Frequency, Type, and Degree of Potential Harm of Adverse Safety Events among Pediatric Emergency Medical Services Encounters.

BACKGROUND: Only 5-10% of emergency medical services (EMS) patients are children, and most pediatric encounters are low-acuity. EMS chart review has been used to identify adverse safety events (ASEs) in high-acuity and high-risk pediatric encounters. The objective of this work was to evaluate the frequency, type, and potential harm of ASEs in varied acuity pediatric EMS encounters. METHODS: This cross-sectional study evaluated pediatric (ages 0-18 years) prehospital records from 15 EMS agencies among three states (Colorado, Connecticut, and Rhode Island) between November 2019 and October 2021. Research associates used a previously validated tool to analyze electronic EMS and hospital records. Adverse safety events were recorded in six care categories, grouped into four levels for analysis: assessment/diagnosis/clinical decision-making, procedures, medication administration (including O2), and fluid administration, and defined across five types of ASEs: Unintended injuries or consequences, Near misses, Suboptimal actions, Errors, and Management complications (UNSEMs). Type and frequency of ASEs in each category were rated in three harm severities: Harm Unlikely, Mild/Temporary, or Permanent/Severe. Three physicians verified ASEs determined by research associates. Frequency of ASEs and harm likelihood are reported. RESULTS: Records for 508 EMS patients were reviewed, with 63 (12.4%) transported using lights and sirens. At least one clinical intervention beyond assessment/diagnosis/clinical decision-making was documented for 183 (36.1%, 95% CI: 31.8, 40.4) patients. A total of 162 ASEs were identified for 112 patients (22.1%, 95% CI: 18.5, 25.7). Suboptimal actions were the most frequent UNSEM (n = 66, 40.7%; 95% CI: 33.1, 48.3). For ASEs, (n = 162), the most frequent associations were with procedures 39.5% (95% CI: 32.0, 47.0) or assessment/diagnosis/clinical decision making, 32.1%, (95% CI: 24.9, 39.3). Among care categories, fluid administration was associated with significantly more UNSEMs (58.1%, 95% CI:53.8, 62.4). Most ASEs were determined to be 'Harm Unlikely' 62.4% (95% CI: 54.4, 70.4), with assessment/diagnosis/clinical decision making having significantly fewer ASEs with documented harm (22.4%, 95% CI: 10.7, 34.1) compared to other care categories. CONCLUSION: Over 20% of pediatric EMS encounters had an identified ASE, and most were unlikely to cause harm. Most frequent ASEs were likely to be associated with procedures and assessment/diagnosis/clinical decision-making.

An Evaluation of Prehospital Adenosine Use.

BACKGROUND: Adenosine has been safely used by paramedics for the treatment of stable supraventricular tachycardia since the mid-1990s. However, there continues to be variability in paramedics' ability to identify appropriate indications for adenosine administration. As the first of a planned series of studies aimed at improving the accuracy of SVT diagnosis and successful administration of adenosine by paramedics, this study details the current usage patterns of adenosine by paramedics. METHODS: This cross-sectional retrospective study investigated adenosine use within a large northeast EMS region from January 1, 2019, through September 30, 2021. Excluding pediatric and duplicate case reports, we created a dataset containing patient age, sex, and vital signs before, during, and after adenosine administration; intravenous line location; and coded medical history from paramedic narrative documentation, including a history of atrial fibrillation, suspected arrhythmia diagnosis, and effect of adenosine. In cases with available prehospital electrocardiograms (EKGs) for review, two physicians independently coded the arrhythmia diagnosis and outcome of adenosine administration. Statistical analysis included interrater reliability with Cohen's kappa statistic. RESULTS: One hundred eighty-three cases were included for final analysis, 84 did not have a documented EKG for review. Categorization of presenting rhythms in these cases occurred by a physician reviewing EMS narrative and documentation. Forty of these 84 cases (48%) were adjudicated as SVT likely, 32 (38%) as SVT unlikely and 12 (14%) as uncategorized due to lack of supporting documentation. Of the 99 cases with EKGs available to review, there was substantial agreement of arrhythmia diagnosis interpretation between physician reviewers (Cohen's kappa 0.77-1.0); 54 cases were adjudicated as SVT by two physician reviewers. Other identified cardiac rhythms included atrial fibrillation (16), sinus tachycardia (11), and ventricular tachycardia (2). Adenosine cardioversion occurred in 47 of the 99 cases with EKGs available for physician review (47.5%). Adenosine cardioversion was also deemed to occur in 87% (47/54) of cases when the EKG rhythm was physician adjudicated SVT. CONCLUSIONS: This study supports the use of adenosine as a prehospital treatment for SVT while highlighting the need for continued efforts to improve paramedics' identification and management of tachyarrhythmias.

