Improving Pediatric Intensive Care Unit Discharge Timeliness of Infants with Bronchiolitis Using Clinical Decision Support

Abstract

Background Identifying children ready for transfer out of the pediatric intensive care unit (PICU) is an area that may benefit from clinical decision support (CDS). We previously implemented a quality improvement (QI) initiative to accelerate the transfer evaluation of non–medically complex PICU patients with viral bronchiolitis receiving floor-appropriate respiratory support.

Objectives Design a CDS tool adaptation of this QI initiative to further accelerate transfer evaluation of appropriate patients.

Methods The original initiative focused on identifying for transfer evaluation otherwise healthy children admitted to the PICU with bronchiolitis who had been receiving floor-appropriate levels of respiratory support for at least 6 hours. However, this initiative required that clinicians manually track the respiratory support of qualifying patients. We designed an electronic health record (EHR)–based CDS tool to automate identification of transfer-ready candidates. The tool parses EHR data to identify children meeting prior QI initiative criteria and alerts clinicians to assess transfer readiness once a child has been receiving floor-appropriate respiratory support for 6 hours. We compared time from reaching floor-appropriate support to placement of the transfer order (“time-to-transfer”), PICU length of stay (LOS), and hospital LOS between patients admitted prior to our QI initiative (December 1, 2018–October 19, 2019, “pre-QI phase”), during the initiative but before CDS tool implementation (October 20, 2019–February 7, 2022, “QI phase”), and after CDS implementation (February 8–November 11, 2022, “CDS phase”).

Results CDS-phase patients (n = 131) had a shorter median time-to-transfer of 5.23 (interquartile range [IQR], 3.38–10.0) hours compared with QI-phase patients (n = 304) at 5.93 (IQR, 4.23–12.2) hours (p = 0.04). PICU and hospital LOS values decreased from the pre-QI (n = 150) to QI phase. Though LOS reductions were sustained during the CDS phase, further reductions from QI to CDS phase were not statistically significant.

Conclusion An EHR-based CDS adaptation of a prior QI initiative facilitated timely identification of PICU patients with bronchiolitis ready for transfer evaluation. Such tools might allow PICU clinicians to focus on other high-acuity tasks while accelerating transfer evaluation of appropriate patients.

Keywords

bronchiolitis - pediatric critical care - clinical decision support - electronic health record - de-escalation

Background and Significance

Though several clinical decision support (CDS) tools have been developed to facilitate timely identification of children at risk for deterioration,[1] [2] [3] [4] less attention has been paid to the development of tools able to identify children with an improving clinical trajectory who may be ready for transfer out of the pediatric intensive care unit (PICU). Such “de-escalation” CDS tools could improve patient care by reducing PICU length of stay (LOS), hospital LOS, and hospital and patient costs and by allowing providers on inpatient acute care wards to practice at the top of their clinical scope.

Viral bronchiolitis can cause acute respiratory failure and is one of the most common causes for admission to a PICU.[5] [6] [7] Such critically ill children are often supported with noninvasive positive pressure ventilation (NIPPV) or high-flow nasal cannula (HFNC) respiratory support modalities. At the time of this report, children in our institution 3 months old to 3 years old must be managed in the PICU if requiring more than 8 L per minute (LPM) of HFNC or if they require NIPPV. Conversely, they can be managed on the acute care inpatient ward if clinically stable on 8 LPM or less of HFNC.

In prior work,[8] we learned that transfer of many children with viral bronchiolitis from our institution's PICU to the acute care ward was not occurring despite hours of clinical stability while receiving 8 LPM or less of HFNC. Manual chart review revealed that non–medically complex children admitted for viral bronchiolitis who remained stable on ≤8 LPM of HFNC for at least 6 hours were very unlikely to require re-escalation of respiratory support and were likely safe for floor transfer. During our team's initial quality improvement (QI) initiative, we further defined this “non–medically complex” viral bronchiolitis population as those children who (1) have no chronic medical conditions, (2) were not intubated during their PICU encounter, and (3) were not managed with continuous albuterol. The last criterion prevented inclusion of children whose primary problem was status asthmaticus. Though the specific definition of a chronic medical condition was left to the discretion of the managing provider in the PICU, provided examples included any history of cardiovascular disease, chronic lung disease, neurologic/neuromuscular disease, and gestational age < 36 weeks at delivery.

