Differences In Severity Of Illness Research Articles

Infant feeding and criticality in children.

Data support the protective effects of human breast milk (HBM) feeding in acute illness but little is known about the impact of HBM feeding on the criticality of infants. To explore the relationship between early HBM feeding and severity of illness and recovery in critically ill children requiring intubation and mechanical ventilation for acute respiratory failure (ARF). Prospective cohort study of mothers of patients aged 1-36 months who participated in the acute and follow-up phases of the Randomized Evaluation of Sedation Titration for Respiratory Failure (RESTORE) clinical trial. Participants completed a survey describing HBM dose fed during their infant's first month of life. Of 138 patients, 70 (51%) received exclusive HBM feedings (90%-100% total feeds) and 68 (49%) did not. We found no group differences in severity of illness on paediatric intensive care unit (PICU) admission or severity of paediatric acute respiratory distress syndrome (PARDS) within the first 24-48 h of intubation/mechanical ventilation (Pediatric Risk of Mortality [PRISM] III-12 score median: 5 vs. 5, p = .88; moderate/severe PARDS: 53% vs. 54%, p = .63). While median time to recovery from ARF was reduced by 1 day in patients who received exclusive HBM feedings, the difference between groups was not statistically significant (median 1.5 vs. 2.6 days, hazard ratio 1.40 [95% confidence interval, 0.99-1.97], p = .06). Human breast milk dose was not associated with severity of illness on PICU admission in children requiring mechanical ventilation for ARF. Data support the protective effects of HBM during acute illness and data from this study support a clinically important reduction in time to recovery of ARF. Paediatric nurses should continue to champion HBM feeding to advance improvements in infant health.

Patient sex and outcomes in children with life-threatening hemorrhage.

Recent data suggest female sex imparts a survival benefit after trauma in adults. The independent associations between patient sex and age with outcomes have not been examined in children with life-threatening hemorrhage (LTH) from all etiologies. In a secondary analysis of a multicenter prospective observational study of children with LTH, Massive Transfusion in Children (MATIC), we analyzed if patient sex and age were associated with differences in severity of illness, therapies, and outcomes. Primary outcomes were 24 hour mortality and weight-adjusted transfusion volume during LTH. Kruskal-Wallis, chi-square testing, and multivariable linear regression were used for adjusted analyses. Of 449 children, 45% were females and 55% were males. Females were more commonly younger, white, and with less trauma as the etiology of LTH compared to males. Markers of clinical severity were similar between groups, except injury severity score (ISS) was higher in females in the trauma subgroup. In terms of resuscitative practices, females received greater weight-adjusted total transfusion volumes compared to males (76 (40-150) mL/kg vs. 53 (24-100) mL/kg), as well as increased red blood cells (RBCs), plasma, and platelets compared to males. After adjustment for confounders, female sex and age 0-11 years were independently associated with increased transfusion volume during LTH. There were no differences in mortality or adverse outcomes according to patient sex. Patient sex and age may impact factors associated with LTH and therapies received. Studies in developmental hemostasis are needed to determine the optimal transfusion strategy for LTH according to patient sex and age.

Association between goal nutrition and intubation in patients with bronchiolitis on noninvasive ventilation: A retrospective cohort study.

Acute bronchiolitis causes many hospitalizations in children younger than 2 years. Early enteral nutrition is associated with improved outcomes in these patients. However, often nutrition is withheld when patients require noninvasive respiratory support because of the risk of aspiration worsening respiratory failure, possibly requiring intubation. We hypothesize that achieving goal energy intake is associated with a lower intubation rate in hospitalized children with bronchiolitis who require noninvasive ventilation. This retrospective cohort study examined the association between goal enteral nutrition (60% of dietary reference energy intake) and intubation rates. We grouped patients by severity of illness and compared intubation rates in those who met goal energy to those who did not. We use stratified analysis methods (for both level of respiratory support and feeding route) to evaluate progression to intubation. Of the 272 patients, 215 met goal feeds. These groups had similar demographics, but the goal-feeds group started on higher respiratory support in the pediatric intensive care unit. We found that 4.65% of the patients who met goal feeds required intubation compared with 24.6% of patients who did not meet goal feeds (P < 0.0001), even after controlling for respiratory status at admission and time of feed initiation and feeding route. We observed when adjusting for severity, feeding route, and respiratory support, achieving goal energy intake remained associated with a lower rate of intubation, without higher rates of aspiration. Confounding factors include practice variation and difference in severity of illness that objective scoring may have missed.