Emergency Medical Services Provider Acceptance of and Attitudes About Pediatric SimBox Simulations.

OBJECTIVE: SimBox simulations allow for high-frequency open-access health care education, overcoming cost and resource barriers. Prehospital paramedics and emergency medical technician (EMT) care for children infrequently. In this study, prehospital providers evaluated pediatric SimBox simulations. METHODS: This was a cross-sectional study of EMS professionals participating in a series of simulations conducted in a larger project assessing improvement of the quality of pediatric care in the prehospital setting. Participants were teams of two, which comprised a paramedic/paramedic, paramedic/EMT, or 2 EMTs. The simulations used facilitator resources, debriefing prompts, video depictions of patients and vital signs, and a low-fidelity manikin. Pediatric emergency care coordinators, EMS training officers, and/or emergency physicians facilitated simulations of seizure, sepsis with respiratory failure, and child abuse, followed by debriefings. Participants completed an online survey after the simulation and rated it in 4 domains: prebriefing, scenario content, debriefing, and overall. Ratings were trifold: "strongly agree," "somewhat agree," or "do not agree." Data were analyzed by case type, participant type, location, participant reaction to simulation elements, and the debriefing. Net Promoter Scores were calculated to assess participant endorsement of SimBox. RESULTS: There were 121 participants: 103 (87%) were paramedics, and 18 (13%) were EMTs. Participant agreement of simulation benefit for clinical practice was high, for example, "I am more confident in my ability to prioritize care and interventions" (98.4% strongly or somewhat agree), and 99.2% of participants agreed the postsimulation debriefing with facilitators "provided opportunities to self-reflect on my performance during simulation." Overall, 97.5% strongly or somewhat agreed that the simulations "improved my comfort in pediatric acute care." Net Promoter Score showed 65.3% were promoters of and 24% were passive about SimBox. CONCLUSION: SimBox simulations are associated with improved self-efficacy of prehospital care providers for care of acutely ill or injured children. The majority promotes SimBox as a learning tool.

Modified Delphi Method Derivation of the FAMILY (Family Assessment of Medical Interventions & Liaisons with the Young) EMS Instrument.

BACKGROUND: Though family satisfaction with prehospital care is a surrogate for quality and patient outcomes, there are no tools available to measure family satisfaction. OBJECTIVE: To develop the EMS Family Assessment of Medical Interventions & Liaisons with the Young (FAMILY) instrument. METHODS: Components of family experiences with pediatric prehospital care were identified with a modified Delphi method. The expert panel included Emergency Medical Technicians, paramedics, family representatives, and EMS leaders from Colorado, Connecticut, and Rhode Island. An online survey was used to assess proposed questions from each of five candidate domains from national guidelines, including Safety, Communication, Family Presence, Cultural Awareness, Children with Special Healthcare Needs and Overall Satisfaction. Round-1 items were scored on a five-point Likert scale. Inclusion in the final instrument required 70% agreement ranking items as "include" or "definitely include." In Round-2, participants assessed proposed refinements. This resulted in FAMILY Version-1, with sections for family members and EMS care providers. EMSC Family Action Network (FAN) representatives evaluated the FAMILY, leading to Version-2. Suggestions from the national FAN about content, clarity, and whether the instrument captured their experiences with pediatric EMS care led to the final FAMILY version. Bilingual speakers translated the instrument into Spanish, while assessing the content for semantic, idiomatic, experiential, and conceptual equivalence between the English and Spanish versions. RESULTS: There were 22 experts in Round-1, and 20 continued into Round-2 .The Delphi process yielded 12 questions in six domains with 14 recommended modifications. Two questions were excluded. Five domains reached 70% agreement in Round-1. Cultural Awareness reached 75% agreement after Round-2. Six FAN representatives evaluated Version-1, leading to changes for clarity, content and cultural sensitivity. Seventeen FAN representatives evaluated Version-2 leading to additional refinement. The assessment of the equivalence between the English and Spanish survey versions resulted in changes in the Spanish language content for equivalent meaning. CONCLUSION: A panel of EMS and family stakeholders successfully developed an instrument to assess family satisfaction with pediatric EMS care. Further validation is required in a large respondent population. Assessing family satisfaction with pediatric EMS encounters is an important step toward improving prehospital care.

