Search

ObjectiveTo determine the interobserver agreement of history and physical examination findings in children undergoing evaluation in the emergency department (ED) for headaches.Study designWe conducted a prospective, cross-sectional study of children aged 2-17years evaluated at 3 tertiary-care pediatric EDs for non-traumatic headaches. Two clinicians independently completed a standardized assessment of each child and documented the presence or absence of history and physical examination variables. Unweighted κ statistics were determined for 68 history and 24 physical examination variables.ResultsWe analyzed 191 paired observations; median age was 12years, with 19 (9.9%) children younger than 7years. Interrater reliability was at least moderate (κ≥0.41) for 41 (60.3%) patient history variables. Eleven (61.1%) of 18 physical examination variables for which κ statistics could be calculated had a κ that was at least moderate.ConclusionsA substantial number of history and physical examination findings demonstrated at least moderate κ statistic values when assessed in children with headaches in the ED. These variables may be generalizable across different types of clinicians for evaluation of children with headaches. If also found to predict the presence or absence of emergent intracranial abnormalities, the more reliable clinical findings may be helpful in the development of clinical prediction rules or risk stratification models that could be used across settings for children with headaches.

Objective To identify self-reported pain scores that best represent categories of no pain, mild, moderate, and severe pain in children, and a pain score that accurately represents a child’s perceived need for medication, that is, a minimum pain score at which a child would want an analgesic. Study design Prospective cross-sectional cohort study of children aged 6-17 years presenting to a pediatric emergency department with painful and nonpainful conditions. Pain was measured using the 10-point Verbal Numerical Rating Scale. Receiver operating characteristic –based methodology was used to determine pain scores that best differentiated no pain from mild pain, mild pain from moderate pain, and moderate pain from severe pain. Descriptive statistics were used to determine the perceived need for medication. Results We analyzed data from 548 children (51.3% female, 61.9% with a painful condition). The scores that best represent categories of pain intensity are as follows: 0-1 for no pain; 2-5 for mild pain; 6-7 for moderate pain; and 8-10 for severe pain. The area under the curve for the cut points differentiating each category ranged from 0.76 to 0.88. The median pain score representing the perceived need for medication was 6 (IQR, 4-7; range, 0-10). Conclusions We identified population-level self-reported pain scores in children associated with categories of pain intensity that differ from scores conventionally used. Implementing our findings may provide a more accurate representation of the clinical meaning of pain scores and reduce selection bias in research. Our findings do not support the use of pain scores in isolation for clinical decision making or the use of a pain score threshold to represent a child’s perceived need for medication.

Objective To determine the risk of serious bacterial infections (SBIs) in young febrile infants with and without viral infections. Study design Planned secondary analyses of a prospective observational study of febrile infants 60 days of age or younger evaluated at 1 of 26 emergency departments who did not have clinical sepsis or an identifiable site of bacterial infection. We compared patient demographics, clinical, and laboratory findings, and prevalence of SBIs between virus-positive and virus-negative infants. Results Of the 4778 enrolled infants, 2945 (61.6%) had viral testing performed, of whom 1200 (48.1%) were virus positive; 44 of the 1200 had SBIs (3.7%; 95% CI, 2.7%-4.9%). Of the 1745 virus-negative infants, 222 had SBIs (12.7%; 95% CI, 11.2%-14.4%). Rates of specific SBIs in the virus-positive group vs the virus-negative group were: UTIs (33 of 1200 [2.8%; 95% CI, 1.9%-3.8%] vs 186 of 1745 [10.7%; 95% CI, 9.2%-12.2%]) and bacteremia (9 of 1199 [0.8%; 95% CI, 0.3%-1.4%] vs 50 of 1743 [2.9%; 95% CI, 2.1%-3.8%]). The rate of bacterial meningitis tended to be lower in the virus-positive group (0.4%) than in the viral-negative group (0.8%); the difference was not statistically significant. Negative viral status (aOR, 3.2; 95% CI, 2.3-4.6), was significantly associated with SBI in multivariable analysis. Conclusions Febrile infants ≤60 days of age with viral infections are at significantly lower, but non-negligible risk for SBIs, including bacteremia and bacterial meningitis.