Characteristics of Emergency Medicine Specimen Bank Participants Compared to the Overall Emergency Department Population.

Biorepositories lack diversity both demographically and with regard to the clinical complaints of patients enrolled. The Emergency Medicine Specimen Bank (EMSB) seeks to enroll a diverse cohort of patients for discovery research in acute care conditions. Our objective in this study was to determine the differences in demographics and clinical complaints between participants in the EMSB and the overall emergency department (ED) population. This was a retrospective analysis of participants of the EMSB and the entire UCHealth at University of Colorado Anschutz Medical Center (UCHealth AMC) ED population across three periods: peri-EMSB; post-EMSB; and COVID-19. We compared patients consented to the EMSB to the entire ED population to determine differences in age, gender, ethnicity, race, clinical complaints, and severity of illness. We used chi-square tests to compare categorical variables and the Elixhauser Comorbidity Index to determine differences in the severity of illness between the groups. Between February 5, 2018-January 29, 2022, there were 141,670 consented encounters in the EMSB, representing 40,740 unique patients and over 13,000 blood samples collected. In that same time, the ED saw approximately 188,402 unique patients for 387,590 encounters. The EMSB had significantly higher rates of participation from the following: patients 18-59 years old (80.3% vs 77.7%); White patients (52.3% vs 47.8%), and women (54.8% vs 51.1%) compared to the overall ED population. The EMSB had lower rates of participation from patients ≥70 years, Hispanic patients, Asian patients, and men. The EMSB population had higher mean comorbidity scores. During the six months after Colorado's first COVID-19 case, the rate of consented patients and samples collected increased. The odds of consent during the COVID-19 study period were 1.32 (95% CI 1.26-1.39), and the odds of sample capture were 2.19 (95% CI 2.0-2.41). The EMSB is representative of the overall ED population for most demographics and clinical complaints.

Use of Aerosolized Prostacyclins in Critically Ill Patients and Association With Clinical Outcomes.

To compare the effects of inhaled epoprostenol and inhaled iloprost in critically ill patients with refractory hypoxia. We performed a retrospective cohort analysis of patients admitted to the ICUs at the University of Oklahoma Health Sciences Center between 2015 and 2018. Adult patients who received aerosolized epoprostenol or iloprost for more than 4 hours were included in the analysis. The primary endpoint measured was to compare the change in Pao2/Fio2 ratio between patients treated with iloprost compared with epoprostenol. Secondary outcomes measured were 90-day in-hospital mortality and improvement in vasopressor requirements. A total of 126 patients were included in the study, 95 of whom received iloprost (75%) and 31 patients (25%) received epoprostenol. There were significant improvements in Pao2/Fio2 ratio in both the iloprost and epoprostenol group. Patients in the epoprostenol group appeared to have a higher 90-day mortality compared with the iloprost group. However, our study was not powered to detect a mortality difference and this finding likely represents a sicker population in the epoprostenol group and prescription bias. The use of iloprost was associated with higher vasopressor requirements in the first 12 hours of administration, an association was not observed in the epoprostenol group. In this retrospective cohort analysis, use of both pulmonary vasodilators was associated with similar improvement in gas exchange. The mortality difference observed likely represents difference in severity of illness. Further studies are needed to corroborate these findings.

The need for integrated clinical and administrative data models for risk adjustment in assessment of the cost transplant care.

