IMPORTANCE:A better understanding of the relationship between sedative dose and PICU delirium is crucial for delirium prevention. OBJECTIVES:To identify risk factors for delirium in critically ill children and characterize the dose-response relationship between benzodiazepine administration and next-day delirium in the PICU. DESIGN, SETTING, AND PARTICIPANTS:A secondary analysis of 10,644 patient-days from prospectively collected data, including sedative and analgesic use and delirium screening from the multicenter, randomized PICU Up! pilot trial, including seven PICUs and 1,006 patients. MAIN OUTCOMES AND MEASURES:Delirium odds were evaluated using a multivariable logistic regression model adjusted for known delirium risk factors. RESULTS:Benzodiazepines, opioids, and dexmedetomidine were commonly used in the PICU. A higher previous-day benzodiazepine dose was associated with increased odds of delirium on a given day with an adjusted odds ratio (aOR) of 1.12 per midazolam milligram equivalent per kilogram (95% CI, 1.02-1.25; p = 0.020). The most pronounced increase in delirium risk occurred at the lowest doses, between 0 and 0.01 mg/kg (p = 0.010), with risk continuing to rise at higher doses. Previous-day opioid dose was not associated with changes in the odds of delirium (aOR, 1.00; 95% CI, 1.00-1.00; p = 0.892). Any previous-day dexmedetomidine administration was associated with increased odds of next-day delirium (aOR, 1.26; 95% CI, 1.04-1.52; p = 0.019). Additionally, same-day physical restraint use was associated with increased odds of delirium (aOR, 1.72; 95% CI, 1.27-2.31; p < 0.001). CONCLUSIONS AND RELEVANCE:No safe benzodiazepine threshold was identified in critically ill children, with delirium risk increased even at minimal doses. The greatest increase in odds occurred at the lowest doses. Hence, regardless of dose, benzodiazepine use should be minimized or eliminated whenever possible. Physical restraints and dexmedetomidine use were also associated with increased risk of delirium; thus, a tailored approach for each patient with an emphasis on judicious use is critical.

OBJECTIVES:To determine the patient-level factors associated with performing daily delirium screening in PICUs with established delirium screening practices. DESIGN:A secondary analysis of 2019-2020 prospective data from the baseline phase of the PICU Up! pilot stepped-wedge multicenter trial (NCT03860168). SETTING:Six PICUs in the United States. PATIENTS:One thousand sixty-four patients who were admitted to a PICU for 3 or more days. INTERVENTIONS:None. MEASUREMENTS AND MAIN RESULTS:Of 1064 patients, 74% (95% CI, 71-76%) underwent delirium screening at least once during their PICU stay. On 57% of the 8965 eligible patient days, screening was conducted. The overall prevalence of delirium was 46% across all screened days, and 64% of screened patients experienced delirium at some point during their PICU stay. Factors associated with greater adjusted odds ratio (aOR) of increased daily delirium screening included PICU stay longer than 15 days compared with 1-3 days (aOR 3.36 [95% CI, 2.62-4.30]), invasive mechanical ventilation as opposed to room air (aOR 1.67 [95% CI, 1.32-2.12]), dexmedetomidine infusions (aOR 1.23 [95% CI, 1.04-1.44]) and propofol infusions (aOR 1.55 [95% CI, 1.08-2.23]). Conversely, decreased aOR of daily delirium screening was associated with female gender (aOR 0.78 [95% CI, 0.63-0.96]), and the administration of continuous infusions of opioids (aOR 0.75 [95% CI, 0.63-0.90]) or ketamine (aOR 0.48 [95% CI, 0.29-0.79]). Neither patient age, the presence of family or physical restraints, or benzodiazepine infusions were associated with daily delirium screening rates. CONCLUSIONS:In the 2019-2020 PICU UP! cohort, across six PICUs, delirium screening occurred on only 57% of days, despite the presence of established practices. Female gender, patients in the early stages of their PICU stay, and patients not receiving mechanical ventilation were associated with lower odds of daily delirium screening. Our results highlight the need for structured quality improvement processes to both standardize and increase the frequency of delirium screening.