Medication Dosing Safety for Pediatric Patients: Recognizing Gaps, Safety Threats, and Best Practices in the Emergency Medical Services Setting. A Position Statement and Resource Document from NAEMSP.

BACKGROUND: Millions of patients receive medications in the Emergency Medical Services (EMS) setting annually, and dosing safety is critically important. The need for weight-based dosing in pediatric patients and variability in medication concentrations available in the EMS setting may require EMS providers to perform complex calculations to derive the appropriate dose to deliver. These factors can significantly increase the risk for harm when dose calculations are inaccurate or incorrect. METHODS: We conducted a scoping review of the EMS, interfacility transport and emergency medicine literature regarding pediatric medication dosing safety. A priori, the authors identified four research topics: (1) what are the greatest safety threats that result in significant dosing errors that potentially result in harm to patients, (2) what practices or technologies are known to enhance dosing safety, (3) can data from other settings be extrapolated to the EMS environment to inform dosing safety, and (4) what impact could standardization of medication formularies have on enhancing dosing safety. To address these topics, 17 PICO (Patient, Intervention, Comparison, Outcome) questions were developed and a literature search was performed. RESULTS: After applying exclusion criteria, 70 articles were reviewed. The methods for the investigation, findings from these articles and how they inform EMS medication dosing safety are summarized here. This review yielded 11 recommendations to improve safety of medication delivery in the EMS setting. CONCLUSION: These recommendations are summarized in the National Association of EMS Physicians® position statement: Medication Dosing Safety for Pediatric Patients in Emergency Medical Services.

Sound and Air: Ultrasonographic Measurements of Pediatric Chest Wall Thickness and Implications for Needle Decompression of Tension Pneumothorax.

OBJECTIVES: Needle decompression is potentially life-saving in cases of tension pneumothorax. Although Advanced Trauma Life Support recommends an 8-cm needle for decompression for adults, no detailed pediatric guidelines exist, specifically regarding needle length or site of decompression. METHODS: Point-of-care ultrasound was used to measure chest wall thickness (CWT), the distance between skin and pleural line, bilaterally at the second intercostal midclavicular line and the fourth intercostal anterior axillary line in children of various ages and sizes. Patients were grouped based on Broselow tape weight categories. Measurements were compared between left versus right sides at the 2 anatomic sites. Interclass correlation coefficients were calculated to assess for interrater reliability. RESULTS: A convenience sample of 163 patients from our emergency department was enrolled. For patients who fit into Broselow tape categories, CWT at the second intercostal midclavicular line ranged from 1.11 to 1.91 cm and at the fourth intercostal anterior axillary line ranged from 1.13 to 1.92 cm. In patients larger than the largest Broselow category, 77% had a CWT less than the length of a standard 1.25-in (3.175 cm) catheter. There were no significant differences in the measurements of CWT based on laterality nor anatomic site. CONCLUSIONS: The standard 1.25-in (3.175 cm) catheters are sufficient to treat most tension pneumothoraces in pediatric patients.

Canadian Association of Radiologists White Paper on De-Identification of Medical Imaging: Part 1, General Principles.