Value-based purchasing requires accurate techniques to appropriately measure both outcomes and cost with robust adjustment for differences in severity of illness. Traditional methods to adjust cost estimates have exclusively used administrative data derived from billing claims to identify comorbidity and complications. Transplantation uniquely has accurate national clinical registry data that can be used to supplement administrative data. Administrative claims from the Vizient, Inc, Clinical Data Base (CDB) were linked with clinical records from the Scientific Registry for Transplant Recipients for 76 liver and 109 kidney transplant programs. Using either or both datasets, we fitted a regression model to the total direct cost of care for 16,649 kidney and 6058 liver transplants. The proportion of variation explained by these risk-adjustment models increased significantly when combined administrative and clinical data were used for kidney (administrative only R2 =.069, clinical only R2 =.047, combined R2 =.14, p<.0001) and liver (administrative only R2 =.28, clinical only R2 =.25, combined R2 =.33, p<.0001). Incorporating accurate clinical data into risk-adjustment methodologies can improve risk adjustment methodologies; however, as majority of variation in cost remains unexplained by these risk-adjustment models further work is needed to accuracy assess transplant value.

INFLUENCE OF DEMOGRAPHIC FACTORS ON RESEARCH PARTICIPATION

INFLUENCE OF DEMOGRAPHIC FACTORS ON RESEARCH PARTICIPATION

Characteristics of Primary Care Providers' Consultations With a Statewide Child Psychiatry Access Program Regarding Autism Spectrum Disorder.

Child Psychiatry Access Programs (CPAPs) provide consultative support for pediatric primary care providers (PCPs) to treat co-morbid mental health (MH) symptoms among patients with Autism Spectrum Disorder (ASD). We examined differences in illness severity, comorbidity, and psychotropic medication use between patients with and without ASD for whom primary care providers sought consultation from Maryland's CPAP. We examined N=3641 Maryland's CPAP consultations from 2012 to 2019; n=311 were consultations for ASD. Demographics, treatment recommendations, diagnoses, and Clinical Global Impression-Severity scores were collected. Patients who received psychotropic medication or psychotherapy by any mental health provider were defined as comanaged. Descriptive statistics and logistic regression were conducted. Sample size for regression was N=1854. Compared with non-ASD, patients with consults for ASD were more often male (P < 0.001), aged 0-5 years (P < 0.001), severely ill (Clinical Global Impression-Severity > 4) (P < 0.001), and prescribed stimulants, nonstimulant attention-deficit hyperactivity disorder medications, and antipsychotic medications (P < 0.001). Controlling for key covariates, consultations for youth displaying aggression were 3.02 times (P < 0.001) more likely, with Clinical Global Impression-Severity > 4 were 2.36 times (P < 0.001) more likely, and prescribed antipsychotics were 4.30 times more likely to concern an ASD patient (P<0.001). A larger proportion of ASD patients (vs. non-ASD) had comorbid psychiatric diagnoses of attention-deficit hyperactivity disorder, a learning disability, and disruptive behavior disorder vs. a smaller proportion with major depressive disorder. ASD patients for whom primary care providers sought Maryland's CPAP consultation were more severe and complex than non-ASD patients in terms of comorbid diagnoses and medication regimen.

Nurse staffing levels in critical care: The impact of patient characteristics.

Intensive care is one of the most resource-intensive forms of care because seriously ill patients are cared for in units with high staffing levels. Studies show that the number of registered nurses (RNs) per patient and nurse education level affects patient outcome. However, there is a lack of studies that consider how nurses/patient ratio with an advanced educational level of specialized nurses in intensive care, affect the intensive care performed in different patient populations. To investigate if differences in patient characteristics and nurse-patient ratio have an impact on the quality of care. This is a retrospective observational study with a review of all patients >15 years receiving care at two general intensive care units with different nurse/patient ratio (unit A, 1:1 nurse/patient ratio and unit B, 0.5:1 nurse/patient ratio). There was no significant difference in the initial severity of illness between the units. However, younger patients, male patients and patients requiring surgery entailed a higher workload and a longer intensive care unit (ICU) stay despite a 1:1 critical care nurse/patient ratio. A small difference, but not significant, with more unplanned re-intubations occurred at unit A compared with unit B. The differences in the nurse/patient ratio did not reflect a difference in the severity of illness among admitted patients but might be explained by patient characteristics with different needs. Health care managers should consider not only the number of nurses but also their educational level, specific competencies and skills mix and nursing-sensitive measures to provide high-quality ICU care in settings with different patient characteristics. Nursing-sensitive patient outcomes should be considered in relation to nurse/patient ratio, as important to measure to ensure a high quality of patient care in the ICU.

