a b s t r a c t

Background: Bronchiolitis is the most common cause for hospitalization in infants. While the use of high flow nasal cannula has increased, it has not uniformly reduced Intubation rates.

Objective: We identified factors associated with respiratory failure in children with bronchiolitis on HFNC. Methods: We conducted a retrospective study of previously healthy children b24 months of age with bronchiol- itis, who were treated with HFNC in two pediatric emergency departments from 1/2014-1/2018. The primary outcome was the identification of Demographic and clinical factors that are associated with intubation after an antecedent trial of HFNC. A multivariable logistic regression model was constructed to identify predictors of re- spiratory failure.

Results: Of 2657 children on HFNC, the median age was 7 months, while the median age of the intubated cohort was 3 months. Ten percent (271) progressed to mechanical ventilation within 48 h of PED presentation. Of the 301 patients that needed escalation to CPAP and/or BiPAP, 91 required intubation. Factors associated with intu- bation were young age and a high respiratory tool score; factors associated with no progression to intubation were a reduction in tachycardia after initiation of HFNC and presentation after day 5 of illness. A secondary anal- ysis also revealed decreased rate of intubation with the use of bronchodilators. We identified demographic, clin- ical, and therapeutic factors that are associated with requiring intubation.

Conclusion: Given the high burden of bronchiolitis in pediatric emergency departments, these factors can be con- sidered upon presentation of children with bronchiolitis to selectively identify children at higher risk for respira- tory failure.

1. Introduction

Bronchiolitis is the most common lower airway tract infection causing respiratory distress in infants and hospitalization of children younger than 2 years of age [1]. N10% of all infants are affected by bron- chiolitis in the first year of life [2], with up to 3% requiring admission to hospitals for respiratory distress, hypoxia, or dehydration [3], and 2-10% requiring intubation for respiratory failure [1].

* Corresponding author.

E-mail addresses: [email protected] (A. Suessman), [email protected] (L.L. Gray), [email protected] (S. Cavenaugh), [email protected] (E.A. Camp), [email protected] (Y. Shi), [email protected] (S.D. Meskill).

Over the last decade, high-flow nasal cannula therapy has emerged as a new method to provide less invasive respiratory sup- port [4]. The theory is that HFNC provides optimal warmth and hu- midity, enhances the liquefaction of secretions while preventing bronchoconstriction, and improves mucociliary clearance to de- crease respiratory distress [5]. Most importantly, this oxygen sup- port modality is well tolerated and can reduce hypoxemia and tachypnea [3,6]. HFNC therapy has been widely adapted and com- monly used in many pediatric emergency departments (PEDs) in re- cent years.

Some retrospective studies have shown HFNC’s possible benefit in reducing the need for intubation [7-11]. However, a randomized clinical trial evaluating HFNC usage did not see a difference in intubation rates [12]. Given the relative paucity of data regarding escalation of care in

https://doi.org/10.1016/j.ajem.2019.12.017 0735-6757/

children with bronchiolitis and the widespread use of HFNC, there is a need to determine factors associated with intubation. The purpose of this study is to determine which factors are associated with the need for intubation among patients presenting with bronchiolitis in the era of HFNC.

1. Methods

This was a retrospective cross-sectional study to determine if patient characteristics and vital signs impact the need for intubation to alleviate respiratory failure in children with bronchiolitis after HFNC use.

We included children b24 months old admitted for bronchiolitis from two PEDs in the greater Houston area between January 1, 2014 and January 1, 2018. One PED is a quaternary care center located in a large medical center with a total patient volume of 80,000 cases per year and an approximate admission rate of 29%, while the other is lo- cated in the community and treats about 40,000 patients annually with an approximate 12% admission rate. Approximately 30% of this study’s cohort represents patients presenting to the community site. Be- tween the two PEDs, there are 120,000 combined annual visits. The PEDs are in urban and suburban environments and are staffed by pedi- atric emergency medicine faculty, as well as trainee physicians from one pediatric residency, three emergency medicine residencies, and pediat- ric emergency medicine fellows. The authors of this paper began a ret- rospective chart review in 2018 and began data collection from 2014, when the institution’s utilization of HFNC for patients with bronchiolitis reached over 60%. Children were included if they received HFNC while in the PED and were therefore admitted.

