 

Although the mechanistic basis for any superiority of HFNC over NIPPV is unclear, presumably HFNC reduces lung injury by maintaining a lower transpulmonary pressure gradient and more lung-protective tidal volumes than NIPPV in spontaneously breathing patients with airspace disease. Thus, earlier initiation of this therapy in the ED would provide a greater benefit. 

The overall goal of our research is to design a prospective multicenter randomized controlled trial to compare the rate of intubation at 72 hours for HFNC and NIPPV in adult ED patients with AHRF. 

The 3 specific aims are to: 

1. Determine the failure rates (rate of intubation) of each therapy to adequately power a multicenter trial 

2. Compare rates of intubation in ED patients with AHRF treated with either HFNC or NIPPV 

3. Assess lung injury and inflammation in these patients before, during, and after treatment with either FHNC or NIPPV 

Recovery From Acute Kidney Injury Requiring Dialysis: Incidence, Timeline, and Risk Prediction Models

Javier A. Neyra, MD, MSCS 

Acute kidney kidney injury (AKI) is a complex systemic syndrome associated with high morbidity and mortality. More than 5 million patients are admitted to hospital intensive care units (ICUs) each year in the United States. Approximately 10% of these patients have AKI requiring dialysis (AKI-D). Since January 1, 2017, healthcare law has included a provision on coverage and payment for outpatient renal dialysis services for individuals with AKI. This legislation constitutes a timely and needed change in practice because it is estimated that a significant proportion of AKI-D survivors can sufficiently recover renal function (without needing dialysis) within the first 90 days after the initial insult. Therefore, there is an urgent need to characterize the biological natural course of AKI-D recovery and to develop clinical risk prediction models to identify patients at high risk for not significantly recovering kidney function by the time of hospital discharge. 

In this context, we propose a multicenter collaborative registry study in the ICU to 1) determine the incidence and timeline occurrence of AKI-D recovery, 2) determine clinical parameters associated with AKI-D recovery, and 3) develop and validate risk prediction models of AKI-D recovery. Our proposed study is a critical step for the development of post-AKI risk stratification tools and the identification of best practices for the optimal care of AKI-D survivors. We propose to evaluate this potential treatment in a randomized placebo-controlled, multicenter trial. This trial will require 40-60 sites. Primary outcome measures will be the number of vasopressor-free days and ventilator-free days. 

Oral Midodrine Hydrochloride in Early Sepsis: Randomized, Double-Blind, and Placebo-Controlled Feasibility Study

Rahul Kashyap, MBBS 

The aim of this patient-centered study is to conduct a feasibility clinical trial on oral midodrine in early sepsis and to seek alternatives to minimize the burden of an intensive care unit (ICU) stay in these patients. 

Sepsis is the second leading cause of death in medical ICUs, carrying a mortality rate of between 25% and 30%. Cardiovascular compromise in sepsis manifests as hypotension due to arterial vasodilation between 25% and 30%. Cardiovascular compromise in sepsis manifests as hypotension due to arterial vasodilation. Hypotension can persist despite initial resuscitation, prompting additional fluid boluses and subsequent central venous catheterization for the infusion of intravenous vasopressor agents. Both excess fluid boluses and central venous catheterization may expose patients to risk, harm, and discomfort. 

Midodrine is an oral vasopressor approved for treating orthostatic hypotension. Preliminary data during the past several years suggest a markedly increased off-label use as a vasopressor-sparing agent in critically ill patients. However, no randomized trials have been conducted to evaluate the safety and efficacy of this practice. 

The central hypothesis is that administering oral midodrine to patients with sepsis who have received initial fluid resuscitation and appropriate antimicrobial treatment will mitigate systemic hypotension and decrease the need for additional fluids and vasopressors. The proposed multicenter pilot trial is necessary to test the feasibility of enrollment, appropriate population, timing, effect size to determine the need, and sample size for a subsequent phase II pragmatic clinical trial. 