Blackman, Isabella1; De Armas, Rachel2; Padilla, Sofia3; Goldberg, Sarah4; Abu-Sultaneh, Samer5; Piole, Michelle6; Geno, Megan7; Manning, Mary-Jeanne7; Madden, Kate7; Madsen, Erik6; Behrens, Deanna8; Maue, Danielle9; Mennie, Colleen10; Furniturewala, Sukaina11; Awojoodu, Ronke10; Needham, Dale10; Kudchadkar, Sapna12; Azamfirei, Razvan10 Author Information

Azamfirei, Razvan; Mennie, Colleen; Awojoodu, Ronke; Behrens, Deanna; Padilla, Sofia; Madden, Kate; Goldberg, Sarah; Geno, Megan; Manning, Mary-Jeanne; Piole, Michelle; Madsen, Erik; Maue, Danielle; Abu-Sultaneh, Samer; Needham, Dale; Kudchadkar, Sapna Author Information

Introduction: NORSE refers to new onset refractory status epilepticus in patients without known epilepsy or seizure etiology. When patients fail conventional therapy, alternative methods are pursued. Electroconvulsive therapy (ECT) has been occasionally used, but rarely in children. We report a previously healthy 8-year-old female with NORSE whose epileptic burden improved with aggressive therapies including ECT. Description: Our patient presented with a week of headaches and fever before developing multiple focal seizures that rapidly evolved into RSE. Extensive infectious, autoimmune, metabolic, neoplastic, and genetic work up was unrevealing. Sequential brain MRIs remained normal. Aggressive anti-seizure medications (ASMs), pentobarbital coma, high dose propofol and ketamine infusions, pulse-dose steroids, IVIg, ketogenic diet, cannabidiol, and rituximab were initiated during her first 12 hospital days. Despite these measures, EEG remained highly epileptic prior to ECT initiation on day 12. ECT was performed daily for 7 days with two stimulations per session using a Thymatron device with bitemporal electrode placement. ECT achieved seizures in all treatments with median post-ictal suppression of 70.5%, indicating robust seizure quality. All ASM infusions were weaned one day post-ECT completion, and the patient was extubated shortly thereafter. About one week after extubation, a vagal nerve stimulator was placed (hospital day 29), after which point she has remained essentially seizure-free. Discussion: Our 8-year-old with NORSE responded well to a 7-day course of ECT. Multiple ASMs and infusions failed to improve overall epileptic burden. Efficacy of ECT is evidenced by a high post-ictal suppression but is limited by close succession of other treatments including rituximab, cannabidiol, and ketosis.

De Armas, Rachel; Blackman, Isabella; Mennie, Colleen; Awojoodu, Ronke; Maue, Danielle; Behrens, Deanna; Madsen, Erik; Madden, Kate; Manning, Mary-Jeanne; Geno, Megan; Piole, Michelle; Abu-Sultaneh, Samer; Goldberg, Sarah; Padilla, Sofia; Needham, Dale; Kudchadkar, Sapna; Azamfirei, Razvan Author Information

OBJECTIVES:We aimed to determine which characteristics and management approaches were associated with postoperative invasive mechanical ventilation (IMV) and with a prolonged course of IMV in children post liver transplant as well as describing the utilization of critical care resources.DESIGN:Retrospective, multicenter, cohort study of children who underwent an isolated liver transplantation between January 2017 and December 2018.SETTING:Twelve U.S., pediatric, liver transplant centers.PATIENTS:Three hundred thirty children post liver transplant admitted to the ICU.INTERVENTIONS:None.MEASUREMENTS AND MAIN RESULTS:Six patients died in our cohort. The median length of PICU stay was 4.5 days (interquartile range [IQR], 2.9–8.2 d). Most patients were initially monitored with arterial catheters (96%), central venous pressures (95%), and liver ultrasound (93%). Anticoagulation (80%), blood product administration (52.4%), and vasoactive agents (23.0%) were commonly used therapies in the first 7 days. In multivariable logistic regression analysis, age (adjusted odds ratio [aOR] 0.9 [0.86–0.95]), open fascia (aOR 7.0 [95% CI, 2.6–18.9]), large center size (aOR 4.3 [95% CI 2.2–8.3]), and higher Model for End-Stage Liver Disease/Pediatric End-Stage Liver Disease scores (aOR 1.04 [95% CI, 1.01–1.06]) were associated with postoperative IMV. In multivariable logistic regression analysis, postoperative day 0 peak inspiratory pressure (PIP) (aOR 1.2 [95% CI, 1.1–1.3]), large center size (aOR 2.9 [95% CI, 1.6–5.4]), and age (aOR 0.89 [95% CI, 0.85–0.95]) were associated with length of IMV greater than 24 hours. Length of IMV greater than 24 hours was associated with bleeding complications (p= 0.03), infections (p= 0.03), graft loss (p= 0.02), and reoperation (p= 0.03).CONCLUSIONS:Younger age, preoperative hospitalization, large center size, and open fascia are associated with use of IMV, and younger age, large center size, and postoperative day 0 PIP are associated with prolonged IMV on multivariable analysis. Longer IMV is associated with negative outcomes, making it an important clinical marker.

