Factors Associated With Prolonged Mental Health Admissions at US Children's Hospitals.

BACKGROUND AND OBJECTIVES: Mental health (MH) hospitalizations at medical hospitals are associated with longer length of stay (LOS) compared with non-MH hospitalizations, but patient factors and costs associated with prolonged MH hospitalizations are unknown. The objective of this paper is to assess patient clinical and demographic factors associated with prolonged MH hospitalizations and describe variation in MH LOS across US children's hospitals. METHODS: We studied children aged 5 to 20 years hospitalized with a primary MH diagnosis during 2021 and 2022 across 46 children's hospitals using the Pediatric Health Information System database. Generalized estimating equations, clustered on hospital, tested associations between patient characteristics with prolonged MH hospitalization, defined as those in the 95th percentile or above (>14 days). RESULTS: Among 42 654 primary MH hospitalizations, most were aged 14 to 18 (62.4%), female (68.5%), and non-Hispanic white (53.8%). The most common primary MH diagnoses were suicide/self-injury (37.4%), depressive disorders (16.6%), and eating disorders (10.9%). The median (interquartile range) LOS was 2 days (1-5), but 2169 (5.1%) experienced a hospitalization >14 days. In adjusted analyses, race and ethnicity, category of MH diagnosis, and increasing medical and MH complexity were associated with prolonged hospitalization. CONCLUSIONS: Our results emphasize several diagnoses and clinical descriptors for targeted interventions, such as behavioral and inpatient MH resources and discharge planning. Expanded investment in both community and inpatient MH supports have the potential to improve health equity and reduce prolonged MH hospitalizations.

Spending on public benefit programs and exposure to adverse childhood experiences.

BACKGROUND: Adverse childhood experiences are associated with poverty, and public benefit programs are increasingly used as primary prevention for negative child outcomes. OBJECTIVE: To estimate the association between spending on benefit programs and cumulative exposure to ACEs among children. PARTICIPANTS AND SETTING: Children aged 0-17 years in the United States during 2016-17 as reported in National Survey of Children's Health. METHODS: We examined the sum of state and federal spending on 5 categories of public benefit programs at the state-level. The primary exposure was mean annual spending per person living below the Federal poverty limit across 2010-2017 Federal fiscal years. The primary outcome was children <18 years old having ever been exposed to ≥ 4 ACEs. RESULTS: Nationally, 5.7 % (95 % confidence interval [CI] 5.3 % - 6.0 %) of children had exposure to ≥ 4 ACEs. After adjustment for children's race and ethnicity, total spending on benefit programs was associated with lower exposure to ≥ 4 ACEs (odds 0.96 [95 % CI: 0.95, 0.97]; p < 0.001). Increased spending in each individual benefit category was also associated with decreased cumulative ACEs exposure (all p < 0.05). Inverse associations were largely consistent when children were stratified by race and ethnicity and income strata. CONCLUSIONS: Investments in public benefit programs may not only decrease poverty but also have broad positive effects on near- and long-term child well-being beyond the programs' stated objectives. Findings support federal and state efforts to prioritize families' economic stability as part of a public health model to prevent ACEs.

Pilot study using machine learning to improve estimation of physical abuse prevalence.

BACKGROUND: International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM) codes have been shown to underestimate physical abuse prevalence. Machine learning models are capable of efficiently processing a wide variety of data and may provide better estimates of abuse. OBJECTIVE: To achieve proof of concept applying machine learning to identify codes associated with abuse. PARTICIPANTS AND SETTING: Children <5 years, presenting to the emergency department with an injury or abuse-specific ICD-10-CM code and evaluated by the child protection team (CPT) from 2016 to 2020 at a large Midwestern children's hospital. METHODS: The Pediatric Health Information System (PHIS) and the CPT administrative databases were used to identify the study sample and injury and abuse-specific ICD-10-CM codes. Subjects were divided into abused and non-abused groups based on the CPT's evaluation. A LASSO logistic regression model was constructed using ICD-10-CM codes and patient age to identify children likely to be diagnosed by the CPT as abused. Performance was evaluated using repeated cross-validation (CV) and Reciever Operator Characteristic curve. RESULTS: We identified 2028 patients evaluated by the CPT with 512 diagnosed as abused. Using diagnosis codes and patient age, our model was able to accurately identify patients with confirmed PA (mean CV AUC = 0.87). Performance was still weaker for patients without existing ICD codes for abuse (mean CV AUC = 0.81). CONCLUSIONS: We built a model that employs injury ICD-10-CM codes and age to improve accuracy of distinguishing abusive from non-abusive injuries. This pilot modelling endeavor is a steppingstone towards improving population-level estimates of abuse.

