Artificial Intelligence in Emergency Medicine: Benefits, Risks, and Recommendations.

BACKGROUND: Artificial intelligence (AI) can be described as the use of computers to perform tasks that formerly required human cognition. The American Medical Association prefers the term 'augmented intelligence' over 'artificial intelligence' to emphasize the assistive role of computers in enhancing physician skills as opposed to replacing them. The integration of AI into emergency medicine, and clinical practice at large, has increased in recent years, and that trend is likely to continue. DISCUSSION: AI has demonstrated substantial potential benefit for physicians and patients. These benefits are transforming the therapeutic relationship from the traditional physician-patient dyad into a triadic doctor-patient-machine relationship. New AI technologies, however, require careful vetting, legal standards, patient safeguards, and provider education. Emergency physicians (EPs) should recognize the limits and risks of AI as well as its potential benefits. CONCLUSIONS: EPs must learn to partner with, not capitulate to, AI. AI has proven to be superior to, or on a par with, certain physician skills, such as interpreting radiographs and making diagnoses based on visual cues, such as skin cancer. AI can provide cognitive assistance, but EPs must interpret AI results within the clinical context of individual patients. They must also advocate for patient confidentiality, professional liability coverage, and the essential role of specialty-trained EPs.

Ranking the Attributes of Effective Disaster Responders and Leaders.

OBJECTIVE: To investigate the relative importance of 10 attributes identified in prior studies as essential for effective disaster medical responders and leaders. METHODS: Emergency and disaster medical response personnel (N=220) ranked 10 categories of disaster worker attributes in order of their importance in contributing to the effectiveness of disaster responders and leaders. RESULTS: Attributes of disaster medical leaders and responders were rank ordered, and the rankings differed for leaders and responders. For leaders, problem-solving/decision-making and communication skills were the highest ranked, whereas teamwork/interpersonal skills and calm/cool were the highest ranked for responders. CONCLUSIONS: The 10 previously identified attributes of effective disaster medical responders and leaders include personal characteristics and general skills in addition to knowledge of incident command and disaster medicine. The differences in rank orders of attributes for leaders and responders suggest that when applying these attributes in personnel recruitment, selection, and training, the proper emphasis and priority given to each attribute may vary by role. (Disaster Med Public Health Preparedness. 2019;13:700-703).

Characteristics of Effective Disaster Responders and Leaders: A Survey of Disaster Medical Practitioners.

OBJECTIVE: To identify key attributes of effective disaster/mass casualty first responders and leaders, thereby informing the ongoing development of a capable disaster health workforce. METHODS: We surveyed emergency response practitioners attending a conference session, the EMS State of the Science: A Gathering of Eagles. We used open-ended questions to ask participants to describe key characteristics of successful disaster/mass casualty first responders and leaders. RESULTS: Of the 140 session attendees, 132 (94%) participated in the survey. All responses were categorized by using a previously developed framework. The most frequently mentioned characteristics were related to incident command/disaster knowledge, teamwork/interpersonal skills, performing one's role, and cognitive abilities. Other identified characteristics were related to communication skills, adaptability/flexibility, problem solving/decision-making, staying calm and cool under stress, personal character, and overall knowledge. CONCLUSIONS: The survey findings support our prior focus group conclusion that important characteristics of disaster responders and leaders are not limited to the knowledge and skills typically included in disaster training. Further research should examine the extent to which these characteristics are consistently associated with actual effective performance of disaster response personnel and determine how best to incorporate these attributes into competency models, processes, and tools for the development of an effective disaster response workforce. (Disaster Med Public Health Preparedness. 2016;page 1 of 4).

Gifts to physicians from industry: the debate evolves.

In October 2009, the board of directors of the American College of Emergency Physicians (ACEP) approved a major revision to ACEP's "Gifts to Emergency Physicians from Industry" policy. The revised policy is a response to increasing debate and calls for restriction of the long-standing biomedical industry practice of giving promotional gifts to individual physicians. This article outlines the history of professional attention to gift giving and reviews recent contributions to the ongoing debate over its justifiability, including professional association recommendations for limitation or prohibition of the practice. The article concludes with a description of the provisions of the revised ACEP gifts policy and brief reflection on the future of this practice.