During the QI initiative, we established standardized timing for transfer assessment for such children in which the bedside nurse or respiratory therapist (RT) would contact the managing PICU provider to request a clinical evaluation for transfer readiness once the patient had been receiving HFNC of 8 LPM or less for at least 6 hours. If the patient had stable work of breathing, was not requiring suctioning or other nursing interventions more frequently than every 2 hours, and had stable vital signs, then the provider was encouraged to transfer the patient out of the PICU. This new process required the nurse and provider to determine upon admission if the child met the initiative's inclusion/exclusion criteria, the nurse and/or RT to keep track of the amount of time on 8 LPM or less, and for the nurse/RT to contact the provider once 6 hours had elapsed.

This QI initiative was well received, and the median time from reaching 8 LPM of HFNC until the time at which the transfer order was entered (hereafter referred to as the “time-to-transfer”) was reduced by 46% (from 14.4 to 7.8 hours) among qualifying children. In addition, there was no increase in rapid response team activations nor unplanned PICU readmissions; however, PICU and hospital LOS were not significantly different during the initiative. Targeted interviews with PICU clinicians revealed two key challenges. First, providers and nurses often felt they had inadequate time to manually review the patient's medical history in enough depth to confirm that a patient did not have any chronic medical conditions (thus meeting initiative inclusion/exclusion criteria) potentially limiting the population of bronchiolitis patients to whom the intervention was applied. Second, bedside nurses and RTs often found it challenging to contact the managing provider exactly at the 6-hour mark of stability on 8 LPM or less of HFNC given their clinical responsibilities for other high-acuity or deteriorating patients in the PICU.

Given these challenges, we hypothesized that an automated, electronic health record (EHR)–integrated CDS tool adaptation of this QI initiative could further enhance transit outcomes by improving identification of patients meeting our prior initiative's inclusion/exclusion criteria and facilitating timelier evaluation for transfer readiness.

Objectives

The study objective was to develop and implement an EHR-based bronchiolitis transfer CDS tool that would result in decreases in time-to-transfer, PICU LOS, and total hospital LOS, with no increase in unplanned PICU readmission rate, among non–medically complex children admitted to our institution's PICU with viral bronchiolitis.

Methods

Patient Population

We convened a multidisciplinary team that included PICU physicians, advanced practice providers (APPs), nurses, RTs, and informaticists to develop a bronchiolitis transfer CDS tool. Goals of the team included development of a CDS tool capable of automating the identification of patients meeting prior QI initiative inclusion/exclusion criteria, creation of multimodal visualizations of clinical trajectory to aid with patient tracking and patient flow/bed management, and delivery of targeted alerts to accelerate clinical evaluation of transfer readiness.

To automate identification of children meeting the original QI initiative criteria, we collaborated with our Epic software analyst team to design CDS software linked in near real time to our institution's EHR. Patients were identified for inclusion based on their age at the time of PICU admission as listed in the EHR and whether the EHR “Principal Problem” (the primary reason for hospitalization) chosen by the admitting provider was a diagnosis included within a predefined list of viral bronchiolitis diagnoses ([Supplementary Table S1], available in the online version). The CDS tool identified children who required intubation and invasive mechanical ventilation via parsing of each child's lines, drains, and airways (LDA) report for presence of an endotracheal tube. Similarly, the CDS tool parsed the medication administration record for continuous albuterol administration during the PICU encounter and up to 14 days prior to PICU admission. To automate identification of children with chronic medical conditions, we applied an adapted version of the pediatric complex chronic condition (PCCC) criteria developed by Feudtner et al.[9] Specifically, children were excluded if they were identified as having a PCCC based on International Classification of Diseases, 10th revision (ICD-10) procedure and diagnosis codes from prior encounters in our health care system per the PCCC algorithm.[10]