Racial and ethnic disparities in diagnostic imaging for child physical abuse.

IMPORTANCE: Racial bias may affect occult injury testing decisions for children with concern for abuse. OBJECTIVES: To determine the association of race on occult injury testing decisions at children's hospitals. DESIGN: In this retrospective study, we measured disparities in: (1) the proportion of visits for which indicated diagnostic imaging studies for child abuse were obtained; (2) the proportion of positive tests. SETTING: The Pediatric Health Information System (PHIS) administrative database encompassing 49 tertiary children's hospitals during 2017-2019. PARTICIPANTS: We built three cohorts based on guidelines for diagnostic testing for child abuse: infants with traumatic brain injury (TBI; n = 1952), children <2 years old with extremity fracture (n = 20,842), and children <2 years old who received a skeletal survey (SS; n = 13,081). MAIN OUTCOMES AND MEASURES: For each group we measured: (1) the odds of receiving a specific guideline-recommended diagnostic imaging study; (2) among those with the indicated imaging study, the odds of an abuse-related injury diagnosis. We calculated both unadjusted and adjusted odds ratios (AOR) by race and ethnicity, adjusting for sex, age in months, payor, and hospital. RESULTS: In infants with TBI, the odds of receiving a SS did not differ by racial group. Among those with a SS, the odds of rib fracture were higher for non-Hispanic Black than Hispanic (AOR 2.05 (CI 1.31, 3.2)) and non-Hispanic White (AOR 1.57 (CI 1.11, 2.32)) patients. In children with extremity fractures, the odds of receiving a SS were higher for non-Hispanic Black than Hispanic and non-Hispanic White patients (AOR 1.97 (CI 1.74, 2.23)); (AOR 1.17 (CI 1.05, 1.31)), respectively, and lower for Hispanic than non-Hispanic White patients (AOR 0.59 (CI 0.53, 0.67)). Among those receiving a SS, the rate of rib fractures did not differ by race. In children with skeletal surveys, the odds of receiving neuroimaging did not differ by race. Among those with neuroimaging, the odds of a non-fracture, non-concussion TBI were lower in non-Hispanic Black than Hispanic patients (AOR 0.7 (CI 0.57, 0.86)) and were higher among Hispanic than non-Hispanic White patients (AOR 1.23 (CI 1.02, 1.47)). CONCLUSIONS AND RELEVANCE: We did not identify a consistent pattern of race-based disparities in occult injury testing when considering the concurrent yield for abuse-related injuries.

Public Benefit Programs and Differential Associations With Child Maltreatment by Race and Ethnicity.