Effective presence of antibodies against common human coronaviruses in immunoglobulin medicinal products

BackgroundImmunoglobulin products (for intravenous, intramuscular and subcutaneous administration) prepared from geographically diverse plasma pools were tested for activity against common human coronaviruses (HCoVs). Products from plasma obtained from Germany, Czech Republic, Slovak Republic, USA and Spain were tested for antibodies to common HCoVs: 229E, OC43, NL63 and HKU1. As these products are manufactured from pooled plasma from thousands of donors, the antibodies therein are representative of HCoV exposure in the population at large.MethodsImmunoglobulin products were tested for antibodies to four common HCoVs by enzyme-linked immunosorbent assays (ELISAs). Neutralization assays were conducted using HCoV-229E cultured on to MRC5 cells.ResultsELISAs showed that when expressed as specific activity (anti-HCoV activity/mg immunoglobulin), similar activity against the four common HCoVs was seen across the immunoglobulin products regardless of concentration or geographic origin. Highest anti-HCoV activity was seen against HCoV-229E, followed by HCoV-OC43, HCoV-NL63 and HCoV-HKU1. The neutralization assays showed similar potency for two immunoglobulin products prepared by different processes.ConclusionsTo the authors’ knowledge, this is the first demonstration of antibodies to common HCoVs in immunoglobulin products. These results may explain the cross-reactivity seen with pre-pandemic immunoglobulin products and severe acute respiratory syndrome coronavirus-2, and contribute to differences in severity of illness between patients.

Is Discretionary Care Associated with Safety Among Medicare Beneficiaries Undergoing Spine Surgery?

Spine surgery and its corresponding costs have increased in recent years and are variable across geographic regions. Discretionary care is the component of spending variation that is independent of illness severity, age, and regional pricing. It is unknown whether greater discretionary care is associated with improved safety for patients undergoing spine surgery, as we would expect from value-based health care. We conducted an analysis of 5 spine surgery cohorts based on Medicare claims from 2013 to 2017. Patients were grouped into quintiles based on the Dartmouth Atlas End-of-Life Inpatient Care Index (EOL), reflecting regional spending variation attributed to discretionary care. Multivariable regression examined the association between discretionary care and safety measures while controlling for age, sex, race, comorbidity, and hospital features. We observed a threefold to fourfold variation in 90-day episode-of-care cost across regions, depending on the cohort. Spine-specific spending was correlated with EOL quintile, confirming that spending variation is due more to discretionary care than it is to pricing, age, or illness severity. Greater spending across EOL quintiles was not associated with improved safety, and, in fact, was associated with poorer safety in some cohorts. For example, all-cause readmission was greater in the high-spending EOL quintile relative to the low-spending EOL quintile among the "fusion, except cervical" cohort (14.2% vs. 13.1%; OR = 1.10; 95% CI = 1.05 to 1.20), the "complex fusion" cohort (28.0% vs. 25.4%; OR = 1.15; 95% CI = 1.01 to 1.30), and the "cervical fusion" cohort (15.0% vs. 13.6%; OR = 1.12; 95% CI = 1.05 to 1.20). Wide variation in spending was not explained by differences in illness severity, age, or pricing, and increased discretionary care did not enhance safety. These findings point to inefficient use of health-care resources, a potential focus of reform. Economic and Decision Analysis Level IV. See Instructions for Authors for a complete description of levels of evidence.