Patients with bronchiolitis were identified by ICD-9 and ICD-10 codes (Appendix A) in the first three encounter diagnoses. The elec- tronic health record was used to abstract demographic characteristics (age, sex, ethnicity, race and insurance status). Clinical information in- cluded gestational age in weeks, weight, Acuity level, day of illness, PED length of stay , vital signs, time on respiratory support modal- ity (HFNC, continuous positive airway pressure [CPAP], bilevel positive airway pressure [BiPAP], mechanical ventilator), medications adminis- tered (including beta agonists, racemic epinephrine, systemic cortico- steroids), Clinical Respiratory Score (CRS [13], Appendix B), vital signs, viral studies, and co-diagnoses. Data that could not be extracted were obtained through manual chart review and entered into REDCap(R) (Vanderbilt University; Nashville, TN). For clinical and demographic variables, we included the percent positive of the number of children in whom that variable was documented. Ten percent of all manually ex- tracted records were examined by a second reviewer to ensure accuracy with Cohen’s kappa for inter-rater reliability on each element demon- strating excellent agreement (k = 0.79-0.88).

Children with diagnoses of congenital heart disease, neuromuscular disease, cystic fibrosis, do-not-intubate (DNI) status, immunodeficiency status, and airway anomaly or tracheostomy history as identified by In- ternational Classification of Disease (ICD)-9 or ICD-10 codes were ex- cluded. Children who were intubated by EMS or another hospital prior to arrival were also excluded. However, neither pneumonia nor prema- turity constituted exclusion criteria.

The primary outcome, escalation of support, was defined as requir- ing mechanical ventilation within 48 h of presentation. Intubation, rather than ICU admission, was chosen as a surrogate for respiratory failure to increase generalizability as ICU admission criteria vary among all institutions. The 48-h timeframe for intubation from arrival was chosen to decrease inclusion of mechanical ventilation required due to hospital-acquired complications.

Our primary objective was to identify demographic and clinical fac- tors associated with intubation of children with bronchiolitis admitted on HFNC. Our secondary objective was to identify therapeutic interven- tions that differed among intubated versus non-intubated children with bronchiolitis. This study was approved by the local institutional review board.

Descriptive comparisons between intubation statuses were ana- lyzed using the Pearson Chi-Square test or Fisher’s Exact Test for cate- gorical variables and the Mann-Whitney test for skewed, continuous data. Clinically relevant and statistically significant factors (P- value <= 0.20) were included in the initial logistic regression model. A priori, we evaluated risk of intubation in neonates (less than two months of age) compared to older children. To determine indicators for intubation, we used a backward step approach to create the final model. In the backward step approach, the factors with the highest P- value are removed one-by-one from the model until only factors with a P-value b 0.05 remain. For secondary predictors (medications and in- terventions), associations with intubation were calculated using unad- justed binary Regression modeling. All analyses used the Statistical Package for the Social Sciences (SPSS), version 25 (IBM Corp.; Armonk, NY).

1. Results

Over the 48-month study period, 3329 patients were admitted with a primary diagnosis of bronchiolitis. Of these, 2657 infants met inclusion criteria (Fig. 1); of these, 10.2% required mechanical ventilation within 48 h. More than half of all intubations occurred within 6 h of presenta- tion to the hospital. Based on significance and a priori selections, the following factors were selected for regression modeling: insurance, gen- der, race, ethnicity, gestational age, age at arrival in pre-defined groups, CRS in PED, vital signs at presentation (systolic blood pressure [SBP], temperature, heart rate [HR], respiratory rate [RR], oxygen saturation [SpO2]), vital signs before and after placement on HFNC (HR, RR, SpO2), RSV positive testing result, pneumonia diagnosis, and day of ill- ness (Table 1). Of the six factors found to be significant in the model, four were found to be clinically significant as predictors for intubation within 48 h of PED arrival in HFNC patients: younger age at presenta- tion, highest CRS, higher heart rate (HR) and presentation on day 4 of ill- ness (Table 2).