Importance: Public benefit programs, including state spending on local, state, and federal-state partnership programs, have consistently been associated with overall reductions in child protective services (CPS) involvement. Inequities in eligibility and access to benefit programs may contribute to varying associations by race and ethnicity. Objective: To determine whether associations between state spending on benefit programs and rates of CPS investigations differ by race and ethnicity. Design, Setting, and Participants: This cross-sectional ecological study used repeated state-level measures of child maltreatment from the National Child Abuse and Neglect Data System and population estimates from the US Census Bureau for all Black, Hispanic, and White children. All 50 US states from October 1, 2009, through September 30, 2019 (fiscal years 2010-2019), were included. Data were collected and analyzed from May 13, 2022, to March 2, 2023. Exposures: Annual state spending on benefit programs per person living below the federal poverty limit, total and by the following subcategories: (1) cash, housing, and in-kind; (2) housing infrastructure; (3) child care assistance; (4) refundable earned income tax credit; and (5) medical assistance programs. Main Outcomes and Measures: Race- and ethnicity-specific rates of CPS investigations. Generalized estimating equations, with repeated measures of states, an interaction between race and spending, and estimated incidence rate ratios (IRRs) and 95% CIs for incremental changes in spending of US $1000 per person living below the federal poverty limit were calculated after adjustment for federal spending, race- and ethnicity-specific child poverty rate, and year. Results: A total of 493 state-year observations were included in the analysis. The association between total spending and CPS investigations differed significantly by race and ethnicity: there was an inverse association between total state spending and CPS investigations for White children (IRR, 0.94 [95% CI, 0.91-0.98]) but not for Black children (IRR, 0.98 [95% CI, 0.94-1.02]) or Hispanic children (IRR, 0.99 [95% CI, 0.95-1.03]) (P = .02 for interaction). Likewise, inverse associations were present for only White children with respect to all subcategories of state spending and differed significantly from Black and Hispanic children for all subcategories except the refundable earned income tax credit (eg, IRR for medical assistance programs for White children, 0.89 [95% CI, 0.82-0.96]; P = .005 for race and spending interaction term). Conclusions and Relevance: These results raise concerns that benefit programs may add relative advantages for White children compared with Black and Hispanic children and contribute to racial and ethnic disparities in CPS investigations. States' eligibility criteria and distribution practices should be examined to promote equitable effects on adverse child outcomes.

Clinical guideline highlights for the hospitalist: Evaluation for bleeding disorders in suspected child abuse.

GUIDELINE TITLE: Evaluation for Bleeding Disorders in Suspected Child Abuse RELEASE DATE: October 2022 PRIOR VERSION(S): 2013 DEVELOPER: American Academy of Pediatrics FUNDING SOURCE: American Academy of Pediatrics TARGET POPULATION: Children and infants with bruising or bleeding suspicious for physical abuse.

Diagnostic testing for and detection of physical abuse in infants with brief resolved unexplained events.

BACKGROUND: A Brief Resolved Unexplained Event (BRUE) can be a sign of occult physical abuse. OBJECTIVES: To identify rates of diagnostic testing able to detect physical abuse (head imaging, skeletal survey, and liver transaminases) at BRUE presentation. The secondary objective was to estimate the rate of physical abuse diagnosed at initial BRUE presentation through 1 year of age. PARTICIPANTS AND SETTING: Infants who presented with a BRUE at one of 15 academic or community hospitals were followed from initial BRUE presentation until 1 year of age for BRUE recurrence or revisits. METHODS: This study was part of the BRUE Research and Quality Improvement Network, a multicenter retrospective cohort examining infants with BRUE. Generalized estimating equations assessed associations with performance of diagnostic testing (adjusted odds ratio (aOR)). RESULTS: Of the 2036 infants presenting with a BRUE, 6.2 % underwent head imaging, 7.0 % skeletal survey, and 12.1 % liver transaminases. Infants were more likely to undergo skeletal survey if there were physical examination findings concerning for trauma (aOR 8.23, 95 % CI [1.92, 35.24], p < 0.005) or concerning social history (aOR 1.89, 95 % CI [1.13, 3.16], p = 0.015). There were 7 (0.3 %) infants diagnosed with physical abuse: one at BRUE presentation, one <3 days after BRUE presentation, and five >30 days after BRUE presentation. CONCLUSION: There were low rates of diagnostic testing and physical abuse identified in infants presenting with BRUE. Further study including standardized testing protocols is warranted to identify physical abuse in infants presenting with a BRUE.

Differences Between Viral Meningitis and Abusive Head Trauma.

