Optimizing Advanced Imaging of the Pediatric Patient in the Emergency Department: Policy Statement.

Advanced imaging, including ultrasonography, computed tomography, and magnetic resonance imaging (MRI), is an integral component to the evaluation and management of ill and injured children in the emergency department. As with any test or intervention, the benefits and potential impacts on management must be weighed against the risks to ensure that high-value care is being delivered. There are important considerations specific to the pediatric patient related to the ordering and interpretation of advanced imaging. This policy statement provides guidelines for institutions and those who care for children to optimize the use of advanced imaging in the emergency department setting and was coauthored by experts in pediatric and general emergency medicine, pediatric radiology, and pediatric surgery. The intent is to guide decision-making where children may access care.

Optimizing Advanced Imaging of the Pediatric Patient in the Emergency Department: Policy Statement.

Advanced imaging, including ultrasonography, computed tomography, and magnetic resonance imaging, is an integral component to the evaluation and management of ill and injured children in the emergency department. As with any test or intervention, the benefits and potential impacts on management must be weighed against the risks to ensure that high-value care is being delivered. There are important considerations specific to the pediatric patient related to the ordering and interpretation of advanced imaging. This policy statement provides guidelines for institutions and those who care for children to optimize the use of advanced imaging in the emergency department setting and was coauthored by experts in pediatric and general emergency medicine, pediatric radiology, and pediatric surgery. The intent is to guide decision-making where children may access care.

Optimizing Advanced Imaging of the Pediatric Patient in the Emergency Department: Technical Report.

Advanced diagnostic imaging modalities, including ultrasonography, computed tomography, and magnetic resonance imaging, are key components in the evaluation and management of pediatric patients presenting to the emergency department. Advances in imaging technology have led to the availability of faster and more accurate tools to improve patient care. Notwithstanding these advances, it is important for physicians, physician assistants, and nurse practitioners to understand the risks and limitations associated with advanced imaging in children and to limit imaging studies that are considered low value, when possible. This technical report provides a summary of imaging strategies for specific conditions where advanced imaging is commonly considered in the emergency department. As an accompaniment to the policy statement, this document provides resources and strategies to optimize advanced imaging, including clinical decision support mechanisms, teleradiology, shared decision-making, and rationale for deferred imaging for patients who will be transferred for definitive care.

Optimizing Advanced Imaging of the Pediatric Patient in the Emergency Department: Policy Statement.

Advanced imaging, including ultrasonography, computed tomography, and magnetic resonance imaging (MRI), is an integral component to the evaluation and management of ill and injured children in the emergency department. As with any test or intervention, the benefits and potential impacts on management must be weighed against the risks to ensure that high-value care is being delivered. There are important considerations specific to the pediatric patient related to the ordering and interpretation of advanced imaging. This policy statement provides guidelines for institutions and those who care for children to optimize the use of advanced imaging in the emergency department setting and was coauthored by experts in pediatric and general emergency medicine, pediatric radiology, and pediatric surgery. The intent is to guide decision-making where children may access care.

Optimizing Advanced Imaging of the Pediatric Patient in the Emergency Department: Technical Report.

Advanced diagnostic imaging modalities, including ultrasonography, computed tomography, and magnetic resonance imaging (MRI), are key components in the evaluation and management of pediatric patients presenting to the emergency department. Advances in imaging technology have led to the availability of faster and more accurate tools to improve patient care. Notwithstanding these advances, it is important for physicians, physician assistants, and nurse practitioners to understand the risks and limitations associated with advanced imaging in children and to limit imaging studies that are considered low value, when possible. This technical report provides a summary of imaging strategies for specific conditions where advanced imaging is commonly considered in the emergency department. As an accompaniment to the policy statement, this document provides resources and strategies to optimize advanced imaging, including clinical decision support mechanisms, teleradiology, shared decision-making, and rationale for deferred imaging for patients who will be transferred for definitive care.

Emergency Department Evaluation of Young Infants With Head Injury.