Clinical Decision Support Tool Function and Alerts

Once a patient met the above-outlined criteria, the software would begin digesting data from the respiratory support flowsheets within the EHR to determine when the patient was placed on HFNC of 8 LPM or less (or a lower support modality such as low-flow nasal cannula, simple face mask, or room air). The tool then tracked the duration of this “floor-appropriate” level of respiratory support and alerted the managing PICU providers (resident, critical care fellow, and/or APP) to perform an assessment for transfer readiness after 6 hours. The primary method for notifying providers of potential transfer readiness was an alert pushed to the phones of the providers assigned to that patient as listed in the EHR ([Fig. 1]).

Fig. 1 Sample PICU clinician phone displays after CDS tool activation. (A) An alert recommending transfer assessment for a patient with viral bronchiolitis displays at the top of the clinician's phone screen. (B) Alert information and review of initiative inclusion/exclusion criteria is displayed after the clinician navigates into the alert. (C) Appearance of the alert in the alert history tab after the alert has been reviewed by the clinician.

In addition to a phone-based alert, we designed multiple visual alert tools to assist PICU clinicians with tracking the clinical progress of eligible bronchiolitis patients. In our institution's PICU, a central display monitor lists patients by room number and provides the name and contact information for the managing provider, nurse, and RT. These central monitors also display when a transfer order has been placed and when handoff to the receiving team has been completed. We added to this display an additional column to reflect the status of eligible bronchiolitis patients ([Fig. 2A]). Specifically, if a patient meets criteria for the QI initiative but is not yet on floor-appropriate respiratory support settings, a gray dot is displayed. Once the patient has been weaned to floor-appropriate respiratory support (8 LPM of HFNC or less), a yellow dot is displayed to indicate improving respiratory status. After 6 hours on this level of support without escalation above 8 LPM, the dot turns green indicating the patent's appropriateness for transfer assessment.

Fig. 2 Screenshots displaying addition of the new “Bronchiolitis Transfer Status” column added to: (A) the central PICU display monitor and (B) clinician desktop EHR patient lists. Presence of a gray dot indicates the patient meets inclusion and exclusion criteria for the QI initiative but has not yet weaned to floor-allowed respiratory support settings. The yellow dot indicates the patient has weaned to floor-allowed settings but has been on such settings for less than 6 hours. The green dot indicates the patient has been on floor-allowed respiratory settings for at least 6 hours and is ready for transfer assessment. Fictional patient data are used to protect patient privacy.

A similar visualization with identical color-coding scheme was added to the patient list display within the desktop Epic environment for use by PICU clinicians ([Fig. 2B]). The combination of these two visualizations was intended to optimize dissemination of information on patient clinical trajectory and readiness for transfer assessment to managing providers and nurses, as well as to charge nurses and PICU physicians tasked with optimizing patient flow and bed management.

Study Design and Statistical Analysis

We included patients admitted to our institution's PICU between December 1, 2018, and November 11, 2022. As our initial QI initiative began on October 20, 2019, patients admitted between December 1, 2018, and October 19, 2019, were included and analyzed to determine baseline characteristics (hereafter referred to as “pre-QI-phase” patients). Patients admitted from October 20, 2019, through February 7, 2022, were managed during our initial QI initiative (and prior to deployment of the EHR-based CDS tool) and are hereafter referred to as “QI-phase” patients. The CDS tool was deployed in the EHR on February 8, 2022, and patients admitted from February 8 to November 11, 2022, are hereafter referred to as “CDS-phase” patients.