OBJECTIVES: To evaluate the hypothesis that viral meningitis may mimic abusive head trauma (AHT) by comparing the history of present illness (HPI) and clinical presentation of young children with proven viral meningitis to those with AHT and those with subdural hemorrhage (SDH) only. We hypothesized that significant differences would exist between viral meningitis and the comparison groups. METHODS: We performed a 5-year retrospective case-control study of subjects aged <2 years, comparing those with confirmed viral meningitis (controls) to those with SDH evaluated by the hospital child abuse pediatrics team (cases). Cases were classified as SDH with concomitant suspicious injuries (AHT) and without concomitant suspicious injuries (SDH-only). Groups were compared across demographic (5 measures), HPI (11 measures), and clinical (9 measures) domains. Odds ratios were calculated for measures within each domain. RESULTS: Of 550 subjects, there were 397 viral meningitis, 118 AHT, and 35 SDH-only subjects. Viral meningitis differed significantly from AHT subjects on all demographic measures, and from SDH-only subjects on age. Viral meningitis differed significantly from AHT subjects in all HPI measures with odds ratios ranging from 2.7 to 322.5, and from SDH-only subjects in 9 HPI measures with odds ratios ranging from 4.6 to 485.2. In the clinical domain, viral meningitis differed significantly from AHT subjects in all measures, with odds ratios ranging from 2.5 to 74.0, and from SDH-only subjects in 5 measures with odds ratios ranging from 2.9 to 16.8. CONCLUSIONS: Viral meningitis is not supported as a mimic of AHT.

Racial and ethnic differences in pediatric unintentional injuries requiring hospitalization.

BACKGROUND/OBJECTIVE: This study aims to comprehensively examine racial and ethnic differences in pediatric unintentional injuries requiring hospitalization by age across injury mechanisms. STUDY DESIGN: This was a retrospective, nationally representative cross-sectional analysis of discharge data within the 2016 Kids' Inpatient Database for 98,611 children ≤19 years with unintentional injuries resulting in hospitalization. Injury categories included passengers and pedestrians injured in a motor vehicle crash, falls, drownings, burns, firearms, drug and nondrug poisonings, suffocations, and other injuries. Relative risk (RR) for injuries requiring hospitalization were calculated for children of Black, Hispanic, and Other races and ethnicities compared with White children, and then RR were further stratified by age. Excessive hospitalizations were calculated as the absolute number of hospitalizations for each race and ethnicity group that would have been avoided if each group had the same rate as White children. RESULTS: Black children were significantly more likely to be hospitalized compared with White children for all injury mechanisms except falls, and in nearly all age groups with the greatest RR for firearm injuries (RR 9.8 [95% confidence interval: 9.5-10.2]). Differences were associated with 6263 excessive hospitalizations among all racial and ethnic minority children compared with White children. CONCLUSIONS: Racial and ethnic minority children represent populations at persistent disproportionate risk for injuries resulting in hospitalization; risk that varies in important ways by injury mechanism and children's age. These findings suggest the importance of the environmental and societal exposures that may drive these differences, but other factors, such as provider bias, may also contribute.

Child Opportunity Index and Changes in Pediatric Acute Care Utilization in the COVID-19 Pandemic.

BACKGROUND: Pediatric acute care utilization decreased dramatically during the coronavirus disease 2019 (COVID-19) pandemic. This study examined the association between the Child Opportunity Index (COI), a multidimensional neighborhood measure of childhood opportunity, and changes in acute care utilization at US pediatric hospitals during the COVID-19 pandemic compared with the previous 3 years. METHODS: This observational study used administrative data across 41 US-based pediatric hospitals. Children aged 0 to 17 years with emergency department (ED) encounters during the study period were included. The COVID-19 pandemic time period (March 15, 2020-March 14, 2021) was the primary exposure. The primary outcome was the relative volume drop in ED encounters and observation/inpatient admissions through the ED by COI quintile. RESULTS: Of 12 138 750 encounters, 3 705 320 (30.5%) were among the very low COI quintile. Overall, there was a 46.8% relative volume reduction in the pandemic period compared with the prepandmic period. This drop in volume occurred disproportionately among the very low COI quintile (51.1%) compared with the very high COI quintile (42.8%). The majority of clinical diagnosis groups demonstrated larger relative volume drops among the very low COI quintile. CONCLUSIONS: Acute care utilization decreased the most among children from very low COI neighborhoods, narrowing previously described acute care utilization disparities. Additional study of patient perspectives on health care needs and access during this period is required to understand these changes.

