Variation in Length of Stay by Level of Neonatal Care Among Moderate and Late Preterm Infants.

BACKGROUND AND OBJECTIVES: Moderate and late preterm infants are a growing subgroup of neonates with increased care needs after birth, yet standard protocols are lacking. We aim to describe variation in length of stay (LOS) by gestational age (GA) across hospitals within the same level of neonatal care and between different levels of neonatal care. METHODS: Retrospective cohort study of hospitalizations for moderate (32-33 weeks GA) and late (34-36 weeks GA) preterm infants in 2019 Kid's Inpatient Database. We compared adjusted LOS in this cohort and evaluated variation within hospitals of the same level and across different levels of neonatal care. RESULTS: This study includes 217 051 moderate (26.2%) and late (73.8%) preterm infants from level II (19.7%), III (66.3%), and IV (11.1%) hospitals. Patient-level (race and ethnicity, primary payor, delivery type, multiple gestation, birth weight) and hospital-level (birth region, level of neonatal care) factors were significantly associated with LOS. Adjusted mean LOS varied for hospitals within the same level of neonatal care with level II hospitals showing the greatest variability among 34- to 36- week GA infants when compared with level III and IV hospitals (P < .01). LOS also varied significantly between levels of neonatal care with the greatest variation (0.9 days) seen in 32-week GA between level III and level IV hospitals. CONCLUSIONS: For moderate and late preterm infants, the level of neonatal care was associated with variation in LOS after adjusting for clinical severity. Hospitals providing level II neonatal care showed the greatest variation and may provide an opportunity to standardize care.

Social Determinants of Health ICD-10 Code Use in Inpatient Pediatrics.

BACKGROUND AND OBJECTIVES: The American Academy of Pediatrics endorses screening for social determinants of health (SDOH) and providing families resources for unmet needs. A systematic response to unmet needs requires identification, documentation, and provision of resources. Our goal was to compare SDOH International Classification of Diseases, 10th Revision (ICD-10), code use for pediatric inpatients after policy changes in 2018 permitting coding by nonphysicians. METHODS: We conducted a retrospective cohort study comparing data from the 2016 and 2019 Kid's Inpatient Database for patients <21 years old. The primary variable was the presence of an SDOH code, defined as an ICD-10 Z-code (Z55-Z65) or 1 of 13 ICD-10 codes recommended by the American Academy of Pediatrics. We compared overall SDOH code usage between 2016 and 2019, and by Z-code category, demographic, clinical, and hospital characteristics using χ2 tests and odds ratios. Using logistic regression, we examined hospital-level characteristics for hospitals with >5% of discharges with an SDOH code. RESULTS: SDOH code documentation increased from 1.4% in 2016 to 1.9% in 2019 (P < .001), with no notable differences based on Z-code category. In both periods, SDOH code documentation was more common in adolescents, Native Americans, and patients with mental health diagnoses. The number of all hospitals using any SDOH code increased nearly 8% between 2016 and 2019. CONCLUSIONS: ICD-10 codes remain underused to track SDOH needs within the inpatient pediatric setting. Future research should explore whether SDOH code documentation is associated with increased response to unmet social needs and, if so, how to improve use of SDOH codes by all providers.

Evaluation of an Educational Outreach and Audit and Feedback Program to Reduce Continuous Pulse Oximetry Use in Hospitalized Infants With Stable Bronchiolitis: A Nonrandomized Clinical Trial.

Importance: National guidelines recommend against continuous pulse oximetry use for hospitalized children with bronchiolitis who are not receiving supplemental oxygen, yet guideline-discordant use remains high. Objectives: To evaluate deimplementation outcomes of educational outreach and audit and feedback strategies aiming to reduce guideline-discordant continuous pulse oximetry use in children hospitalized with bronchiolitis who are not receiving supplemental oxygen. Design, Setting, and Participants: A nonrandomized clinical single-group deimplementation trial was conducted in 14 non-intensive care units in 5 freestanding children's hospitals and 1 community hospital from December 1, 2019, through March 14, 2020, among 847 nurses and physicians caring for hospitalized children with bronchiolitis who were not receiving supplemental oxygen. Interventions: Educational outreach focused on communicating details of the existing guidelines and evidence. Audit and feedback strategies included 2 formats: (1) weekly aggregate data feedback to multidisciplinary teams with review of unit-level and hospital-level use of continuous pulse oximetry, and (2) real-time 1:1 feedback to clinicians when guideline-discordant continuous pulse oximetry use was discovered during in-person data audits. Main Outcomes and Measures: Clinician ratings of acceptability, appropriateness, feasibility, and perceived safety were assessed using a questionnaire. Guideline-discordant continuous pulse oximetry use in hospitalized children was measured using direct observation of a convenience sample of patients with bronchiolitis who were not receiving supplemental oxygen. Results: A total of 847 of 1193 eligible clinicians (695 women [82.1%]) responded to a Likert scale-based questionnaire (71% response rate). Most respondents rated the deimplementation strategies of education and audit and feedback as acceptable (education, 435 of 474 [92%]; audit and feedback, 615 of 664 [93%]), appropriate (education, 457 of 474 [96%]; audit and feedback, 622 of 664 [94%]), feasible (education, 424 of 474 [89%]; audit and feedback, 557 of 664 [84%]), and safe (803 of 847 [95%]). Sites collected 1051 audit observations (range, 47-403 per site) on 709 unique patient admissions (range, 31-251 per site) during a 3.5-month period of continuous pulse oximetry use in children with bronchiolitis not receiving supplemental oxygen, which were compared with 579 observations (range, 57-154 per site) from the same hospitals during the baseline 4-month period (prior season) to determine whether the strategies were associated with a reduction in use. Sites conducted 148 in-person educational outreach and aggregate data feedback sessions and provided real-time 1:1 feedback 171 of 236 times (72% of the time when guideline-discordant monitoring was identified). Adjusted for age, gestational age, time since weaning from supplemental oxygen, and other characteristics, guideline-discordant continuous pulse oximetry use decreased from 53% (95% CI, 49%-57%) to 23% (95% CI, 20%-25%) (P < .001) during the intervention period. There were no adverse events attributable to reduced monitoring. Conclusions and Relevance: In this nonrandomized clinical trial, educational outreach and audit and feedback deimplementation strategies for guideline-discordant continuous pulse oximetry use among hospitalized children with bronchiolitis who were not receiving supplemental oxygen were positively associated with clinician perceptions of feasibility, acceptability, appropriateness, and safety. Evaluating the sustainability of deimplementation beyond the intervention period is an essential next step. Trial Registration: ClinicalTrials.gov Identifier: NCT04178941.