As part of our prior QI initiative, our team built a PICU bronchiolitis dashboard to help track important metrics and clinical outcomes for children admitted to our PICU meeting initiative criteria. From this dashboard, we extracted relevant EHR patient data for analysis. This included information on patient demographics, PICU LOS, hospital LOS, and readmission to the PICU within 24 hours of transfer to the floor. We compared medians of continuous variables using Mann–Whitney U tests and proportions of categorical variables using Fisher's exact tests. Since the original QI initiative utilized statistical process control (SPC) rules to determine when a change had occurred, these rules[11] were also applied to this project's data for completeness (see [Supplementary Methods], available in the online version only). All analyses were performed using the R programming language version 4.2.1. Approval for the study was obtained through our institution's Organizational Research Risk and Quality Improvement Review Panel.

Clinical Decision Support Tool Silent Validation

To evaluate the performance of the bronchiolitis transfer CDS tool prior to active deployment in the EHR, we first performed a silent validation in which the alerts that would have been sent out by the tool were analyzed without transmission to clinicians. We performed manual chart review of all PICU admissions during the month of December 2021 to determine for which patients alerts would have fired or not fired correctly and incorrectly. The sensitivity and positive predictive value of the CDS tool in identifying the target population were both 92%. Two of 24 (8%) patients identified by the tool as meeting initiative criteria were false-positives. The first child was admitted for diabetic ketoacidosis but was incorrectly included as a result of a diagnosis of acute viral bronchiolitis listed in the patient problem list despite being categorized as a “nonhospital problem.” The CDS software was subsequently edited to only consider problems categorized by the provider as “hospital problems.” The second patient was a child admitted for acute viral bronchiolitis who arrived to the PICU intubated and receiving invasive mechanical ventilation. The CDS tool had identified this patient as meeting initiative criteria after 6 hours as no respiratory support or LDA information had been documented in the EHR flowsheets at the 6-hour mark. We addressed this problem by requiring documentation of the respiratory support level prior to CDS tool activation. Given the overall satisfactory performance of the CDS tool, it was actively deployed in our EHR on February 8, 2022.

Results

A total of 585 children met inclusion and exclusion criteria during the study period of which 150 (26%) were admitted during the pre-QI phase, 304 (52%) during the QI phase, and 131 (22%) during the CDS phase. Patient demographic variables were similar between the three phases ([Table 1]; [Supplementary Table S2], available in the online version).

Table 1 Demographic characteristics of patients managed during the three phases of the project

Demographic category	All patients	Pre-QI phase	QI phase	CDS phase
Total patients, n (%)	585 (100)	150 (26)	304 (52)	131 (22)
Age in months, median (IQR)	13.4 (7.88–17.9)	13.1 (7.56–17.40)	13.7 (7.83–17.7)	13.1 (8.87–19.0)
Race, n (%)
White/Caucasian	366 (63)	105 (70)	180 (59)	81 (62)
Black/African American	43 (7)	8 (5)	27 (9)	8 (6)
Other	78 (13)	20 (13)	38 (12)	20 (15)
Unknown	50 (9)	7 (5)	31 (10)	12 (9)
More than one race	38 (6)	10 (7)	22 (7)	6 (5)
Asian	10 (2)	0 (0)	6 (2)	4 (3)
Ethnicity, n (%)
Not Hispanic or Latino	357 (61)	90 (60)	182 (60)	85 (65)
Hispanic or Latino	188 (32)	53 (35)	96 (32)	39 (30)
Unknown	40 (7)	7 (5)	26 (9)	7 (5)

Abbreviations: CDS, clinical decision support; EHR, electronic medical record; IQR, interquartile range; QI, quality improvement.

Note: Pre-QI-phase patients are those children with viral bronchiolitis admitted to the PICU who would have met study criteria but were managed prior to introduction of the QI initiative and subsequent CDS tool. QI-phase patients are those managed after introduction of the QI initiative but before CDS tool deployment. CDS-phase patients are those managed after deployment of the EHR-based CDS tool. Patient age, race, and ethnicity are shown for completeness and were not variables used in the QI project or the CDS tool.