Late Pre-term Infants with Severe Bronchiolitis and Risk of Asthma by Age 5 Years.

In a prospective, multicenter cohort of infants hospitalized with bronchiolitis, we found infants born late pre-term (ie, gestational age of 34-36.9 weeks) had 35% higher odds of having asthma by age 5 years compared with infants born at full-term.

Prenatal Substance Exposure and Child Maltreatment: A Systematic Review.

State and federal policies regarding substance use in pregnancy, specifically whether a notification to child protective services is required, continue to evolve. To inform practice, policy, and future research, we sought to synthesize and critically evaluate the existing literature regarding the association of prenatal substance exposure with child maltreatment. We conducted a comprehensive electronic search of PubMed, Web of Science, PsycInfo, CHINAL, Social Work Abstracts, Sociological Abstracts, and Social Services Abstracts. We identified 30 studies that examined the association of exposure to any/multiple substances, cocaine, alcohol, opioids, marijuana, and amphetamine/methamphetamine with child maltreatment. Overall, results indicated that substance exposed infants have an increased likelihood of child protective services involvement, maternal self-reported risk of maltreatment behaviors, hospitalizations and clinic visits for suspected maltreatment, and adolescent retrospective self-report of maltreatment compared to unexposed infants. While study results suggest an association of prenatal substance exposure with child maltreatment, there are several methodological considerations that have implications for results and interpretation, including definitions of prenatal substance exposure and maltreatment, study populations used, and potential unmeasured confounding. As each may bias study results, careful interpretation and further research are warranted to appropriately inform programs and policy.

State Spending on Public Benefit Programs and Child Maltreatment.

BACKGROUND: To determine the association between states' total spending on benefit programs and child maltreatment outcomes. METHODS: This was an ecological study of all US states during federal fiscal years 2010-2017. The primary predictor was states' total annual spending on local, state, and federal benefit programs per person living ≤100% federal poverty limit, which was the sum of (1) cash, housing, and in-kind assistance, (2) housing infrastructure, (3) child care assistance, (4) refundable Earned Income Tax Credit, and (5) Medical Assistance Programs. The main outcomes were rates of maltreatment reporting, substantiations, foster care placements, and fatalities after adjustment for relevant confounders. Generalized estimating equations adjusted for federal spending and estimated adjusted incidence rate ratios (IRRs) and 95% confidence intervals (CIs). RESULTS: States' total spending was inversely associated with all maltreatment outcomes. For each additional $1000 states spent on benefit programs per person living in poverty, there was an associated -4.3% (adjusted IRR: 0.9573 [95% CI: 0.9486 to 0.9661]) difference in reporting, -4.0% (adjusted IRR: 0.903 [95% CI: 0.9534 to 0.9672]) difference in substantiations, -2.1% (adjusted IRR: 0.9795 [95% CI: 0.9759 to 0.9832]) difference in foster care placements, and -7.7% (adjusted IRR: 0.9229 [95% CI: 0.9128 to 0.9330]) difference in fatalities. In 2017, extrapolating $1000 of additional spending for each person living in poverty ($46.5 billion nationally, or 13.3% increase) might have resulted in 181 850 fewer reports, 28 575 fewer substantiations, 4168 fewer foster care placements, and 130 fewer fatalities. CONCLUSIONS: State spending on benefit programs was associated with reductions in child maltreatment, which might offset some benefit program costs.

Association of routine school closures with child maltreatment reporting and substantiation in the United States; 2010-2017.