Introduction: This study aimed to determine if a respiratory therapist (RT)-driven high flow nasal cannula (HFNC) protocol could decrease duration of HFNC use, pediatric intensive care unit (PICU) and hospital length of stay (LOS), and duration of continuous albuterol use in pediatric patients with critical asthma. Methods: This was a quality improvement project performed at a quaternary academic PICU. Patients admitted to the PICU between 2 and 18 years of age with a diagnosis of asthma requiring continuous albuterol and HFNC were included. Implementation of a RT-driven HFNC protocol [Plan-Do-Study-Act (PDSA) 1] occurred in October 2017. Additional interventions included weaning continuous albuterol and HFNC simultaneously (PDSA 2; March 2019), adjusting HFNC wean rate (PDSA 3; July 2020), and a HFNC holiday (PDSA 4; October 2021). HFNC duration was the primary outcome. Secondary outcomes included LOS data and continuous albuterol duration. Noninvasive ventilation (NIV), invasive mechanical ventilation (IMV), and 7-day PICU and hospital readmission rates were balancing measures. Results: 410 patients were included. Patient demographics and adjunct therapy use did not differ among the groups. HFNC duration decreased from 26.8 to 18.1 hours, both PICU and hospital LOS were decreased (41 to 31.8 hours, and 86.5 to 68 hours respectively) after PDSA 2. These outcomes remained stable during PDSA 3 and 4. Continuous albuterol duration and NIV use remained stable, while IMV use decreased throughout the study. Conclusions: An RT-driven HFNC protocol led to an improvement in clinical outcomes for pediatric patients with critical asthma without an increase in adverse events.

Background:Vascular thromboses (VT) are life-threatening events after pediatric liver transplantation (LT). Single-center studies have identified risk factors for intra-abdominal VT, but large-scale pediatric studies are lacking.Methods:This multicenter retrospective cohort study of isolated pediatric LT recipients assessed pre- and perioperative variables to determine VT risk factors and anticoagulation-associated bleeding complications.Results:Within seven postoperative days, 31/331 (9.37%) patients developed intra-abdominal VT. Open fascia occurred more commonly in patients with VT (51.61 vs 23.33%) and remained the only independent risk factor in multivariable analysis (OR = 2.84, p = 0.012). Patients with VT received more blood products (83.87 vs 50.00%), had significantly higher rates of graft loss (22.58 vs 1.33%), infection (50.00 vs 20.60%), and unplanned return to the operating room (70.97 vs 16.44%) compared to those without VT. The risk of bleeding was similar (p = 0.2) between patients on and off anticoagulation.Conclusions:Prophylactic anticoagulation did not increase bleeding complications in this cohort. The only independent factor associated with VT was open fascia, likely a graft/recipient size mismatch surrogate, supporting the need to improve surgical techniques to prevent VT that may not be modifiable with anticoagulation.

Early-onset ventilator-associated pneumonia (VAP) is associated with poor outcomes in patients with severe traumatic brain injury (TBI). The primary aim of this study was to describe VAP, including the microbiology of VAP and differences in frequency of VAP when various definitions are applied. The secondary aim was to determine the clinical variables associated with the development of VAP in children with severe TBI. This is a retrospective cohort study at a quaternary referral children’s hospital with a level I trauma center designation. Inclusion criteria were patients aged 0–18 years admitted to the pediatric intensive care unit between 2015 and 2020 with severe TBI requiring at least 2 days of invasive ventilation. VAP was defined by using Center of Disease Control (CDC) definition or clinical VAP, based on physician diagnosis. We compared general demographics, reviewed trauma and injury data, and outcomes to assess any differences between patients with VAP and non-VAP patients. Associations were tested with regression models. After applying all inclusion and exclusion criteria, 90 patients were included in the analysis. Patients with VAP were older (8.5 vs. 5.6 years, P = 0.03). Patients with VAP were less likely to have suffered from abusive head trauma (P = 0.01). Patients who received continuous neuromuscular blockade or targeted temperature management did not have different frequencies of VAP. CDC-defined VAP was diagnosed in 27