The application of big data, radiomics, machine learning, and artificial intelligence (AI) algorithms in radiology requires access to large data sets containing personal health information. Because machine learning projects often require collaboration between different sites or data transfer to a third party, precautions are required to safeguard patient privacy. Safety measures are required to prevent inadvertent access to and transfer of identifiable information. The Canadian Association of Radiologists (CAR) is the national voice of radiology committed to promoting the highest standards in patient-centered imaging, lifelong learning, and research. The CAR has created an AI Ethical and Legal standing committee with the mandate to guide the medical imaging community in terms of best practices in data management, access to health care data, de-identification, and accountability practices. Part 1 of this article will inform CAR members on principles of de-identification, pseudonymization, encryption, direct and indirect identifiers, k-anonymization, risks of reidentification, implementations, data set release models, and validation of AI algorithms, with a view to developing appropriate standards to safeguard patient information effectively.

Canadian Association of Radiologists White Paper on De-identification of Medical Imaging: Part 2, Practical Considerations.

The application of big data, radiomics, machine learning, and artificial intelligence (AI) algorithms in radiology requires access to large data sets containing personal health information. Because machine learning projects often require collaboration between different sites or data transfer to a third party, precautions are required to safeguard patient privacy. Safety measures are required to prevent inadvertent access to and transfer of identifiable information. The Canadian Association of Radiologists (CAR) is the national voice of radiology committed to promoting the highest standards in patient-centered imaging, lifelong learning, and research. The CAR has created an AI Ethical and Legal standing committee with the mandate to guide the medical imaging community in terms of best practices in data management, access to health care data, de-identification, and accountability practices. Part 2 of this article will inform CAR members on the practical aspects of medical imaging de-identification, strengths and limitations of de-identification approaches, list of de-identification software and tools available, and perspectives on future directions.

Comparative Effectiveness of Analgesics to Reduce Acute Pain in the Prehospital Setting.

Objectives: The objectives of this study were to assess comparative effectiveness and harms of opioid and nonopioid analgesics for the treatment of moderate to severe acute pain in the prehospital setting. Methods: We searched MEDLINE®, Embase®, and Cochrane Central from the earliest date through May 9, 2019. Two investigators screened abstracts, reviewed full-text files, abstracted data, and assessed study level risk of bias. We performed meta-analyses when appropriate. Conclusions were made with consideration of established clinically important differences and we graded each conclusion's strength of evidence (SOE). Results: We included 52 randomized controlled trials and 13 observational studies. Due to the absence or insufficiency of prehospital evidence we based conclusions for initial analgesia on indirect evidence from the emergency department setting. As initial analgesics, there is no evidence of a clinically important difference in the change of pain scores with opioids vs. ketamine administered primarily intravenously (IV) (low SOE), IV acetaminophen (APAP) (low SOE), or nonsteroidal anti-inflammatory drugs (NSAIDs) administered primarily IV (moderate SOE). The combined use of an opioid and ketamine, administered primarily IV, may reduce pain more than an opioid alone at 15 and 30 minutes (low SOE). Opioids may cause fewer adverse events than ketamine (low SOE) when primarily administered intranasally. Opioids cause less dizziness than ketamine (low SOE) but may increase the risk of respiratory depression compared with ketamine (low SOE), primarily administered IV. Opioids cause more dizziness (moderate SOE) and may cause more adverse events than APAP (low SOE), both administered IV, but there is no evidence of a clinically important difference in hypotension (low SOE). Opioids may cause more adverse events and more drowsiness than NSAIDs (low SOE), both administered primarily IV. Conclusions: As initial analgesia, opioids are no different than ketamine, APAP, and NSAIDs in reducing acute pain in the prehospital setting. Opioids may cause fewer total side effects than ketamine, but more than APAP or NSAIDs. Combining an opioid and ketamine may reduce acute pain more than an opioid alone but comparative harms are uncertain. When initial morphine is inadequate, giving ketamine may provide greater and quicker acute pain relief than giving additional morphine, although comparative harms are uncertain. Due to indirectness, strength of evidence is generally low, and future research in the prehospital setting is needed.

Automated External Defibrillators in High Schools: Disparities Persist Despite Legislation.