A preliminary naturalistic study of low-dose ketamine for depression and suicide ideation in the emergency department.

We examined the preliminary feasibility, tolerability and efficacy of single-dose, intravenous (i.v.) ketamine in depressed emergency department (ED) patients with suicide ideation (SI). Fourteen depressed ED patients with SI received a single i.v. bolus of ketamine (0.2 mg/kg) over 1-2 min. Patients were monitored for 4 h, then re-contacted daily for 10 d. Treatment response and time to remission were evaluated using the Montgomery-Asberg Depression Rating Scale (MADRS) and Kaplan-Meier survival analysis, respectively. Mean MADRS scores fell significantly from 40.4 (s.e.m.=1.8) at baseline to 11.5 (s.e.m.=2.2) at 240 min. Median time to MADRS score ≤10 was 80 min (interquartile range 0.67-24 h). SI scores (MADRS item 10) decreased significantly from 3.9 (s.e.m.=0.4) at baseline to 0.6 (s.e.m. =0.2) after 40 min post-administration; SI improvements were sustained over 10 d. These data provide preliminary, open-label support for the feasibility and efficacy of ketamine as a rapid-onset antidepressant in the ED.

Community responses to a suicidal crisis: implications for suicide prevention.

We conducted a cross-sectional, random-digit-dial survey to evaluate public responses to a hypothetical question: "If someone you knew was suicidal, what would you do first?" Younger people were more likely to call a suicide hotline, and less likely to go to an emergency room (ER) or call 911; immigrants (in the U.S. < 15 years) were more likely to call 911, and less likely to call a suicide hotline; African Americans were more likely to go to the ER and call 911; Hispanics were more likely to call 911 but less likely to call a suicide hotline. These results suggest that public messages about hotlines and emergency options for suicidal patients need to be tailored to relevant population characteristics including age, education, ethnicity, and language preferences.

Patient-centered integrated networks of emergency care: consensus-based recommendations and future research priorities.

Patient-centered care is defined by the Institute of Medicine (IOM) as care that is responsive to individual patient needs and values and that guides the treatment decisions. This article is a result of a breakout session of the 2010 Academic Emergency Medicine (AEM) consensus conference and describes the process of developing consensus-based recommendations for providing patient-centered emergency care. The objectives of the working group were to identify and describe the critical gaps in the provision of patient-centered care, develop a consensus-based research agenda, and create a list of future research priorities. Using e-mail and in-person meetings, knowledge gaps were identified in the areas of respect for patient preferences, coordination of clinical care, and communication among health care providers. Four consensus-based recommendations were developed on the following themes: enhancing communication and patient advocacy in emergency departments (EDs), facilitating care coordination after discharge, defining metrics for patient-centered care, and placing the locus of control of medical information into patients' hands. The set of research priorities based on these recommendations was created to promote research and advance knowledge in this dimension of clinical care.

Commentary: doctors without boundaries: the ethics of teacher-student relationships in academic medicine.

Possessed of both instinct and intellect, physician teachers are required to be respectful exemplars of professionalism and interpersonal ethics in all environments, be it the hospital, classroom, or outside the educational setting. Sometimes, even while protecting the sanctity of the teacher-student relationship, they may surreptitiously find themselves in the throes of consensual intimacy, boundary violations, student exploitation, or other negative interpersonal and/or departmental dynamics. One may question how an academic can consistently resolve this tension and summon the temperance, humility, charity, and restraint needed to subdue lust, pride, abuse, and incontinence in the workplace. One important answer may lie in an improved understanding of the moral necessity of social cooperation, fairness, reciprocity, and respect that is constitutive of the physician-teacher role. Although normative expectations and duties have been outlined in extant codes of ethics and conduct within academic medicine, to date, few training programs currently teach faculty and residents about the ethics of appropriate pedagogic and intimate relations between teaching staff and students, interns, residents, researchers, and other trainees. This essay highlights examples from history, literature, and medical ethics as one small step toward filling this void.