BACKGROUND:The Critical Care Societies Collaborative included not ordering diagnostic tests at regular intervals as one of their Choosing Wisely initiatives. A reduction in unnecessary chest radiographs (CXRs) can help reduce exposure to radiation and eliminate health care waste. We aimed to reduce daily screening CXRs in a pediatric ICU (PICU) by 20% from baseline within 4 months of implementation of CXR criteria. METHODS:All intubated patients in the PICU were included in this quality improvement project. Patients with tracheostomies were excluded. We developed criteria delineating which patients were most likely to benefit from a daily screening CXR, and these criteria were discussed for each patient on rounds. Patients on extracorporeal membrane oxygenation, on high-frequency oscillatory ventilation, or on high support on conventional mechanical ventilation were included as needing a daily screening CXR. We tracked the percentage of intubated subjects receiving a screening CXR as an outcome measure. Unplanned extubations and the number of non-screening CXRs per intubated subject were followed as balancing measures. RESULTS:The percentage of intubated subjects receiving a daily screening CXR was reduced from 79% to 31%. There was no increase in frequency of unplanned extubations or number of non-screening CXRs. With an estimated subject charge of roughly $270 and hospital cost of $54 per CXR, this project led to an estimated $300,000 in patient charge savings and $60,000 in hospital cost savings. CONCLUSIONS:Adopting criteria to delineate which patients are most likely to benefit from screening CXRs can lead to a reduction in the percentage of intubated patients receiving screening CXRs without appearing to increase harm.

Introduction: Costs for bronchiolitis have risen over the years. Previous studies have shown that respiratory therapist (RT) driven high flow nasal cannula (HFNC) management protocols can decrease length of HFNC as well as pediatric intensive care unit (PICU) and hospital length of stay (LOS). The aim of this study was to determine if modifications to an existing RT-driven HFNC management protocol could further safely decrease length of HFNC and PICU and hospital LOS. Methods: This is a quality improvement project performed at a quaternary academic PICU. This study included patients less than 24 months of age admitted to the PICU with an admission diagnosis of bronchiolitis requiring HFNC. After initial implementation of a (RT)-driven HFNC protocol [Plan-Do-Study-Act (PDSA) 1] in October 2017, additional interventions included adjusting HFNC wean rate (PDSA 2) in July 2020, a HFNC holiday (PDSA 3), and standardized discharge criteria (PDSA 4) in October 2021. Duration of HFNC was used as the primary outcome measure and PICU LOS and hospital LOS were used as secondary outcome measures. Noninvasive ventilation use, invasive mechanical ventilation use and 7-day PICU and hospital readmission rates were used as balancing measures. Results: A total of 1,310 patients were included. Patients in PDSA 2, PDSA 3&4 groups were older compared to pre-intervention and PDSA 1 (median of 9 and 10 months compared to 8 months, p=0.01). HFNC duration decreased from 2.5 to 1.8 days after PDSA 1, then to 1.3 days after PDSA 2. PICU LOS decreased from 2.6 to 2.1 days after PDSA 1, 1.8 days after PDSA 2 and 1.5 days after PDSA 3&4. Hospital LOS decreased from 5.7 days to 4.5 days after PDSA 1, 3.1 days after PDSA 2 and 2.7 days after PDSA 3&4. The use of non-invasive ventilation and invasive mechanical ventilation decreased throughout the study from 23.2% in the pre-intervention group, to 6.9% at end of the project. The 7-day PICU and hospital readmission rates did not increase. The percentage of patients discharged from the PICU increased from 6.2% to 21.5%. Conclusions: Modifications to an existing RT-driven HFNC protocol and standardized discharge criteria led to an improvement in bronchiolitis patients’ outcomes without an increase in adverse events.