BACKGROUND: There exists a presumption that school closures lead to a diminished capacity to detect child maltreatment, but empiric evidence is lacking. OBJECTIVE: To determine if child maltreatment reporting and substantiation differ between periods when schools are routinely closed compared to in session. PARTICIPANTS AND SETTING: Child maltreatment reporting and substantiation among all U.S. States and the District of Columbia from January 1, 2010 through December 31, 2017. METHODS: Two-week intervals during periods of routine school closure (early January, June through mid-August, late November, and late December) were compared to all other 2-week intervals. Negative binomial generalized estimating equations compared rates of reporting and substantiation, resulting in incidence rate ratios (IRR) with 95% confidence intervals (CI). RESULTS: Compared to when school was in session, reporting was 16.0% (IRR 0.84 [95% CI: 0.83, 0.85]) lower during school closures and substantiations were 12.3% (IRR 0.88 [95% CI: 0.86, 0.89]) lower. The largest reductions in reporting were observed among education personnel (-42.1%; IRR 0.58 [95% CI: 0.54, 0.62]), children aged 5-17 years (-18.6%; IRR 0.81 [95% CI: 0.80, 0.83), and for physical abuse (-19.6%; IRR 0.80 [95% CI: 0.79, 0.82]). Reductions during closure periods were not matched by increases during two-week intervals immediately following closure periods. CONCLUSIONS: Results suggest that the detection of child maltreatment may be diminished during periods of routine school closure. Findings may inform prevention planning and risk-benefit analyses for future school closures. Further study should disentangle the issue of decreased detection versus decreased prevalence of maltreatment during school closures.

The Child Opportunity Index 2.0 and Hospitalizations for Ambulatory Care Sensitive Conditions.

BACKGROUND AND OBJECTIVES: Hospitalizations for ambulatory care sensitive conditions (ACSCs) are thought to be avoidable with high-quality outpatient care. Morbidity related to ACSCs has been associated with socioeconomic contextual factors, which do not necessarily capture the complex pathways through which a child's environment impacts health outcomes. Our primary objective was to test the association between a multidimensional measure of neighborhood-level child opportunity and pediatric hospitalization rates for ACSCs across 2 metropolitan areas. METHODS: This was a retrospective population-based analysis of ACSC hospitalizations within the Kansas City and Cincinnati metropolitan areas from 2013 to 2018. Census tracts were included if located in a county where Children's Mercy Kansas City or Cincinnati Children's Hospital Medical Center had >80% market share of hospitalizations for children <18 years. Our predictor was child opportunity as defined by a composite index, the Child Opportunity Index 2.0. Our outcome was hospitalization rates for 8 ACSCs. RESULTS: We included 604 943 children within 628 census tracts. There were 26 977 total ACSC hospitalizations (46 hospitalizations per 1000 children; 95% confidence interval [CI]: 45.4-46.5). The hospitalization rate for all ACSCs revealed a stepwise reduction from 79.9 per 1000 children (95% CI: 78.1-81.7) in very low opportunity tracts to 31.2 per 1000 children (95% CI: 30.5-32.0) in very high opportunity tracts (P < .001). This trend was observed across cities and diagnoses. CONCLUSIONS: Links between ACSC hospitalizations and child opportunity extend across metropolitan areas. Targeting interventions to lower-opportunity neighborhoods and enacting policies that equitably bolster opportunity may improve child health outcomes, reduce inequities, and decrease health care costs.

Identification of physical abuse-related hospitalizations in young children: Impact of the transition to ICD-10-CM coding.

BACKGROUND: Administrative healthcare databases are frequently used for child physical abuse (CPA)-related research and surveillance. In October 2015, the United States transitioned to the International Classification of Diseases, Clinical Modification-10th Revision (ICD-10-CM) coding regimen. ICD-10-CM expands coding related to CPA, including codes to differentiate suspected from confirmed CPA. OBJECTIVE: This study examined the impact of the transition to ICD-10-CM coding on population-level trends in rates of hospitalizations coded for CPA. PARTICIPANTS AND SETTINGS: Hospitalizations coded as related to CPA in children <5 years-of-age from 2010 to 2017 were identified across 19 Statewide Inpatient Databases. METHODS: Interrupted time series analyses were used to assess the impact of the coding transition on hospitalizations coded for CPA, overall and by child race/ethnicity. RESULTS: Of 9715 hospitalizations coded for CPA, 2797 (29%) occurred after the coding transition, including 51% coded for suspected CPA and 49% coded for confirmed CPA. There was a marginally-significant increase in the trend in CPA-related hospitalization after the coding transition among all children (0.09 per 100,000 children-per-quarter, p = 0.06), a significant increase in the trend among white children (0.15 per 100,000 children-per-quarter, p = 0.01), and no change among Black or Hispanic children. After the coding transition, hospitalizations coded for suspected CPA increased significantly overall (0.10 per 100,000 children-per-quarter, p < 0.001), and in particular among white children (0.12 per 100,000 children-per-quarter, p = 0.01) but not among Black or Hispanic children. CONCLUSIONS: The transition to ICD-10-CM differentially impacted trends in hospitalizations coded for CPA by child race/ethnicity. Further research is necessary to discern the reasons for these discrepancies.