OBJECTIVES: We compared the emergency department (ED) evaluation and outcomes of young head-injured infants to older children. METHODS: Using the Pediatric Health Information Systems database, we performed a retrospective, cross-sectional analysis of children <2 years old with isolated head injuries (International Classification of Diseases, 10th Revision, diagnoses) at one of 47 EDs from 2015 to 2019. Our primary outcome was utilization of diagnostic cranial imaging. Secondary outcomes were diagnosis of traumatic brain injury (TBI), clinically important TBI, and mortality. We compared outcomes between the youngest infants (<3 months old) and children 3 to 24 months old. RESULTS: We identified 112 885 ED visits for children <2 years old with isolated head injuries. A total of 62 129 (55%) were by males, and 10 325 (9.1%) were by infants <3 months of age. Compared with older children (12-23 months old), the youngest infants were more likely to: Undergo any diagnostic cranial imaging (50.3% vs 18.3%; difference 31.9%, 95% confidence interval [CI] 35.0-28.9%), be diagnosed with a TBI (17.5% vs 2.7%; difference 14.8%, 95% CI 16.4%-13.2%) or clinically important TBI (4.6% vs 0.5%; difference 4.1%, 95% CI 3.8%-4.5%), and to die (0.3% vs 0.1%; difference 0.2%, 95% CI 0.3%-0.1%). Among those undergoing computed tomography or MRI, TBIs were significantly more common in the youngest infants (26.4% vs 8.8%, difference 17.6%, 95% CI 16.3%-19.0%). CONCLUSIONS: The youngest infants with head injuries are significantly more likely to undergo cranial imaging, be diagnosed with brain injuries, and die, highlighting the need for a specialized approach for this vulnerable population.

Utility of Synovial Fluid Biomarkers for Culture-Positive Septic Arthritis in a Lyme Disease-Endemic Region.

OBJECTIVE: To evaluate the performance of synovial fluid biomarkers to identify children with culture-positive septic arthritis. METHODS: We identified children 6 months to 18 years old presenting to a single emergency department between 2007 and 2022 undergoing evaluation for septic arthritis defined by having a synovial fluid culture obtained. Our primary outcome was septic arthritis defined by a positive synovial fluid culture. We evaluated the ability of synovial fluid biomarkers to identify children with septic arthritis using area under the receiver operating characteristic curve (AUC) analyses. We measured the sensitivity and specificity of commonly used synovial fluid biomarkers. RESULTS: We included 796 children, of whom 79 (10%) had septic arthritis. Compared with synovial white blood cell count (AUC, 0.72; 95% confidence interval [CI], 0.65-0.78), absolute neutrophil count (AUC, 0.72; 95% CI, 0.66-0.79; P = 0.09), percent neutrophils (AUC, 0.66; 95% CI, 0.60-0.71; P = 0.12), and glucose (AUC, 0.78; 95% CI, 0.67-0.90; P = 0.33) performed similarly, whereas protein (AUC, 0.52; 95% CI, 0.40-0.63, P = 0.04) had lower diagnostic accuracy. Synovial fluid white blood cell count ≥50,000 cells/µL had a sensitivity of 62.0% (95% CI, 50.4%-72.7%) and a specificity of 67.0% (95% CI, 63.4%-70.4%), whereas a positive synovial fluid Gram stain had a sensitivity of 48.1% (95% CI, 36.5%-59.7%) and specificity of 99.1% (95% CI, 98.1%-99.7%) for septic arthritis. CONCLUSIONS: None of the routinely available synovial fluid biomarkers had sufficient accuracy to be used in isolation in the identification of children with septic arthritis. New approaches including multivariate clinical prediction rules and novel biomarkers are needed.

Harnessing the Power of Generative AI for Clinical Summaries: Perspectives From Emergency Physicians.