Objective To determine the association between Hispanic ethnicity and limited English proficiency (LEP) and the rates of appendiceal perforation and advanced radiologic imaging (computed tomography and ultrasound) in children with abdominal pain. Study design We performed a secondary analysis of a prospective, cross-sectional, multicenter study of children aged 3-18 years presenting with abdominal pain concerning for appendicitis between March 2009 and April 2010 at 10 tertiary care pediatric emergency departments in the US. Appendiceal perforation and advanced imaging rates were compared between ethnic and language proficiency groups using simple and multivariate regression models. Results Of 2590 patients enrolled, 1001 (38%) had appendicitis, including 36% of non-Hispanics and 44% of Hispanics. In multivariate modeling, Hispanics with LEP had a significantly greater odds of appendiceal perforation (OR, 1.44; 95% CI, 1.20-1.74). Hispanics with LEP with appendiceal perforation of moderate clinical severity were less likely to undergo advanced imaging compared with English-speaking non-Hispanics (OR, 0.64; 95% CI, 0.43-0.95). Conclusion Hispanic ethnicity with LEP is an important risk factor for appendiceal perforation in pediatric patients brought to the emergency department with possible appendicitis. Among patients with moderate clinical severity, Hispanic ethnicity with LEP appears to be associated with lower imaging rates. This effect of English proficiency and Hispanic ethnicity warrants further investigation to understand and overcome barriers, which may lead to increased appendiceal perforation rates and differential diagnostic evaluation.

Objective To describe variations in emergency department (ED) quality measures and determine the association between ED costs and outcomes for 3 pediatric conditions: asthma, gastroenteritis, and simple febrile seizure. Study design This cross-sectional analysis of ED visits used the Pediatric Health Information System database. Children aged ≤18 years who were evaluated in an ED between July 2009 and June 2011 and had a discharge diagnosis of asthma, gastroenteritis, or simple febrile seizure were included. Two quality of care metrics were evaluated for each target condition, and Spearman correlation was applied to evaluate the relationship between ED costs (reflecting overall resource utilization) and admission and revisit rates among institutions. Results More than 250 000 ED visits at 21 member hospitals were analyzed. Among children with asthma, the median rate of chest radiography utilization was 35.1% (IQR, 31.3%-41.7%), and that of corticosteroid administration was 82.6% (IQR, 78.5%-86.5%). For children with gastroenteritis, the median rate of ondansetron administration was 52% (IQR, 43.2%-57.0%), and that of intravenous fluid administration was 18.1% (IQR, 15.3%-21.3%). Among children with febrile seizures, the median rate of computed tomography utilization was 3.1% (IQR, 2.7%-4.3%), and that of lumbar puncture was 4.0% (IQR, 2.3%-5.6%). Increased costs were not associated with lower admission rate or 3-day ED revisit rate for the 3 conditions. Conclusion We observed variation in quality measures for patients presenting to pediatric EDs with common conditions. Higher costs were not associated with lower hospitalization or ED revisit rates.

To describe factors associated with computed tomography (CT) use for children with minor blunt head trauma that are evaluated in emergency departments.Planned secondary analysis of a prospective observational study of children18 years with minor blunt head trauma between 2004 and 2006 at 25 emergency departments. CT scans were obtained at the discretion of treating clinicians. We risk-adjusted patients for clinically important traumatic brain injuries and performed multivariable regression analyses. Outcome measures were rates of CT use by hospital and by clinician training type.CT rates varied between 19.2% and 69.2% across hospitals. Risk adjustment had little effect on the differential rate of CT use. In low- and middle-risk patients, clinicians obtained CTs more frequently at suburban and nonfreestanding children's hospitals. Physicians with emergency medicine (EM) residency training obtained CTs at greater rates than physicians with pediatric residency or pediatric EM training. In multivariable analyses, compared with pediatric EM-trained physicians, the OR for CT use among EM-trained physicians in children2 years was 1.24 (95% CI 1.04-1.46), and for children2 years was 1.68 (95% CI 1.50-1.89). Physicians of all training backgrounds, however, overused CT scans in low-risk children.Substantial variation exists in the use of CT for children with minor blunt head trauma not explained by patient severity or rates of positive CT scans or clinically important traumatic brain injuries.