OBJECTIVES: Automated external defibrillators (AEDs) have demonstrated increased survival in out-of-hospital cardiac arrest, and their prevalence continues to rise. In 2009, Connecticut passed a legislation requiring all schools to have an AED, barring financial barriers. The objectives of this study were (1) to determine if this legislation was associated with an increase in Connecticut high school AEDs and (2) to detect disparities in the availability of AEDs based on school type, student demographics, and school size. STUDY DESIGN: A single researcher conducted a scripted telephone survey of all 54 public and 13 private high schools in New Haven County, Connecticut. RESULTS: A response rate of 100% was achieved. Forty-nine percent of high schools had an AED before the legislation, compared with 88% after (P < 0.001). Before legislation, private schools had a higher percentage of AEDs than public schools (69% vs 44%; P = 0.1). Postlegislation, the difference is less (92% vs 87%; P = 0.4). Small schools (<400 students) are significantly less likely to have an AED than larger schools (40% vs 100%; P < 0.001). Schools with a higher percentage of students with disabilities are also less likely to have an AED (P = 0.005), even when controlling for school size (P = 0.03). CONCLUSIONS: State legislation requiring schools to have an AED, if financially feasible, was associated with a significant increase in AED presence among New Haven County high schools. Small high schools and those with a higher percentage of students with disabilities remain less likely to have an AED despite legislation.

COVID-19: Transatlantic Declines in Pediatric Emergency Admissions.

INTRODUCTION: This cross-sectional study looked at the impact of the SARS-CoV-2/COVID-19 pandemic on pediatric emergency department (PED) attendances and admissions (as a proxy for severity of illness) in the United States and United Kingdom. METHODS: Data were extracted for children and adolescents, younger than 16 years, attending Royal Manchester Children's Hospital (RMCH, United Kingdom), and Yale New Haven Children's Hospital (YNHCH, United States). Attendances for weeks 1 to 20 of 2020 and 2019 were compared, and likelihood of admission was assessed via calculation of odds ratios, using week 13 (lockdown) as a cutoff. RESULTS: Attendance numbers for each PED decreased in 2020 compared with 2019 (RMCH, 29.2%; YNHCH, 24.8%). Odds of admission were significantly higher after lockdown than in 2019-RMCH (odds ratio, 1.26; 95% confidence interval, 1.08-1.46) and YNHCH (odds ratio, 1.60; 95% confidence interval, 1.31-1.98). CONCLUSIONS: Although the absolute numbers of children and adolescents attending the PED and being admitted decreased after lockdown, the acuity of illness of those attending appears to be higher.

The RSNA Pediatric Bone Age Machine Learning Challenge.

Purpose The Radiological Society of North America (RSNA) Pediatric Bone Age Machine Learning Challenge was created to show an application of machine learning (ML) and artificial intelligence (AI) in medical imaging, promote collaboration to catalyze AI model creation, and identify innovators in medical imaging. Materials and Methods The goal of this challenge was to solicit individuals and teams to create an algorithm or model using ML techniques that would accurately determine skeletal age in a curated data set of pediatric hand radiographs. The primary evaluation measure was the mean absolute distance (MAD) in months, which was calculated as the mean of the absolute values of the difference between the model estimates and those of the reference standard, bone age. Results A data set consisting of 14 236 hand radiographs (12 611 training set, 1425 validation set, 200 test set) was made available to registered challenge participants. A total of 260 individuals or teams registered on the Challenge website. A total of 105 submissions were uploaded from 48 unique users during the training, validation, and test phases. Almost all methods used deep neural network techniques based on one or more convolutional neural networks (CNNs). The best five results based on MAD were 4.2, 4.4, 4.4, 4.5, and 4.5 months, respectively. Conclusion The RSNA Pediatric Bone Age Machine Learning Challenge showed how a coordinated approach to solving a medical imaging problem can be successfully conducted. Future ML challenges will catalyze collaboration and development of ML tools and methods that can potentially improve diagnostic accuracy and patient care. © RSNA, 2018 Online supplemental material is available for this article. See also the editorial by Siegel in this issue.

GPS Devices in a Simulated Mass Casualty Event.