STUDY OBJECTIVE: The workload of clinical documentation contributes to health care costs and professional burnout. The advent of generative artificial intelligence language models presents a promising solution. The perspective of clinicians may contribute to effective and responsible implementation of such tools. This study sought to evaluate 3 uses for generative artificial intelligence for clinical documentation in pediatric emergency medicine, measuring time savings, effort reduction, and physician attitudes and identifying potential risks and barriers. METHODS: This mixed-methods study was performed with 10 pediatric emergency medicine attending physicians from a single pediatric emergency department. Participants were asked to write a supervisory note for 4 clinical scenarios, with varying levels of complexity, twice without any assistance and twice with the assistance of ChatGPT Version 4.0. Participants evaluated 2 additional ChatGPT-generated clinical summaries: a structured handoff and a visit summary for a family written at an 8th grade reading level. Finally, a semistructured interview was performed to assess physicians' perspective on the use of ChatGPT in pediatric emergency medicine. Main outcomes and measures included between subjects' comparisons of the effort and time taken to complete the supervisory note with and without ChatGPT assistance. Effort was measured using a self-reported Likert scale of 0 to 10. Physicians' scoring of and attitude toward the ChatGPT-generated summaries were measured using a 0 to 10 Likert scale and open-ended questions. Summaries were scored for completeness, accuracy, efficiency, readability, and overall satisfaction. A thematic analysis was performed to analyze the content of the open-ended questions and to identify key themes. RESULTS: ChatGPT yielded a 40% reduction in time and a 33% decrease in effort for supervisory notes in intricate cases, with no discernible effect on simpler notes. ChatGPT-generated summaries for structured handoffs and family letters were highly rated, ranging from 7.0 to 9.0 out of 10, and most participants favored their inclusion in clinical practice. However, there were several critical reservations, out of which a set of general recommendations for applying ChatGPT to clinical summaries was formulated. CONCLUSION: Pediatric emergency medicine attendings in our study perceived that ChatGPT can deliver high-quality summaries while saving time and effort in many scenarios, but not all.

Surgical Specialty Consultation for Pediatric Facial Laceration Repair: An American and Canadian Survey.

OBJECTIVE: We sought to describe patterns of and indications for surgical specialty consultation for facial laceration repair in pediatric emergency departments (PEDs). METHODS: We performed a multicenter survey of PED leadership throughout the United States and Canada evaluating the practice patterns of surgical specialty consultation for patients presenting for facial lacerations requiring repair. We measured demographics of PEDs, factors influencing the decision to obtain a surgical specialty consultation, and the presence and components of consultation guidelines. Factors related to consultation were ranked on a Likert scale from 1 to 5 (1 = Not at all important, 5 = Extremely important). We evaluated relationships between reported rates of surgical specialty consultation and PED region, annual PED volume, and reported factors associated with PED consultation. RESULTS: Survey responses were received from 67/124 (54%) queried PEDs. The median self-reported rate of surgical specialty consultation for facial lacerations was 10% and ranged from 1% to 70%, with resident physicians performing the repair 71% of the time a subspecialist was consulted. There was regional variability in specialty consultation, with the highest and lowest rate in the Midwest and Canada, respectively (P = 0.03). The top 4 influential factors prompting consultation with the highest percentage of responses of "Extremely Important" or "Very Important" were: discretion of the physician caring for the patient (95%), parental preference (39%), limited PED resources (32%), and patient requires sedation (32%). Surgical specialty consult guidelines were used in only 6% of PEDs with consensus that depth necessitating more than 2-layer repair or involvement of critical structures should prompt consultation. CONCLUSIONS: Surgical specialty usage in the management of patients who present with facial lacerations to PEDs has significant variation related to patient, provider, and department-level factors that influence the decision to consult. Lack of consult guidelines represent a potential opportunity to standardize care delivery to this common presentation.

A Novel Risk Score to Guide the Evaluation of Acute Hematogenous Osteomyelitis in Children.