Variations in Sedated Echocardiography and Association With Repeat Echocardiography in Nonrefractory Kawasaki Disease.

OBJECTIVES: Echocardiography performed under sedation allows for better coronary artery visualization in children with Kawasaki disease. We sought to describe hospital-level variability in the percentage of sedated echocardiography (SE) among children with nonrefractory Kawasaki disease (NRKD) and then test its association with repeat echocardiography, length of stay (LOS), and costs. METHODS: We identified children in the Pediatric Health Information System <36 months of age hospitalized with NRKD from March 2010 to February 2017. Hospital-level percentage of SE was the primary outcome. Secondary outcomes were repeat echocardiography during the same hospitalization, LOS, and costs. We used χ2 and Wilcoxon rank tests to compare differences in repeat echocardiography between SE and nonsedated echocardiography. Pearson correlation was used to test associations between SE and repeat echocardiography, LOS, and costs. RESULTS: There were 2887 NRKD hospitalizations from 40 children's hospitals. Initial SE varied from 0.0% to 87.0% (median 5.9%; interquartile range 1.4%-21.1%). Of initial echocardiographies, 22.4% were sedated and 10.1% of all hospitalizations had a repeat echocardiography. Use of SE at the hospital level was associated with a lower likelihood for repeat echocardiography (r = -0.32; 95% confidence interval -0.58 to -0.01; P = .042). Absolute risk reduction was 3.5% and the number needed to sedate was 29. Initial SE was not associated with LOS or hospital-level costs but was associated with increased patient-level costs. CONCLUSIONS: Significant variation exists in the use of SE for children <36 months of age with NRKD at children's hospitals. Our results suggest that determination of SE should not be driven by concern for repeat echocardiography or LOS considerations.

Variation in Care and Clinical Outcomes Among Infants Hospitalized With Hyperbilirubinemia.

OBJECTIVES: To assess hospital-level variation in laboratory testing and intravenous fluid (IVF) use and examine the association between these interventions and hospitalization outcomes among infants admitted with neonatal hyperbilirubinemia. METHODS: We performed a retrospective multicenter study of infants aged 2 to 7 days hospitalized with a primary diagnosis of hyperbilirubinemia from December 1, 2016, to June 30, 2018, using the Pediatric Health Information System. Hospital-level variation in laboratory and IVF use was evaluated after adjusting for clinical and demographic factors and associated with hospital-level outcomes by using Pearson correlation. RESULTS: We identified 4396 infants hospitalized with hyperbilirubinemia. In addition to bilirubin level, the most frequently ordered laboratories were direct antiglobulin testing (45.7%), reticulocyte count (39.7%), complete blood cell counts (43.7%), ABO blood type (33.4%), and electrolyte panels (12.9%). IVFs were given to 26.3% of children. Extensive variation in laboratory testing and IVF administration was observed across hospitals (all P < .001). Increased use of laboratory testing but not IVFs was associated with a longer length of stay (P = .007 and .162, respectively). Neither supplementary laboratory use nor IVF use was associated with either readmissions or emergency department revisits. CONCLUSIONS: Substantial variation exists among hospitals in the management of infants with hyperbilirubinemia. With our results, we suggest that additional testing outside of bilirubin measurement may unnecessarily increase resource use for infants hospitalized with hyperbilirubinemia.