Children managed following deployment of our CDS tool (CDS-phase patients) demonstrated a shorter median time-to-transfer of 5.23 (interquartile range [IQR], 3.38–10.0) hours compared with 11.1 (IQR, 5.70–21.8) hours for patients in the pre-QI phase (p < 0.001) and 5.93 (4.24–12.2) hours in the QI phase (p = 0.04; [Table 2]). Though PICU and hospital LOS decreased from the pre-QI phase to the QI phase, further reductions seen during the CDS phase were not statistically significant. Changes in these outcomes over time by study phase are illustrated in modified I-charts ([Fig. 3]). Application of SPC rules did not identify special cause variation in time-to-transfer, PICU LOS, or total LOS during the CDS phase when compared with the QI phase. Rates of unplanned readmission to the PICU within 24 hours of transfer were not significantly different between CDS-phase and QI-phase patients ([Table 2]).

Fig. 3 Modified I-charts illustrating the change over time in median values for (A) time-to-transfer, (B) PICU LOS, and (C) total hospital LOS, by phase of study period. Shaded regions represent the interquartile range for each month's values. Months with <5 patients were omitted. x-bar represents the phase median. Note that figure x-bar (median) phase values differ from those of [Table 2] as a result of excluding months with <5 patients. HFNC, high-flow nasal cannula; LCL, lower control limit; LOS, length of stay; UCL, upper control limit.

Table 2 Clinical outcomes of children with viral bronchiolitis in the PICU managed during the three project phases

Clinical outcome	Pre-QI phase (N = 150)	QI phase (N = 304)	CDS phase (N = 131)	p-Value (QI vs pre-QI)	p-Value (CDS vs QI)
Time-to-transfer[a] in hours, median (IQR)	11.1 (5.70–21.8)	5.93 (4.23–12.2)	5.23 (3.38–10.0)	<0.001	0.04
PICU LOS in hours, median (IQR)	42.2 (28.0–68.0)	34.2 (22.5–47.9)	27.8 (19.8–48.1)	<0.001	0.14
Total hospital LOS in hours, median (IQR)	86.4 (61.9–132)	69.1 (53.5–95.4)	68.2 (49.2–111)	0.001	0.81
Unplanned PICU readmissions, n (%)	1 (1)	1 (0.3)	3 (2)	0.55	0.07

Abbreviations: CDS, clinical decision support; EHR, electronic medical record; IQR, interquartile range; LOS, length of stay; PICU, pediatric intensive care unit; QI, quality improvement.

^a Time-to-transfer indicates the amount of time that transpires between the patient reaching floor-allowed respiratory support and the transfer order being placed in the EHR.

Discussion

This report details the development and implementation of a CDS tool that automates identification of non–medically complex children with viral bronchiolitis in the PICU ready for transfer evaluation. Children managed after deployment of this CDS tool demonstrated a lower median time-to-transfer relative to patients managed during the initial QI project. However, special cause variation in time-to-transfer was not identified during application of SPC rules. Improvements in PICU and hospital LOS seen during the QI phase relative to the pre-QI phase were sustained following CDS deployment. Though LOS metrics were lower during the CDS phase than during the QI phase, these reductions were not statistically significant. While improvements in LOS outcomes following CDS tool deployment beyond those observed during the QI phase did not rise to the level of statistical significance, it is notable that this novel CDS tool was able to maintain such robust improvements in PICU and total hospital LOS without the additional cognitive load imposed on clinicians by the original QI initiative. That is, PICU clinicians no longer need spend additional time determining patient inclusion/exclusion criteria nor additional mental energy tracking floor-appropriate respiratory support settings. Instead, identification of the target population is now automated, allowing clinicians to leverage alerts to identify a patient's transfer-readiness in a timely manner.