OBJECTIVES: To identify independent predictors of and derive a risk score for acute hematogenous osteomyelitis (AHO) in children. METHODS: We conducted a retrospective matched case-control study of children >90 days to <18 years of age undergoing evaluation for a suspected musculoskeletal (MSK) infection from 2017 to 2019 at 23 pediatric emergency departments (EDs) affiliated with the Pediatric Emergency Medicine Collaborative Research Committee. Cases were identified by diagnosis codes and confirmed by chart review to meet accepted diagnostic criteria for AHO. Controls included patients who underwent laboratory and imaging tests to evaluate for a suspected MSK infection and received an alternate final diagnosis. RESULTS: We identified 1135 cases of AHO matched to 2270 controls. Multivariable logistic regression identified 10 clinical and laboratory factors independently associated with AHO. We derived a 4-point risk score for AHO using (1) duration of illness >3 days, (2) history of fever or highest ED temperature ≥38°C, (3) C-reactive protein >2.0 mg/dL, and (4) erythrocyte sedimentation rate >25 mm per hour (area under the curve: 0.892, 95% confidence interval [CI]: 0.881 to 0.901). Choosing to pursue definitive diagnostics for AHO when 3 or more factors are present maximizes diagnostic accuracy at 84% (95% CI: 82% to 85%), whereas children with 0 factors present are highly unlikely to have AHO (sensitivity: 0.99, 95% CI: 0.98 to 1.00). CONCLUSIONS: We identified 10 predictors for AHO in children undergoing evaluation for a suspected MSK infection in the pediatric ED and derived a novel 4-point risk score to guide clinical decision-making.

Patterns and Predictors of Health Care Utilization After Pediatric Concussion: A Retrospective Cohort Study.

OBJECTIVE: To characterize types, duration, and intensity of health care utilization following pediatric concussion and to identify risk factors for increased post-concussion utilization. METHODS: A retrospective cohort study of children 5 to 17 years old diagnosed with acute concussion at a quaternary center pediatric emergency department or network of associated primary care clinics. Index concussion visits were identified using International Classification of Diseases, Tenth Revision, Clinical Modification codes. We analyzed patterns of health care visits 6 months before and after the index visit using interrupted time-series analyses. The primary outcome was prolonged concussion-related utilization, defined as having ≥1 follow-up visits with a concussion diagnosis more than 28 days after the index visit. We used logistic regressions to identify predictors of prolonged concussion-related utilization. RESULTS: Eight hundred nineteen index visits (median [interquartile range] age, 14 [11-16] years; 395 [48.2%] female) were included. There was a spike in utilization during the first 28 days after the index visit compared to the pre-injury period. Premorbid headache/migraine disorder (adjusted odds ratio (aOR) 2.05, 95% confidence interval [CI] 1.09-3.89) and top quartile pre-injury utilization (aOR 1.90, 95% CI 1.02-3.52) predicted prolonged concussion-related utilization. Premorbid depression/anxiety (aOR 1.55, 95% CI 1.31-1.83) and top quartile pre-injury utilization (aOR 2.29, 95% CI 1.95-2.69) predicted increased utilization intensity. CONCLUSIONS: Health care utilization is increased during the first 28 days after pediatric concussion. Children with premorbid headache/migraine disorders, premorbid depression/anxiety, and high baseline utilization are more likely to have increased post-injury health care utilization. This study will inform patient-centered treatment but may be limited by incomplete capture of post-injury utilization and generalizability.

Biomarkers for Pediatric Bacterial Musculoskeletal Infections in Lyme Disease-Endemic Regions.

OBJECTIVES: Bacterial musculoskeletal infections (MSKIs) are challenging to diagnose because of the clinical overlap with other conditions, including Lyme arthritis. We evaluated the performance of blood biomarkers for the diagnosis of MSKIs in Lyme disease-endemic regions. METHODS: We conducted a secondary analysis of a prospective cohort study of children 1 to 21 years old with monoarthritis presenting to 1 of 8 Pedi Lyme Net emergency departments for evaluation of potential Lyme disease. Our primary outcome was an MSKI, which was defined as septic arthritis, osteomyelitis or pyomyositis. We compared the diagnostic accuracy of routinely available biomarkers (absolute neutrophil count, C-reactive protein, erythrocyte sedimentation rate, and procalcitonin) to white blood cells for the identification of an MSKI using the area under the receiver operating characteristic curve (AUC). RESULTS: We identified 1423 children with monoarthritis, of which 82 (5.8%) had an MSKI, 405 (28.5%) Lyme arthritis, and 936 (65.8%) other inflammatory arthritis. When compared with white blood cell count (AUC, 0.63; 95% confidence interval [CI], 0.55-0.71), C-reactive protein (0.84; 95% CI, 0.80-0.89; P < .05), procalcitonin (0.82; 95% CI, 0.77-0.88; P < .05), and erythrocyte sedimentation rate (0.77; 95% CI, 0.71-0.82; P < .05) had higher AUCs, whereas absolute neutrophil count (0.67; 95% CI, 0.61-0.74; P < .11) had a similar AUC. CONCLUSIONS: Commonly available biomarkers can assist in the initial approach to a potential MSKI in a child. However, no single biomarker has high enough accuracy to be used in isolation, especially in Lyme disease-endemic areas.