OBJECTIVE: The aim of this study was to assess the staff perception of a global positioning system (GPS) as a patient tracking tool at an emergency department (ED) receiving patients from a simulated mass casualty event. METHODS: During a regional airport disaster drill a plane crash with 46 pediatric patients was simulated. Personnel from airport fire, municipal fire, law enforcement, emergency medical services, and emergency medicine departments were present. Twenty of the 46 patient actors required transport for medical evaluation, and we affixed GPS devices to 12 of these actors. At the hospital, ED staff including attending physicians, fellows and nurses working in the ED during the time of the drill accessed a map through an application that provided real-time geolocation of these devices. The primary outcome was staff reception of the GPS device as assessed via Likert scale survey after the event. The secondary outcomes were free text feedback from staff and event debriefing observations. RESULTS: Queried registered nurses, attending physicians, and pediatric emergency medicine fellows perceived the GPS device as an advantage for patient care during a disaster. The GPS device allowed multiple-screen real-time tracking and improved situational awareness in cases with and without EMS radio communication prior to arrival at the hospital. CONCLUSION: ED staff reported that the use of GPS trackers in a disaster improved real-time tracking and could potentially improve patient management during a mass casualty event.

Correlation Between Paramedic Disaster Triage Accuracy in Screen-Based Simulations and Immersive Simulations.

Background: Disaster triage is an infrequent, high-stakes skill set used by emergency medical services (EMS) personnel. Screen-based simulation (SBS) provides easy access to asynchronous disaster triage education. However, it is unclear if the performance during a SBS correlates with immersive simulation performance. Methods: This was a nested cohort study within a randomized controlled trial (RCT). The RCT compared triage accuracy of paramedics and emergency medical technicians (EMTs) who completed an immersive simulation of a school shooting, interacted with an SBS for 13 weeks, and then completed the immersive simulation again. The participants were divided into two groups: those exposed vs. those not exposed to 60 Seconds to Survival© (60S), a disaster triage SBS. The aim of the study was to measure the correlation between SBS triage accuracy and immersive simulation triage accuracy. Improvements in triage accuracy were compared among participants in the nested study before and after interacting with 60S, and with improvements in triage accuracy in a previous study in which immersive simulations were used as an educational intervention. Results: Thirty-nine participants completed the SBS; 26 (67%) completed at least three game plays and were included in the evaluation of outcomes of interest. The mean number of plays was 8.5 (SD =7.4). Subjects correctly triaged 12.4% more patients in the immersive simulation at study completion (73.1% before, 85.8% after, P = 0.004). There was no correlation between the amount of improvement in overall SBS triage accuracy, instances of overtriage (P = 0.101), instances of undertriage (P = 0.523), and improvement in the second immersive simulation. A comparison of the pooled data from a previous immersive simulation study with the nested cohort data showed similar improvement in triage accuracy (P = 0.079). Conclusions: SBS education was associated with a significant increase in triage accuracy in an immersive simulation, although triage accuracy demonstrated in the SBS did not correlate with the performance in the immersive simulation. This improvement in accuracy was similar to the improvement seen when immersive simulation was used as the educational intervention in a previous study.

Emergency Medical Responders and Adolescents With Autism Spectrum Disorder.

OBJECTIVES: Because of the high prevalence of Autism Spectrum Disorder (ASD) and wandering behavior, emergency medical responders (EMRs) will likely encounter children and adolescents with ASD. The objectives were to describe interactions between EMRs and children and adolescents with ASD, to evaluate EMRs' ability to recognize ASD in a simulated trauma setting, and to determine if EMRs' demographic characteristics affected their interactions with ASD youth. METHODS: A study of 75 videos of a simulated school bus crash was performed. The simulation included an adolescent with ASD portrayed by an actor. Videos were coded based on 5 domains: (1) reassurance attempts by the EMR, (2) quality of the EMR's interactions, (3) EMR's elicitation of information, (4) EMR's interactions with others, and (5) EMR's recognition of a disability. Two clinicians coded the videos independently, and consensus was reached for any areas of disagreement. RESULTS: Of 75 interactions, 27% provided reassurance to the adolescent with ASD, 1% elicited information, 11% asked bystanders for information or assistance, and 35% suggested a disability with 13% considering ASD. No differences across domains were found based on the EMR's sex. Emergency medical responders with greater than or equal to 5 years of experience were significantly more likely to elicit information than those with less than 5 years of experience, and paramedics had significantly higher total performance scores than paramedic students or those with EMT-Basic. CONCLUSIONS: Few EMRs in this study optimally interacted with adolescents with ASD or recognized a disability. These findings suggest a strong need for targeted educational interventions.