Pre-resuscitation factors associated with mortality in 49,130 cases of in-hospital cardiac arrest: a report from the National Registry for Cardiopulmonary Resuscitation.

AIM: To evaluate key pre-arrest factors and their collective ability to predict post-cardiopulmonary arrest mortality. CPR is often initiated indiscriminately after in-hospital cardiopulmonary arrest. Improved understanding of pre-arrest factors associated with mortality may inform advance care planning. METHODS: A cohort of 49,130 adults who experienced pulseless cardiopulmonary arrest from January 2000 to September 2004 was obtained from 366 US hospitals participating in the National Registry for Cardiopulmonary Resuscitation (NRCPR). Logistic regression with bootstrapping was used to model in-hospital mortality, which included those discharged in unfavorable and severely worsened neurologic state (Cerebral Performance Category >/=3). RESULTS: Overall in-hospital mortality was 84.1%. Advanced age, black race, non-cardiac, non-surgical illness category, pre-existing malignancy, acute stroke, trauma, septicemia, hepatic insufficiency, general floor or Emergency Department location, and pre-arrest use of vasopressors or assisted/mechanical ventilation were independently predictive of in-hospital mortality. Retained peri-arrest factors including cardiac monitoring, and shockable initial pulseless rhythms, were strongly associated with survival. The validation model's AUROC curve (0.77) revealed fair performance. CONCLUSIONS: Predictive pre-resuscitation factors may supplement patient-specific information available at bedside to assist in revising resuscitation plans during the patient's hospitalization.

Factors associated with neurologically intact survival for patients with acute heart failure and in-hospital cardiac arrest.

BACKGROUND: Hospitalized patients with heart failure are at risk for cardiac arrest. The ability to predict who may survive such an event with or without neurological deficit would enhance the information on which patients and providers establish resuscitative preferences. METHODS AND RESULTS: We identified 13 063 adult patients with acute heart failure who had cardiac arrest at 457 hospitals participating in the National Registry of Cardiopulmonary Resuscitation between January 1, 2000 and December 31, 2007. Neurological status was determined on admission and discharge by cerebral performance category with neurologically intact survival (NIS)=cerebral performance category 1 (no) or 2 (moderate dysfunction) and non-NIS=cerebral performance category 3 (severe dysfunction), 4 (coma), or 5 (brain death). Factors available prearrest (demographics, preexisting conditions, and interventions in-place) were assessed for association with NIS using multivariable logistic regression, initially without then with adjustment for arrest-related variables and hospital characteristics. NIS occurred in 2307 patients (17.7%) and was associated by adjusted odds ratio with 18 prearrest factors; 4 positively and 14 negatively. The association (odds ratio; 95% CI) was strongest for 4 specific variables: acute stroke (0.38; 0.25 to 0.58), history of malignancy (0.49; 0.39 to 0.63), vasopressor use (0.50; 0.43 to 0.59), and assisted or mechanical ventilation (0.53; 0.45 to 0.61). CONCLUSIONS: A number of prearrest factors seem to be associated with NIS, the majority inversely. Consideration of these before cardiac arrest could enhance the resuscitative decision-making process for patients with acute heart failure.

The influence of a major disaster on suicide risk in the population.

The authors investigated the relationship between the September 11, 2001 terrorist attacks and suicide risk in New York City from 1990 to 2006. The average monthly suicide rate over the study period was 0.56 per 100,000 people. The monthly rate after September 2001 was 0.11 per 100,000 people lower as compared to the rate in the period before. However, the rate of change in suicide was not significantly different before and after the disaster, and regression discontinuity analysis indicated no change at this date. There was no net change in the suicide rate in New York City attributable to this disaster, suggesting that factors other than exposure to traumatic events (e.g., cultural norms, availability of lethal methods) may be key drivers of suicide risk in this context.

Removing bridge barriers stimulates suicides: an unfortunate natural experiment.