Comprehensiveness of Testing Among Herpes Simplex Virus Infected Infants: A Multicenter Cohort Study.

Despite clear testing recommendations for herpes simplex virus (HSV) infection in infants, few data exist on the comprehensiveness of HSV testing in practice. In a 23-center study of 112 infants with confirmed HSV disease, less than one-fifth had all recommended testing performed, highlighting the need for increased awareness of and adherence to testing recommendations for this vulnerable population.

A Clinical Prediction Rule for Bacterial Musculoskeletal Infections in Children with Monoarthritis in Lyme Endemic Regions.

STUDY OBJECTIVE: Children with a bacterial musculoskeletal infection (MSKI) require prompt identification and treatment. In Lyme disease endemic areas, children with an MSKI can present similarly to those with Lyme arthritis. Our goal was to derive a clinical prediction rule to accurately identify children at a low risk for an MSKI. METHODS: We enrolled children with monoarthritis presenting to 1 of 6 Pedi Lyme Net centers and performed a procalcitonin (PCT) and a first-tier Lyme C6 enzyme immunoassay (EIA) test. Our primary outcome was an MSKI (septic arthritis, osteomyelitis, or pyomyositis). Using recursive partitioning with k-fold cross validation, we derived a clinical prediction rule to identify children at a low risk of an MSKI. We calculated the accuracy of our novel rule in a derivation cohort. RESULTS: Of the 735 children in the derivation cohort with an available research biosample, 39 (5%) had an MSKI (18 had septic arthritis, 20 had osteomyelitis, and 1 had pyomyositis), 260 (37%) had Lyme arthritis, and 436 (53%) had other inflammatory arthritis. Children with a PCT level of more than or equal to 0.50 ng/mL and those with a C-reactive protein (CRP) level of more than or equal to 0.6 mg/dL with a negative Lyme C6 EIA were classified as not low risk for an MSKI. Of the 451 (61%) children categorized as low risk, none had an MSKI (sensitivity 100%, 95% confidence interval 91.0% to 100%; specificity 74.2%, 95% confidence interval 70.5% to 77.6%). CONCLUSION: A novel clinical decision rule that includes PCT, CRP, and a first-tier Lyme EIA was highly sensitive for MSKIs. Although broader external validation is required, the application of this rule may safely reduce invasive testing, procedures, and treatment for low risk children.

Paediatric post-concussive symptoms: symptom clusters and clinical phenotypes.

OBJECTIVE: To assess the co-occurrence and clustering of post-concussive symptoms in children, and to identify distinct patient phenotypes based on symptom type and severity. METHODS: We performed a secondary analysis of the prospective, multicentre Predicting and Preventing Post-concussive Problems in Pediatrics (5P) cohort study, evaluating children 5-17 years of age presenting within 48 hours of an acute concussion. Our primary outcome was the simultaneous occurrence of two or more persistent post-concussive symptoms on the Post-Concussion Symptom Inventory at 28 days post-injury. Analyses of symptom and patient clusters were performed using hierarchical cluster analyses of symptom severity ratings. RESULTS: 3063 patients from the parent 5P study were included. Median age was 12.1 years (IQR: 9.2-14.6 years), and 1857 (60.6%) were male. Fatigue was the most common persistent symptom (21.7%), with headache the most commonly reported co-occurring symptom among patients with fatigue (55%; 363/662). Headache was common in children reporting any of the 12 other symptoms (range: 54%-72%). Physical symptoms occurred in two distinct clusters: vestibular-ocular and headache. Emotional and cognitive symptoms occurred together more frequently and with higher severity than physical symptoms. Fatigue was more strongly associated with cognitive and emotional symptoms than physical symptoms. We identified five patient groups (resolved/minimal, mild, moderate, severe and profound) based on symptom type and severity. CONCLUSION: Post-concussive symptoms in children occur in distinct clusters, facilitating the identification of distinct patient phenotypes based on symptom type and severity. Care of children post-concussion must be comprehensive, with systems designed to identify and treat distinct post-concussion phenotypes.