Canadian Association of Radiologists White Paper on Ethical and Legal Issues Related to Artificial Intelligence in Radiology.

Artificial intelligence (AI) software that analyzes medical images is becoming increasingly prevalent. Unlike earlier generations of AI software, which relied on expert knowledge to identify imaging features, machine learning approaches automatically learn to recognize these features. However, the promise of accurate personalized medicine can only be fulfilled with access to large quantities of medical data from patients. This data could be used for purposes such as predicting disease, diagnosis, treatment optimization, and prognostication. Radiology is positioned to lead development and implementation of AI algorithms and to manage the associated ethical and legal challenges. This white paper from the Canadian Association of Radiologists provides a framework for study of the legal and ethical issues related to AI in medical imaging, related to patient data (privacy, confidentiality, ownership, and sharing); algorithms (levels of autonomy, liability, and jurisprudence); practice (best practices and current legal framework); and finally, opportunities in AI from the perspective of a universal health care system.

Canadian Association of Radiologists White Paper on Artificial Intelligence in Radiology.

Artificial intelligence (AI) is rapidly moving from an experimental phase to an implementation phase in many fields, including medicine. The combination of improved availability of large datasets, increasing computing power, and advances in learning algorithms has created major performance breakthroughs in the development of AI applications. In the last 5 years, AI techniques known as deep learning have delivered rapidly improving performance in image recognition, caption generation, and speech recognition. Radiology, in particular, is a prime candidate for early adoption of these techniques. It is anticipated that the implementation of AI in radiology over the next decade will significantly improve the quality, value, and depth of radiology's contribution to patient care and population health, and will revolutionize radiologists' workflows. The Canadian Association of Radiologists (CAR) is the national voice of radiology committed to promoting the highest standards in patient-centered imaging, lifelong learning, and research. The CAR has created an AI working group with the mandate to discuss and deliberate on practice, policy, and patient care issues related to the introduction and implementation of AI in imaging. This white paper provides recommendations for the CAR derived from deliberations between members of the AI working group. This white paper on AI in radiology will inform CAR members and policymakers on key terminology, educational needs of members, research and development, partnerships, potential clinical applications, implementation, structure and governance, role of radiologists, and potential impact of AI on radiology in Canada.

Pediatric Disaster Triage: Multiple Simulation Curriculum Improves Prehospital Care Providers' Assessment Skills.

OBJECTIVE: Paramedics and emergency medical technicians (EMTs) triage pediatric disaster victims infrequently. The objective of this study was to measure the effect of a multiple-patient, multiple-simulation curriculum on accuracy of pediatric disaster triage (PDT). METHODS: Paramedics, paramedic students, and EMTs from three sites were enrolled. Triage accuracy was measured three times (Time 0, Time 1 [two weeks later], and Time 2 [6 months later]) during a disaster simulation, in which high and low fidelity manikins and actors portrayed 10 victims. Accuracy was determined by participant triage decision concordance with predetermined expected triage level (RED [Immediate], YELLOW [Delayed], GREEN [Ambulatory], BLACK [Deceased]) for each victim. Between Time 0 and Time 1, participants completed an interactive online module, and after each simulation there was an individual debriefing. Associations between participant level of training, years of experience, and enrollment site were determined, as were instances of the most dangerous mistriage, when RED and YELLOW victims were triaged BLACK. RESULTS: The study enrolled 331 participants, and the analysis included 261 (78.9%) participants who completed the study, 123 from the Connecticut site, 83 from Rhode Island, and 55 from Massachusetts. Triage accuracy improved significantly from Time 0 to Time 1, after the educational interventions (first simulation with debriefing, and an interactive online module), with a median 10% overall improvement (p < 0.001). Subgroup analyses showed between Time 0 and Time 1, paramedics and paramedic students improved more than EMTs (p = 0.002). Analysis of triage accuracy showed greatest improvement in overall accuracy for YELLOW triage patients (Time 0 50% accurate, Time1 100%), followed by RED patients (Time 0 80%, Time 1 100%). There was no significant difference in accuracy between Time 1 and Time 2 (p = 0.073). CONCLUSION: This study shows that the multiple-victim, multiple-simulation curriculum yields a durable 10% improvement in simulated triage accuracy. Future iterations of the curriculum can target greater improvements in EMT triage accuracy.