Race, insurance status, and traumatic brain injury outcomes before and after enactment of the Affordable Care Act.

BACKGROUND: The Affordable Care Act aims to improve patient outcomes. Race/ethnicity and insurance status impact outcomes after traumatic brain injury. We sought to gauge the Affordable Care Act's effect on outcomes after traumatic brain injury, as graded by race/ethnicity and insurance status. METHODS: The National Trauma Data Bank was utilized to identify traumatic brain injury patients before and after the Affordable Care Act. Patient outcomes comprised of hospital duration of stay, in-hospital mortality, discharge to rehabilitation, and surgical procedures. Using regression analysis, we evaluated the impact of race/ethnicity and insurance status on traumatic brain injury outcomes, then compared them before and after the Affordable Care Act. RESULTS: Mortality decreased for blacks (odds ratio = 0.96 [confidence interval 0.83-1.10] to odds ratio = 0.79 [confidence interval = 0.70-0.89], and Hispanics (odds ratio = 1.03 [confidence interval = 0.90-1.17] to odds ratio = 0.79 [confidence interval = 0.70-0.89]). Mortality increased for the uninsured (odds ratio = 1.28 [confidence interval = 1.11-1.47] to odds ratio = 1.40 [confidence interval = 1.24-1.58]). Medicaid patients underwent decreased duration of stay, (coefficient = 2.75 [confidence interval = 2.49-3.02] to coefficient = 2.17, [confidence interval = 1.98-2.37]), discharge to rehabilitation (odds ratio = 1.15, [confidence interval = 1.04-1.26] to odds ratio = 0.95 [confidence interval = 0.87-1.03]), and surgical procedures (odds ratio = 1.28 [confidence interval = 1.13-1.45] to odds ratio = 1.18, [confidence interval = 1.07-1.30]), while mortality remained unchanged. CONCLUSION: After the Affordable Care Act traumatic brain injury mortality decreased for blacks and Hispanics, but increased for the uninsured. Decreasing trends in resource consumption were also evident, especially for Medicaid patients. These results may illustrate altered delivery of care.

Insurance status and race affect treatment and outcome of traumatic brain injury.

BACKGROUND: There is increasing evidence that race and socioeconomic factors affect patient outcomes after traumatic brain injury (TBI). Our goal was to assess the effect of race, ethnicity and insurance status on hospital length of stay, procedures performed, mortality, and discharge disposition after TBI. METHODS: This was a retrospective cohort study using the National Trauma Data Bank (2002-2012) to analyze patients aged 14-89 y with one of five closed head injuries. Univariate regressions identified demographic and injury characteristics that were significant predictors of outcomes. These variables were then included in multivariate regression models. RESULTS: We analyzed 187,354 TBI patients. The sample was 78% white, 9% black, 9% Hispanic, 3% Asian, and 1% native American, and included 42% Medicare, 30% private insurance, 12% uninsured, 8% other insurance, and 8% Medicaid. Compared with white patients, black and Hispanic patients were more likely to have a TBI procedure (blacks odds ratio [OR] = 1.19, P < 0.001; Hispanics OR = 1.33, P < 0.001), had longer hospital stays (blacks coeff = 1.02, P < 0.001; Hispanics coeff = 0.61, P < 0.001), were less likely to die in the hospital (blacks OR = 0.90, P = 0.006; Hispanics OR = 0.90, P = 0.007), and more (black OR = 1.09, P = 0.001) or less likely (Hispanic OR = 0.76, P < 0.001) to be discharged to rehabilitation. Compared with the privately insured, the uninsured were less likely to have a TBI procedure (OR = 0.90, P = 0.001), had longer hospital stays (coeff = 0.24, P < 0.001), were more likely to die in the hospital (OR = 1.37, P < 0.001), and less likely to be discharged to rehabilitation (OR = 0.53, P < 0.001). CONCLUSIONS: Race/ethnicity and insurance status significantly affect TBI patient outcomes, even after controlling for demographic and injury characteristics.