Validation of Septic Knee Monoarthritis Prediction Rule in a Lyme Disease Endemic Area.

OBJECTIVE: In Lyme disease endemic areas, Lyme and septic arthritis often present similarly. A published septic knee arthritis clinical prediction rule includes 2 high-risk predictors: absolute neutrophil count of 10,000 cells/mm3 or greater and erythrocyte sedimentation rate of 40 mm/h or greater. The objective of the study was to externally validate this prediction rule in a multicenter prospective cohort. METHODS: We enrolled a prospective cohort of children with knee monoarthritis undergoing evaluation for Lyme disease at 1 of 8 Pedi Lyme Net emergency departments located in endemic areas. We defined a case of septic arthritis with a positive synovial fluid culture or a synovial fluid white blood cell count of 50,000 or greater per high powered field with a positive blood culture and Lyme arthritis with a positive or equivocal C6 EIA, followed by a positive supplemental immunoblot. Other children were classified as having inflammatory arthritis. We report the performance of the septic arthritis clinical prediction rule in our study population. RESULTS: Of the 543 eligible children, 13 had septic arthritis (2.4%), 234 Lyme arthritis (43.1%), and 296 inflammatory arthritis (54.5%). Of the 457 children (84.2%) with available laboratory predictors, all children with septic arthritis were classified as high risk (sensitivity, 100%; 95% confidence interval [CI], 77.2%-100%; specificity, 68.1%; 95% CI, 63.6-73.3; negative predictive value, 278/278 [100%]; 95% CI, 98.6%-100%). Of the 303 low-risk children, 52 (17.2%) underwent diagnostic arthrocentesis. CONCLUSIONS: The septic knee arthritis clinical prediction rule accurately distinguished between septic and Lyme arthritis in an endemic area. Clinical application may reduce unnecessary invasive diagnostic procedures.

Trends in Management of Simple Febrile Seizures at US Children's Hospitals.

OBJECTIVE: We sought to measure trends in evaluation and management of children with simple febrile seizures (SFSs) before and after the American Academy of Pediatrics updated guidelines published in 2011. METHODS: In this retrospective, cross-sectional analysis, we used the Pediatric Health Information System database comprising 49 tertiary care pediatric hospitals in the United States from 2005 to 2019. We included children aged 6 to 60 months with an emergency department visit for first SFS identified using codes from the International Classification of Diseases, Ninth Revision, and International Classification of Diseases 10th Revision. RESULTS: We identified 142 121 children (median age 21 months, 42.4% female) with an emergency department visit for SFS. A total of 49 668 (35.0%) children presented before and 92 453 (65.1%) after the guideline. The rate of lumbar puncture for all ages declined from 11.6% (95% confidence interval [CI], 10.8% to 12.4%) in 2005 to 0.6% (95% CI, 0.5% to 0.8%) in 2019 (P < .001). Similar reductions were noted in rates of head computed tomography (10.6% to 1.6%; P < .001), complete blood cell count (38.8% to 10.9%; P < .001), hospital admission (19.2% to 5.2%; P < .001), and mean costs ($1523 to $601; P < .001). Reductions in all outcomes began before, and continued after, the publication of the American Academy of Pediatrics guideline. There was no significant change in delayed diagnosis of bacterial meningitis (preperiod 2 of 49 668 [0.0040%; 95% CI, 0.00049% to 0.015%], postperiod 3 of 92 453 [0.0032%; 95% CI, 0.00066% to 0.0094%]; P = .99). CONCLUSIONS: Diagnostic testing, hospital admission, and costs decreased over the study period, without a concomitant increase in delayed diagnosis of bacterial meningitis. These data suggest most children with SFSs can be safely managed without lumber puncture or other diagnostic testing.