Prior work has demonstrated that targeted QI interventions can accelerate the transfer of children from the PICU to the floor,[12] including among those admitted for common respiratory conditions such as status asthmaticus and viral bronchiolitis.[8] [13] Karube et al[13] and Alali et al[12] reported that the time interval between a patient becoming clinically ready for transfer and provider awareness of transfer readiness contributed to significant delays in transfer out of the PICU. By providing automated alerts to signal the need for an assessment for transfer readiness, such already successful QI initiatives might maintain improved outcomes without the additional cognitive load that QI projects sometimes add, allowing them to become more sustainable in the long term.

In the current state, specific criteria for PICU admission and discharge vary significantly from one institution to another,[14] as do levels of respiratory support approved for use outside of a PICU.[15] Despite these challenges, CDS systems could be adapted for local use with the same unified goal of identifying patients clinically appropriate for transfer out of the PICU. Broad implementation of locally refined CDS tools may have the potential to safely reduce PICU and hospital LOS for critically ill children across a multitude of institutions.

Limitations

There are several additional factors that should be considered when interpreting the results of this work. First, identification of eligible patients for the CDS tool requires that viral bronchiolitis be listed as the primary admission diagnosis within the EHR “Problem List.” Children with bronchiolitis who otherwise met inclusion criteria could have been missed as a result of clinicians not documenting viral bronchiolitis as the primary problem. Prior studies have demonstrated the utility of other EHR elements such as provider orders in forecasting transfer out of the PICU.[16] Future work should investigate development of viral bronchiolitis computable phenotypes able to leverage flowsheet and provider order information to identify additional bronchiolitis patients. Natural language processing, which has been used to assist in predicting future timing of NICU discharge readiness, could also facilitate improved patient identification.[17] [18] Furthermore, our institution's existing method for identifying bronchiolitis patients, which we leveraged in this project, utilizes bronchiolitis ICD codes attached to the admission diagnosis. Other ontologies such as Systematized Nomenclature of Medicine (SNOMED) could offer a more reliable method for bronchiolitis identification and should be pursued in future work. Second, exclusion of medically complex children relies on the presence of specific ICD codes from prior encounters within the patient's EHR at our institution. Patients with medical complexity who are new to our health care system could have been incorrectly included upon admission for viral bronchiolitis. Additional systems for identifying medical complexity should be explored, evaluated, and compared with the PCCC approach used in this study.

Third, establishing causation between introduction of our CDS tool and continued improvement in goal outcomes is challenging given numerous other factors at play. Changes in the pathophysiology of admitted patients over time, evolution of coronavirus disease 2019 (COVID-19) pandemic effects (both on local viral epidemiology and on PICU census), and staffing shortages all likely affect the clinical outcomes of interest within this population. It should be noted that the QI phase began as the COVID-19 pandemic started, and the CDS phase has mostly included patients during emergence from the pandemic. As such, ongoing evaluation will be crucial to ensure the tool has the desired effects as hospital and patient factors continue to evolve. Additionally, it is certainly possible that clinician behaviors developed during the initial QI project (e.g., the nurse/RT contacting the provider to consider transfer at the 6-hour mark) continued during the CDS phase out of habit and were ongoing drivers of improvement. That the time-to-transfer for some patients was less than 6 hours (the point at which the push notification is delivered) could support this possibility. However, given the sustained improvement observed 9 months after introduction of the CDS tool and cessation of QI activities, it is likely the CDS tool is a major driver of ongoing improvement. Formal evaluation of current clinician workflows in future work will be needed to confirm this. Though a randomized trial could have helped minimize such confounders, this was not feasible given that the project began as a QI initiative.

Additionally, the modest sample size of CDS-phase patients reported thus far may limit our ability to detect significant changes in key outcomes. As time progresses and CDS-phase patient numbers increase, we are hopeful that time-to-transfer and LOS metrics will continue to improve. Finally, a key objective of the CDS tool was to improve the cognitive load that had been placed on PICU clinicians as a result of the original QI initiative. Though we have received general positive feedback from PICU staff on the CDS tool, further work is needed to better understand and quantify this benefit.