1228. Outcomes Associated with Empiric Cefepime or Meropenem for Bloodstream Infections Caused by Ceftriaxone-Resistant, Cefepime-Susceptible Escherichia coli and Klebsiella pneumoniae

BackgroundCefepime is a 4th generation cephalosporin frequently used for empiric sepsis therapy. Dose- and MIC-dependent efficacy of cefepime is supported by the Clinical & Laboratory Standards Institute, however its use in infections due to extended-spectrum beta-lactamase-producing Enterobacterales is controversial. This study aims to compare outcomes in patients given empiric meropenem or cefepime for bloodstream infections (BSI) caused by ceftriaxone-resistant E. coli and K. pneumoniae.MethodsThis single-center retrospective cohort included adults hospitalized from 2010 - 2020 and received empiric cefepime or meropenem for BSI caused by ceftriaxone-resistant E. coli or K. pneumoniae. In the cefepime group, only organisms with MIC ≤ 2 mg/L were included. Patients who received the empiric agent for < 48 hours, or received an additional active agent within 48 hours were excluded. The primary outcome was 30-day mortality; secondary outcomes were recurrent infection, readmission, and time to clinical stability. Chi-squared or Fisher’s exact was used for categorical variables and Mann-Whitney-U for continuous variables. Inverse probability treatment weighing was used to determine the impact of empirical therapy on clinical stability at 48 hours. ResultsFifty-four patients were included: 36 received empiric meropenem, 18 received cefepime. There were no significant differences in baseline severity of illness or comorbid conditions. Urinary source was less common in the meropenem group compared to cefepime (52.8 vs 83.8%, p=0.028) (Table 1). There was no difference in 30-day mortality between meropenem and cefepime (2.8 vs 11.1%, p = 0.255). More patients achieved clinical stability at 48 hours on empiric meropenem compared to cefepime (75 vs 44.4%, p = 0.027), and time to clinical stability was significantly shorter (median 21.3 vs 38.5 hours, p = 0.016). Most patients in the meropenem and cefepime groups completed definitive treatment with a carbapenem (88.9 vs 72.2%, p=0.142). Table 1: Results Summary of primary and secondary outcomesConclusionThere was no difference in mortality between patients receiving empiric cefepime for BSI due to ceftriaxone-resistant Enterobacterales, with cefepime MIC ≤ 2 mg/L, compared to meropenem; however, time to clinical stability was significantly delayed. Disclosures James Kirby, MD, D(ABMM), First Light Biosciences (Board Member)TECAN, Inc. (Research Grant or Support)

Sex differences in illness severity and mortality among adult intensive care patients: A systematic review and meta-analysis.

To investigate the association between sex and illness severity and mortality of ICU patients. We performed systematic searches of MEDLINE and EMBASE for observational studies of adult ICU patients that explicitly examined the association between sex and illness severity or mortality. We used a random effects model to calculate standardised mean differences in illness severity scores and pooled odds ratios for mortality of women compared to men. We identified 21 studies with 505,138 participants in total (43.1% women). There was substantial heterogeneity among studies. Only two studies were at low risk of bias overall. At ICU admission, there was a pattern of higher illness severity scores among women (standardised mean difference 0.04, 95% CI -0.01-0.09). Women had higher risk-adjusted mortality than men at ICU discharge (OR 1.25 95% CI 1.03-1.50) and 1 year (OR 1.08, 95% CI 1.02-1.13), however this finding was not robust to sensitivity analysis. Women tend to have higher illness severity scores at ICU admission. Women also appear to have higher risk-adjusted mortality than men at ICU discharge and at 1 year. Given the heterogeneity and risk of bias in the existing literature, additional studies are needed to confirm or refute these findings.

Abstracts of the 30th Annual Pediatric Pharmacy Association (PPA) Meeting April 21 to 25, 2021

Abstracts of the 30th Annual Pediatric Pharmacy Association (PPA) Meeting April 21 to 25, 2021

Differences in Illness Severity among Circulating Norovirus Genotypes in a Large Pediatric Cohort with Acute Gastroenteritis.