Predictors of Invasive Herpes Simplex Virus Infection in Young Infants.

OBJECTIVES: To identify independent predictors of and derive a risk score for invasive herpes simplex virus (HSV) infection. METHODS: In this 23-center nested case-control study, we matched 149 infants with HSV to 1340 controls; all were ≤60 days old and had cerebrospinal fluid obtained within 24 hours of presentation or had HSV detected. The primary and secondary outcomes were invasive (disseminated or central nervous system) or any HSV infection, respectively. RESULTS: Of all infants included, 90 (60.4%) had invasive and 59 (39.6%) had skin, eyes, and mouth disease. Predictors independently associated with invasive HSV included younger age (adjusted odds ratio [aOR]: 9.1 [95% confidence interval (CI): 3.4-24.5] <14 and 6.4 [95% CI: 2.3 to 17.8] 14-28 days, respectively, compared with >28 days), prematurity (aOR: 2.3, 95% CI: 1.1 to 5.1), seizure at home (aOR: 6.1, 95% CI: 2.3 to 16.4), ill appearance (aOR: 4.2, 95% CI: 2.0 to 8.4), abnormal triage temperature (aOR: 2.9, 95% CI: 1.6 to 5.3), vesicular rash (aOR: 54.8, (95% CI: 16.6 to 180.9), thrombocytopenia (aOR: 4.4, 95% CI: 1.6 to 12.4), and cerebrospinal fluid pleocytosis (aOR: 3.5, 95% CI: 1.2 to 10.0). These variables were transformed to derive the HSV risk score (point range 0-17). Infants with invasive HSV had a higher median score (6, interquartile range: 4-8) than those without invasive HSV (3, interquartile range: 1.5-4), with an area under the curve for invasive HSV disease of 0.85 (95% CI: 0.80-0.91). When using a cut-point of ≥3, the HSV risk score had a sensitivity of 95.6% (95% CI: 84.9% to 99.5%), specificity of 40.1% (95% CI: 36.8% to 43.6%), and positive likelihood ratio 1.60 (95% CI: 1.5 to 1.7) and negative likelihood ratio 0.11 (95% CI: 0.03 to 0.43). CONCLUSIONS: A novel HSV risk score identified infants at extremely low risk for invasive HSV who may not require routine testing or empirical treatment.

Prediction of patient disposition: comparison of computer and human approaches and a proposed synthesis.

OBJECTIVE: To compare the accuracy of computer versus physician predictions of hospitalization and to explore the potential synergies of hybrid physician-computer models. MATERIALS AND METHODS: A single-center prospective observational study in a tertiary pediatric hospital in Boston, Massachusetts, United States. Nine emergency department (ED) attending physicians participated in the study. Physicians predicted the likelihood of admission for patients in the ED whose hospitalization disposition had not yet been decided. In parallel, a random-forest computer model was developed to predict hospitalizations from the ED, based on data available within the first hour of the ED encounter. The model was tested on the same cohort of patients evaluated by the participating physicians. RESULTS: 198 pediatric patients were considered for inclusion. Six patients were excluded due to incomplete or erroneous physician forms. Of the 192 included patients, 54 (28%) were admitted and 138 (72%) were discharged. The positive predictive value for the prediction of admission was 66% for the clinicians, 73% for the computer model, and 86% for a hybrid model combining the two. To predict admission, physicians relied more heavily on the clinical appearance of the patient, while the computer model relied more heavily on technical data-driven features, such as the rate of prior admissions or distance traveled to hospital. DISCUSSION: Computer-generated predictions of patient disposition were more accurate than clinician-generated predictions. A hybrid prediction model improved accuracy over both individual predictions, highlighting the complementary and synergistic effects of both approaches. CONCLUSION: The integration of computer and clinician predictions can yield improved predictive performance.