Pediatric Disaster Triage System Utilization Across the United States.

OBJECTIVES: The study goal was to determine which pediatric disaster triage (PDT) systems are used in US states/territories and whether there is standardization to their use. Secondary goals were to understand user satisfaction with each system, user preferences, and the nature and magnitude of incidents for which the systems are activated. METHODS: A survey was developed regarding PDT systems used in each state/territory, satisfaction with those used, preference for specific systems, and type and magnitude of incidents prompting system activation. The survey was distributed to emergency medical services for children leads in each state/territory. RESULTS: Eighty-six percent of states/territories responded. Eighty-eight percent of respondents used some formal PDT system, 50% of whom reported utilization of multiple systems. JumpSTART was most commonly used, most often in conjunction with other systems. Of formal systems, JumpSTART has been in use the longest. JumpSTART was also preferred by 71% of those stating a preference; it tied with Smart for median satisfaction level. Although types of incidents prompting system activation was similar across responding states/territories, number of patients prompting activation varied from 1 to 3 to greater than 20, median range of 4 to 7. CONCLUSIONS: Most states/territories use some formal PDT system; few have 1 standardized approach. JumpSTART is predominantly used and is preferred by most respondents. With all systems, there is marked variation in number of patients prompting activation although the reported nature of incidents prompting activation is similar.

Development of a Pediatric Mass Casualty Triage Algorithm Validation Tool.

BACKGROUND: Rapid, accurate evaluation and sorting of victims in a mass casualty incident (MCI) is crucial, as over-triage of victims may overwhelm a trauma system and under-triage may lead to an increase in morbidity and mortality. At this time, there is no validation tool specifically developed for the pediatric population to test an MCI algorithm's inherent capabilities to correctly triage children. OBJECTIVE: To develop a set of criteria for outcomes and interventions to be used as a validation tool for testing an MCI algorithm's ability to correctly triage patients from a cohort of pediatric trauma patients. METHODS: Expert opinion and literature review was used to formulate an initial Criteria Outcomes Tool (COT) that retrospectively categorizes pediatric (≤14 years of age) MCI victims based on resource utilization and clinical outcomes using the classic Red to Black MCI triage designations: Red - cardiopulmonary or mental status compromise needing intervention, Yellow - stable cardiopulmonary status but may require life or limb therapy, Green - minimally injured, and Black - deceased or likely to die given the circumstances. Using an anatomic approach, a list of criteria were defined and a modified-Delphi approach was used to create a summative COT that was reviewed by the American Academy of Pediatrics Disaster Preparedness Advisory Council. The resulting COT was independently applied to a weighted retrospective cohort of 25 pediatric victims from a single Level I trauma center by two reviewers to determine reproducibility. RESULTS: We created a Criteria Outcomes Tool (COT) with 47 outcomes and interventions to validate an MCI algorithm's triage designation. When the COT was applied to a cohort of 25 weighted pediatric charts, we identified the following resource utilization and outcome based triage designations: six Red, six Yellow, six Green, and seven Black triage outcomes. The 100% agreement was obtained between the two reviewers in each of the four categories. CONCLUSIONS: We designed an outcomes and resource utilization tool, the COT, to evaluate the ability of an MCI algorithm to correctly triage pediatric patients. Our tool has good reproducibility on initial study. KEY WORDS: pediatric; disaster; validation tools; triage algorithms; emergency.