Norovirus is a major pathogen identified in children with acute gastroenteritis (AGE), little is known about the strain’s diversity and their clinical severity. Stool and/or rectal swabs were collected from children ≤18 years of age recruited at emergency departments (ED), and a provincial nursing advice phone line due to AGE symptoms in the province of Alberta, Canada between December 2014 and August 2018. Specimens were tested using a reverse transcription real time PCR and genotyped by Sanger sequencing. The Modified Vesikari Scale score (MVS) was used to evaluate the disease severity. The objectives are to identify the Genogroup and Genotype distribution and to compare illness severity between the GI and GII genogroups and to complete further analyses comparing the GII genotypes identified. GII.4 was the genotype most commonly identified. Children with GII.4 had higher MVS scores (12.0 (10.0, 14.0; p = 0.002)) and more prolonged diarrheal (5 days (3.0, 7.8)) and vomiting (3.2 days (1.7, 5.3; p < 0.001)) durations compared to other non GII.4 strains. The predominant strain varied by year with GII.4 Sydney[P31] predominant in 2014/15, GII.4 Sydney[P16] in 2015/16 and 2017/18, and GII.3[P12] in 2016/17. Genogroup II norovirus strains predominated in children with AGE with variance between years; clinical severity associated with different strains varied with episodes being most severe among GII.4 infected children.

HIV and risk of dementia in older veterans.

People living with HIV (PLWH) may be at an increased risk for dementia as they age. Surprisingly, it remains unclear whether PLWH have a higher risk of developing dementia in late life than those without. We explored whether HIV-infection is associated with incident dementia diagnosis in older U.S. veterans accounting for potential confounders and competing risk of death. We included 1114 veterans diagnosed with HIV, ages at least 55 years (mean = 62 years, SD = 6), followed in the Veterans Health Administration healthcare system from 2004 to 2015, and a propensity-matched comparison group (n = 1114) without HIV. HIV and dementia diagnoses were determined using electronic medical records. Using Fine-Gray proportional hazards models, we examined whether HIV status was associated with a greater risk of incident dementia. During follow-up (mean = 7 years, SD = 4 from date of HIV diagnosis), 5% of veteran PLWH developed dementia compared with 3% without (P = 0.01). Accounting for the competing risk of death and adjusted for demographics, substance use, education and income, PLWH remained 50% more likely to receive a dementia diagnosis [adjusted hazard ratio (aHR) = 1.50, 95% confidence interval 0.96-2.35]. Although combination antiretroviral therapy (cART) exposure was associated with an increased risk of incident dementia, this was driven by differences in illness severity as captured by CD4 cell count. There was no evidence of a differential effect by cART class. In a cohort of older USA veterans, HIV infection increased risk of dementia by 50%, while exposure to cART did not offset this risk. It is critical to understand the mechanisms by which HIV increases risk for developing dementia in later life.

S243. RACIAL AND ETHNIC DIFFERENCES IN PATHWAY TO CARE AND BASELINE CHARACTERISTICS IN EARLY INTERVENTION SERVICES FOR PSYCHOSIS IN NEW YORK STATE

BackgroundThe racial and ethnic background of individuals with psychosis may shape their pathway to early intervention services and clinical presentation at admission. Studies from Europe and Canada demonstrate that black minority patients with first-episode psychosis experience a more adverse and coercive pathway to care. The extent to which these findings can be extrapolated to the US context is unknown. The aims of this study are (1) to compare baseline contextual and clinical characteristics, and (2) to examine care pathways by race and ethnicity among young people with psychosis in early intervention services.MethodsThis study included individuals with a recent-onset (<2 years) psychosis aged 16 to 30 years enrolled at 19 early intervention programs across New York State. Clinicians collected data on pathway to care, demographic, social and clinical variables at program entry. Level of functioning was assessed using the social, occupational and symptomatic functioning subscales of the MIRECC GAF.ResultsThe sample included 767 individuals with a non-Hispanic white (n=209, 27.2%), non-Hispanic black (277, 36.1%), Hispanic (218, 28.4%), or Asian (63, 8.2%) racial/ethnic background. Compared to non-Hispanic white, minority individuals were more likely to have public or no insurance and, overall, had a lower level of completed education. In terms of pathway to care, a lower proportion of non-Hispanic black (65.7%) and Asian (58.7%) participants had previously used mental health services compared to the non-Hispanic white group (78.0%). In contrast, psychiatric hospital or emergency department admissions in the 90 days prior to program enrollment were more frequent among all minority groups in comparison to the non-Hispanic white group. There were no significant differences by race and ethnicity in the level of symptoms or social functioning at baseline.DiscussionOur findings suggest a pattern of mental health service use among minority groups with psychosis characterized by less mental health contacts but more inpatient and emergency care prior to the initiation of early intervention services. This trend could be partly explained by racial and ethnic patterning at the contextual level, including financial barriers to care, less so by racial/ethnic differences in illness severity. Our findings are consistent with evidence demonstrating an overrepresentation of minority individuals, especially African-Americans, in psychiatric emergency services suggesting a gap in unmet mental health need among minority populations in the US.