Worldwide Assessment of Separation of Patients From Ventilatory Assistance (WEAN SAFE)

Philippe R. Bauer, MD, PhD 

Successful weaning of patients from invasive mechanical ventilation (IMV) represents a crucial step in the recovery process following severe respiratory failure and is a key clinical challenge for intensive care unit (ICU) clinicians. Many of the serious complications of IMV are directly related to the duration of ventilation. Failure to successfully liberate patients from IMV contributes directly to poorer patient outcomes such as longer duration of ventilation, longer ICU and hospital lengths of stay, and higher mortality. Patients spend a considerable amount of time being liberated from IMV. The systematic utilization of approaches to reduce the duration of ventilation are therefore of fundamental importance. 

Despite the importance of the weaning period, this process is not rigorously defined, with wide variations in definitions and practices. In addition, the specific impact of weaning difficulties on patient outcomes is still poorly understood. While guidelines exist on the classification of weaning, a key recent study has shown that these are not applicable to all patients. Moreover, different practices exist in regard to weaning procedures, and some confusion exists even in what should be considered the beginning of the weaning process. This is an important problem because general recommendations regarding the weaning process may encompass completely different causes and consequences of its prolongation and therefore may be totally inappropriate for certain patients. 

The WEAN SAFE study aims to address key issues relating to weaning from IMV. WEAN SAFE will have a structure similar to LUNG SAFE, in that a large set of patients receiving IMV will be enrolled without setting weaning as an inclusion criterion but rather attempting to identify the weaning process retrospectively. 

WEAN SAFE aims to describe the current procedures for weaning in a large population of ICU patients; the applicability of existing classification systems to real-world situations; weaning from IMV; and center, management, and patient characteristics associated with duration of weaning. It will answer the following questions: 

• What is the frequency of delayed weaning from IMV? 
• What approaches are currently used for weaning patients from IMV? 
• What factors are used to determine when patients are in the weaning phase? 
• What are the barriers to effective weaning from IMV? 
• What factors (patient, institutional, medical practice) contribute to failed attempts to wean from IMV? 
• What is the impact of sedation management on weaning from IMV? 
• What is the impact of premorbid conditions and frailty on weaning from IMV? 
• What is the utility of existing classifications for weaning from IMV? 
• What is the impact of early versus delayed and/or failed weaning from IMV? 
• What regional or geo-economic differences exist regarding weaning from IMV? 
• What therapeutic resources are used in patients with delayed weaning from IMV? 

Development and Implementation of Clinical Prediction Models for Seizures During Post-Cardiac Arrest Care

Teresa May, DO, MS 

Cardiac arrest is a major public health problem, affecting more than 500,000 Americans annually. Approximately 100,000 of these patients survive cardiopulmonary resuscitation and are admitted to hospitals, where they are treated in intensive care units (ICUs). Despite this specialized care, the hospital mortality rates of these cardiac arrest survivors are high, ranging from 20% to 60%. Most of these deaths are caused by hypoxic-ischemic encephalopathy (HIE), a brain injury resulting from inadequate blood flow to the brain during the arrest. HIE is accompanied by a high frequency of seizures, which are thought to exacerbate the acute brain damage caused by the arrest. Prevention of seizures after cardiac arrest is an important and novel treatment target with considerable potential to improve outcomes in this vulnerable population. 

The aims of this research are to: 

• Develop a clinical prediction model of post-anoxic seizures to identify patients at risk for seizures at the time of hospital admission for a cardiac arrest using the existing International Cardiac Arrest Registry (INTCAR) database 
• Validate the prediction model using the INTCAR neurology subset of patients collected after 2018 
• Develop a prediction model of existing data from 1,987 patients with cardiac arrest who were monitored on continuous EEG in INTCAR from 2008 to May 2017 to identify patients most likely to develop seizures 
• Validate the model on non-overlapping patients undergoing EEG from the INTCAR neurology data collected after May 2017 
• Develop a seizure prediction interface using an electronic health record (EHR)-based tool easily accessible at the time of hospital admission 
• Disseminate it at 5 centers to evaluate its utility in the clinical setting 

The overall goal of this research is to facilitate the established INTCAR neurology prospective data collection and assist in the selection of centers and implementation of an EHR prediction tool. The concepts and expertise learned from this project will be readily applicable to other disease decisions in both cardiac arrest and the neuro-ICU.