OBJECTIVES: To decrease length of high-flow nasal cannula (HFNC), PICU, and hospital length of stay (LOS). DESIGN: Quality improvement project. SETTING: A quaternary academic PICU. PATIENTS: Patients with bronchiolitis less than 24 months old. INTERVENTIONS: After initial implementation of a respiratory therapist (RT)-driven HFNC protocol (Plan-Do-Study-Act [PDSA] 1) in October 2017, additional interventions included adjusting HFNC wean rate (PDSA 2) in July 2020, a HFNC holiday (PDSA 3), and standardized discharge criteria (PDSA 4) in October 2021. MEASUREMENTS AND MAIN RESULTS: Duration of HFNC was used as the primary outcome measure. PICU LOS and hospital LOS were used as secondary outcome measures. Noninvasive ventilation use, invasive mechanical ventilation use, and 7-day PICU and hospital readmission rates were used as balancing measures. A total of 1,310 patients were included in this study. Patients in PDSA 2, PDSA 3 and 4 groups were older compared with pre-intervention and PDSA 1 (median of 9 and 10 mo compared with 8 mo; p = 0.01). HFNC duration decreased from 2.5 to 1.8 days after PDSA 1, then to 1.3 days after PDSA 2. PICU LOS decreased from 2.6 to 2.1 days after PDSA 1, 1.8 days after PDSA 2, and 1.5 days after PDSA 3 and 4. Hospital LOS decreased from 5.7 to 4.5 days after PDSA 1, 3.1 days after PDSA 2, and 2.7 days after PDSA 3 and 4. The use of noninvasive ventilation and invasive mechanical ventilation decreased throughout the study from 23.2% in the pre-intervention group, to 6.9% at the end of the project. The 7-day PICU and hospital readmission rates did not increase after implementation. The percentage of patients discharged from the PICU increased from 6.2% to 21.5%. CONCLUSIONS: Modifications to an existing RT-driven HFNC protocol and standardization of discharge criteria led to an improvement in outcomes for patients admitted to the PICU with bronchiolitis without an increase in adverse events.

Maue, Danielle1; Martinez, Mercedes2; Beltramo, Fernando3; Alcamo, Alicia4; Ridall, Leslie5; Nares, Michael6; Jeyapalan, Asumthia7; Zinter, Matthew8; Betters, Kristina9; Kamath, Sameer10; Monde, Alexandra11; Kaushik, Shubhi12; Resch, Joseph13; Kang, Elise14; Pike, Francis15; Mangus, Richard16; Rowan, Courtney17 Author Information

Diabetic ketoacidosis (DKA) is known to cause total body potassium depletion, but during initial presentation, very few patients are hypokalemic, and even fewer patients experience clinical effects. As the correction of acidosis and insulin drive potassium intracellularly, measured serum potassium levels decrease and require repletion. This phenomenon is well described, and severe hypokalemia necessitates delaying insulin therapy. Less well described is the kaliuretic nature of treatments of cerebral edema. We present a case of an adolescent male with new-onset type 2 diabetes who presented in DKA with signs of cerebral edema, hyperosmolarity, and hypokalemia. As insulin and cerebral edema therapy were initiated, his hypokalemia worsened despite significant IV repletion, eventually leading to ventricular tachycardia and cardiac arrest. Over the following 36 hours, the patient received >590 milliequivalents (mEq) of potassium. He was discharged home 12 days after admission without sequelae of his cardiac arrest.

Azamfirei, Razvan1; Awojoodu, Ronke1; Abu-Sultaneh, Samer2; Behrens, Deanna3; Geno, Megan4; Goldberg, Sarah4; Madden, Kate4; Madsen, Erik5; Manning, Mary-Jeanne4; Maue, Danielle2; Padilla, Sofia6; Piole, Michelle5; Kudchadkar, Sapna1 Author Information

Advancements in critical care management have led to improvement in pediatric LT outcomes. However, there are no specific guidelines for many aspects of immediate post-LT care. This survey examines practice variations in the immediate postoperative care of pediatric LT patients at a large number of active US centers. This study is a cross-sectional survey of medical directors at PALISI-affiliated PICU in the United States. Centers performing pediatric LT were analyzed. Study measures included PICU practices regarding staffing, composition of the multidisciplinary team, early post-LT graft and patient monitoring, and anticoagulation. Of the thirty-five responding centers, twenty-five had a LT program which accounted for one-half of all US pediatric LTs. For analysis, centers were categorized by volume: high (7), medium (11), and low (7). The majority of PICU teams included an intensivist (80%) and hepatologist (84%). High-volume centers were less likely to have 24-hour in-house attending coverage (29%, compared to 64% (medium) and 100% (low)). High-volume centers were most likely to have pre-printed orders, but least likely to have written PICU management protocols. Most centers utilize routine daily liver ultrasound. Routine prophylactic anticoagulation, and the agent of choice, was variable. There is marked inconsistency in post-LT practice across PALISI centers in regards to team composition and immediate post-LT management. A national US consensus for post-LT PICU practices would facilitate outcomes research and would establish a platform for multicenter studies.

Maue, Danielle; Alcamo, Alicia; Beltramo, Fernando; Betters, Kristina; Kamath, Sameer; Kaushik, Shubhi; Mangus, Richard; Martinez, Mercedes; Monde, Alexandra; Nares, Michael; Resch, Joseph; Ridall, Leslie; Zinter, Matthew; Rowan, Courtney Author Information