The Impact of Step-Down Unit Care on Patient Outcomes After ICU Discharge.

To examine whether and how step-down unit admission after ICU discharge affects patient outcomes. Retrospective study using an instrumental variable approach to remove potential biases from unobserved differences in illness severity for patients admitted to the step-down unit after ICU discharge. Ten hospitals in an integrated healthcare delivery system in Northern California. Eleven-thousand fifty-eight episodes involving patients who were admitted via emergency departments to a medical service from July 2010 to June 2011, were admitted to the ICU at least once during their hospitalization, and were discharged from the ICU to the step-down unit or the ward. None. Using congestion in the step-down unit as an instrumental variable, we quantified the impact of step-down unit care in terms of clinical and operational outcomes. On average, for ICU patients with lower illness severity, we found that availability of step-down unit care was associated with an absolute decrease in the likelihood of hospital readmission within 30 days of 3.9% (95% CI, 3.6-4.1%). We did not find statistically significant effects on other outcomes. For ICU patients with higher illness severity, we found that availability of step-down unit care was associated with an absolute decrease in in-hospital mortality of 2.5% (95% CI, 2.3-2.6%), a decrease in remaining hospital length-of-stay of 1.1 days (95% CI, 1.0-1.2 d), and a decrease in the likelihood of ICU readmission within 5 days of 3.6% (95% CI, 3.3-3.8%). This study shows that there exists a subset of patients discharged from the ICU who may benefit from care in an step-down unit relative to that in the ward. We found that step-down unit care was associated with statistically significant improvements in patient outcomes especially for high-risk patients. Our results suggest that step-down units can provide effective transitional care for ICU patients.

Highlights From the Circulation Family of Journals.

Highlights From the Circulation Family of Journals.

Utilization and outcomes in biventricular assist device support in pediatrics

ObjectivePatients with biventricular assist devices (BiVADs) have worse outcomes than those with left ventricular assist devices (LVADs). It is unclear whether these outcomes are due to device selection or patient factors. We used propensity score matching to reduce patient heterogeneity and compare outcomes in pediatric patients supported with BiVADs with a similar LVAD cohort. MethodsThe Pedimacs registry was queried for patients who were supported with BiVAD or LVAD. Patients were analyzed by BiVAD or LVAD at primary implant and the 2 groups were compared before and after using propensity score matching. ResultsOf 363 patients who met inclusion criteria, 63 (17%) underwent primary BiVAD support. After propensity score matching, differences between cohorts were reduced. Six months after implant, in the BiVAD cohort (LVAD cohort) 52.5% (42.5%) had been transplanted; 32.5% (40%) were alive with device, and 15% (10%) had died. Survival was similar between cohorts (P = .31, log-rank), but patients with BiVADs were more likely to experience a major adverse event in the form of bleeding (P = .04, log-rank). At 1 week and 1 and 3 months' postimplant, the percentage of patients on mechanical ventilation, on dialysis, or with elevated bilirubin was similar between the 2 groups. ConclusionsWhen propensity scores were used to reduce differences in patient characteristics, there were no differences in survival but more major adverse events in the patients with BiVADs, particularly bleeding. Differences in unmatched patient outcomes between LVAD and BiVAD cohorts likely represent differences in severity of illness rather than mode of support.