Conclusion

Introduction of an EHR-based CDS tool was associated with improved transfer readiness identification among PICU patients with viral bronchiolitis when compared with patients managed during the original QI initiative on which the CDS tool was based. Sustained (but not additional) improvements in PICU and total hospital LOS were also observed. Future work should investigate the use of CDS tools to identify critically ill patients ready for transfer assessment who are admitted for other common conditions, as such tools may assist in safely reducing ICU and hospital LOS for appropriate patients.

Clinical Relevance Statement

Development and deployment of a CDS tool based on a prior QI initiative helped accelerate identification of non–medically complex children with viral bronchiolitis in the PICU with improving respiratory support needs who were candidates for transfer to the floor. Introduction of the CDS tool automated the process of identifying children who met the prior QI initiative criteria and alerted the provider team when a patient had clinical improvement and was likely ready for floor transfer. After CDS tool deployment, we observed continued reductions in the time-to-transfer relative to patients managed prior to the initial QI initiative and also observed sustained (but not further improved) reductions in PICU and hospital LOS.

Multiple-Choice Questions

1. After introduction of the bronchiolitis CDS tool, there was a statistically significant improvement (decrease) in which of the following clinical outcomes as compared with the patients managed during the initial QI initiative?

   • PICU length of stay.

   • Hospital length of stay.

   • Time from reaching floor-allowed respiratory support settings to placement of the transfer order (“time-to-transfer”).

   • All of the above.

   Correct Answer: The correct answer is option c. Compared with patients managed during the QI initiative (“QI-phase” patients), children managed during the CDS phase demonstrated shorter median values for time-to-transfer. Though median PICU and total hospital LOS values were lower during the CDS phase compared with the QI phase, these reductions were not statistically significant.

2. Which child would be included in the QI initiative and use of the CDS tool?

   • A previously healthy child admitted with viral bronchiolitis who was intubated for the first 24 hours of their PICU admission.

   • A previously healthy child with viral bronchiolitis who required noninvasive positive pressure ventilation for 2 days and then high-flow nasal cannula for 2 days prior to being transferred to the floor.

   • A child with a history of trisomy 21, congenital heart disease, chronic lung disease, and gastrostomy-tube dependence admitted with viral bronchiolitis who required 10 LPM of high-flow nasal cannula for 12 hours before being weaned to 8 LPM on the second day of PICU admission, at which point the child was transferred to the floor.

   • A previously healthy child admitted with viral bronchiolitis who required high-flow nasal cannula and continuous albuterol for status asthmaticus during his/her PICU encounter.

   Correct Answer: The correct answer is option b. Option a is incorrect as children requiring intubation and mechanical ventilation are excluded from the QI initiative and from CDS tool application. c is incorrect because children with chronic medical conditions (including chronic lung disease) are also excluded. Finally, d is incorrect as patients managed with continuous albuterol during their PICU encounter are excluded as well.

Conflict of Interest

M. B. receives research funding from Merck and Co. on a study that is unrelated to this paper. He has also served on an advisory board for Sanofi Pasteur but has not received any research funding from the company.

The remaining authors have no conflicts of interest to disclose.

Protection of Human and Animal Subjects

The study was performed in compliance with the World Medical Association Declaration of Helsinki on Ethical Principles for Medical Research Involving Human Subjects and was approved by the Children's Hospital Colorado Organizational Research Risk and Quality Improvement Review Panel (ORRQIRP) Review Panel.