OBJECTIVES: Safety barriers to prevent suicide by jumping were removed from Grafton Bridge in Auckland, New Zealand, in 1996 after having been in place for 60 years. This study compared the number of suicides due to jumping from the bridge after the reinstallation of safety barriers in 2003. METHODS: National mortality data for suicide deaths were compared for three time periods: 1991-1995 (old barrier in place); 1997-2002 (no barriers in place); 2003-2006 (after barriers were reinstated). RESULTS: Removal of barriers was followed by a fivefold increase in the number and rate of suicides from the bridge. These increases led to a decision to reinstall safety barriers. Since the reinstallation of barriers, of an improved design, in 2003, there have been no suicides from the bridge. CONCLUSIONS: This natural experiment, using a powerful a-b-a (reversal) design, shows that safety barriers are effective in preventing suicide: their removal increases suicides; their reinstatement prevents suicides.

Epidemiological trends in psychosis-related Emergency Department visits in the United States, 1992-2001.

Mental health visits represented an increasing fraction of all Emergency Department (ED) visits in the U.S. between 1992 and 2001. This study used the National Hospital Ambulatory Medical Care Survey, a 4-staged probability sample of ED visits from geographically diverse hospitals around the U.S., to assess the contribution of all psychosis-related visits to this overall trend. Unlike other mental-health-related ED visits, the rate of psychosis-related visits did not increase. This lack of change is notable in the context of dramatic changes in both healthcare financing and antipsychotic prescribing practices during this period. There was an unexpected decrease in Medicare-funded psychosis-related ED visits at a time of increasing Medicare enrollment overall. An important demographic trend over this decade was the increasing urbanization of psychosis-related ED visits coincident with a relative decrement in such visits within rural areas.

Mental health and emergency medicine: a research agenda.

The burden of mental illness is profound and growing. Coupled with large gaps in extant psychiatric services, this mental health burden has often forced emergency departments (EDs) to become the de facto primary and acute care provider of mental health care in the United States. An expanded emergency medical and mental health research agenda is required to meet the need for improved education, screening, surveillance, and ED-initiated interventions for mental health problems. As an increasing fraction of undiagnosed and untreated psychiatric patients passes through the revolving doors of U.S. EDs, the opportunities for improving the art and science of acute mental health care have never been greater. These opportunities span macroepidemiologic surveillance research to intervention studies with individual patients. Feasible screening, intervention, and referral programs for mental health patients presenting to general EDs are needed. Additional research is needed to improve the quality of care, including the attitudes, abilities, interests, and virtues of ED providers. Research that optimizes provider education and training can help academic settings validate psychosocial issues as core components and responsibilities of emergency medicine. Transdisciplinary research with federal partners and investigators in neuropsychiatry and related fields can improve the mechanistic understanding of acute mental health problems. To have lasting impact, however, advances in ED mental health care must be translated into real-world policies and sustainable program enhancements to assure the uptake of best practices for ED screening, treatment, and management of mental disorders and psychosocial problems.

Cardiac arrests of hospital staff and visitors: experience from the national registry of cardiopulmonary resuscitation.

AIM OF THE STUDY: Approximately 750,000 in-hospital cardiac arrests occur annually in the United States. Many will occur to visitors or staff members within the hospital's public areas. We sought to provide a descriptive analysis of visitor cardiac arrests in hospitals and to compare survival outcomes to matching inpatient arrests. METHODS: We queried the National Registry of Cardiopulmonary Resuscitation (NRCPR)) for all adult cardiac arrests from January 2000 to May 2006 that occurred to visitors or employees anywhere within the hospital. Visitors were matched to inpatient cardiac arrests from within the same NRCPR database for age, gender, race, prior residence and functional status, and presenting rhythms. The compared outcomes were return of spontaneous circulation (ROSC), survival to 24h (S24), and survival to discharge (SHD). RESULTS: 147 visitors suffered a cardiac arrest during the study period. S24 (48% vs. 37%, p=0.011) and SHD (42% vs. 24%, p<0.0001) were both higher in the visitor cohort. However, ROSC did not significantly differ between visitors and controls (57% vs. 51%). Visitor cardiac arrests occurred in a wide variety of locations. CONCLUSION: Cardiac arrest among hospital visitors is a relatively common event. The survival outcomes of hospital visitors compared unfavorably to that of recently published experience with out-of-hospital cardiac arrest victims.