• References

• 1 Seiger N, Maconochie I, Oostenbrink R, Moll HA. Validity of different pediatric early warning scores in the emergency department. Pediatrics 2013; 132 (04) e841-e850
• 2 Duncan H, Hutchison J, Parshuram CS. The Pediatric Early Warning System score: a severity of illness score to predict urgent medical need in hospitalized children. J Crit Care 2006; 21 (03) 271-278
• 3 Jensen CS, Olesen HV, Aagaard H, Svendsen MLO, Kirkegaard H. Comparison of two pediatric early warning systems: a randomized trial. J Pediatr Nurs 2019; 44: e58-e65
• 4 Soeteman M, Kappen T, van Engelen M. et al. Validation of a modified bedside pediatric early warning system score for detection of clinical deterioration in hospitalized pediatric oncology patients: a prospective cohort study. Pediatr Blood Cancer 2023; 70 (01) e30036
• 5 Mahant S, Parkin PC, Thavam T. et al; Canadian Paediatric Inpatient Research Network (PIRN). Rates in bronchiolitis hospitalization, intensive care unit use, mortality, and costs from 2004 to 2018. JAMA Pediatr 2022; 176 (03) 270-279
• 6 Farias JA, Fernández A, Monteverde E. et al; Latin-American Group for Mechanical Ventilation in Children. Mechanical ventilation in pediatric intensive care units during the season for acute lower respiratory infection: a multicenter study. Pediatr Crit Care Med 2012; 13 (02) 158-164
• 7 Linssen RS, Teirlinck AC, van Boven M. et al. Increasing burden of viral bronchiolitis in the pediatric intensive care unit; an observational study. J Crit Care 2022; 68: 165-168
• 8 Fritz CQ, Martin B, Riccolo M. et al. Reducing PICU-to-floor time-to-transfer decision in critically ill bronchiolitis patients using quality improvement methodology. Pediatr Qual Saf 2022; 7 (01) e506
• 9 Feudtner C, Feinstein JA, Zhong W, Hall M, Dai D. Pediatric complex chronic conditions classification system version 2: updated for ICD-10 and complex medical technology dependence and transplantation. BMC Pediatr 2014; 14 (01) 199
• 10 Feinstein JA, Russell S, DeWitt PE, Feudtner C, Dai D, Bennett TD. R Package for pediatric complex chronic condition classification. JAMA Pediatr 2018; 172 (06) 596-598
• 11 Carey RG, Stake LV. Improving Healthcare with Control Charts: Basic and Advanced SPC Methods and Case Studies. Quality Press; 2003
• 12 Alali H, Kazzaz Y, Alshehri A. et al. Reducing unnecessary delays during the transfer of patients from the paediatric intensive care unit to the general ward: a quality improvement project. BMJ Open Qual 2019; 8 (03) e000695
• 13 Karube T, Goins T, Karsies TJ, Gee SW. Reducing avoidable transfer delays in the pediatric intensive care unit for status asthmaticus patients. Pediatr Qual Saf 2022; 7 (01) e527
• 14 Frankel LR, Hsu BS, Yeh TS. et al; Voting Panel. Criteria for critical care infants and children: PICU admission, discharge, and triage practice statement and levels of care guidance. Pediatr Crit Care Med 2019; 20 (09) 847-887
• 15 Panciatici M, Fabre C, Tardieu S. et al. Use of high-flow nasal cannula in infants with viral bronchiolitis outside pediatric intensive care units. Eur J Pediatr 2019; 178 (10) 1479-1484
• 16 Levin SR, Harley ET, Fackler JC. et al. Real-time forecasting of pediatric intensive care unit length of stay using computerized provider orders. Crit Care Med 2012; 40 (11) 3058-3064
• 17 Temple MW, Lehmann CU, Fabbri D. Natural language processing for cohort discovery in a discharge prediction model for the neonatal ICU. Appl Clin Inform 2016; 7 (01) 101-115
• 18 Temple MW, Lehmann CU, Fabbri D. Predicting discharge dates from the NICU using progress note data. Pediatrics 2015; 136 (02) e395-e405

Publikationsverlauf

Eingereicht: 09. September 2022

Angenommen: 13. Februar 2023

Accepted Manuscript online:
15. Februar 2023

Artikel online veröffentlicht:
17. Mai 2023