In secondary unadjusted analysis (Table 3), positive pressure venti- lation was associated with increasing rates of intubation. Positive pres- sure ventilation was needed for 301 (11%) children, of whom 30% required intubation. While systemic steroids were given to a minority of children, receipt was associated with reduced frequency of intuba- tion. Bronchodilator use was also associated with an unadjusted risk re- duction in intubation. To evaluate for confounding from age, we did a logistic regression model with bronchodilators as the exposure, intuba- tion as the outcome, and the association adjusted by age (Table 4). After adjusting for age, the association of bronchodilator use reduced the odds for intubation. Our secondary analysis also found an increased risk of intubation on days 3 and 4 of bronchiolitis illness caused by RSV (Appendix C). 72% of the cohort had viral testing of which 56% tested positive and of those 80% had RSV. This analysis on RSV cohort also showed increased risk of intubation after 1 week of symptoms.

1. Discussion

We evaluated factors associated with the need for mechanical venti- lation in infants receiving HFNC for bronchiolitis. We demonstrated that younger age at presentation and higher respiratory severity scores are associated with respiratory failure progressing to intubation while a de- creasing heart rate after application of HFNC was not associated with progression to mechanical ventilation. Many studies have evaluated clinical factors associated with intubation for children with bronchiolitis [14-17], but we feel that the number of factors we were able to analyze, combined with our large sample size recruited from a quaternary care and a community hospital, add substantially to the existing literature on the role of HFNC for bronchiolitis as well as the potential early iden- tification of infants requiring intubation. To our knowledge, this study represents the largest cohort of children admitted on HFNC with respi- ratory distress due to bronchiolitis.

Fig. 1. CONSORT diagram.

We found that patients presenting at a younger age were at a greater risk of intubation. These results are in agreement with prior studies that found age b2-3 months to be an independent predictor of respiratory decompensation requiring respiratory support beyond HFNC [14-21]. Specifically, our study showed that children 2 months of age and youn- ger had an increased risk of intubation when compared to older chil- dren. This may reflect the diminished respiratory reserves of young infants. Alternatively, a young infant’s increased risk for apnea may prompt intubation [9,15]. Kneyber et al. performed a chart review of 185 patients b12 months of age hospitalized with RSV infection and found that recurrent apnea significantly increased the risk of mechani- cal ventilation, and that age b2 months was the strongest predictor of apnea [18-21]. Since our denominator was children in whom HFNC was initiated, it is possible that the threshold for starting HFNC may have been lower in younger infants, in an attempt to preclude apneic episodes.

We found that a higher clinical respiratory score was associated with

an increased rate of intubation after HFNC. CRS is the respiratory sever- ity score used by our institution for multiple illnesses including bronchi- olitis. This scoring system considers RR, SpO2, wheezing and Accessory muscle use, along with perfusion and mental status to determine degree of respiratory distress. Similarly, the more commonly referenced Modi- fied Tal scoring (MTS) system scores distress based on RR, SpO2, wheez- ing and use of accessory muscles. This scoring system has been shown to be a reliable and reproducible assessment tool for bronchiolitis [22]. A recent study found strong correlation between higher MTS scores at ad- mission and O2 requirements at 48 h, determining MTS a reliable pre- dictor of respiratory failure [23]. Comparably, a recent study looking at CRS demonstrated that the score had excellent predictive value for dis- ease severity with an OR of 5.7 for patients requiring admission for

mild-moderate CRS scores (3-7). While this study looked at ages

1 month to 16 years, the median age was 12 months [IQR 2 to

34.5 months] which would be similar to our cohort [13]. Of note, our CRS is independent of the etiology of respiratory distress, and is used for asthma, bronchiolitis, and other causes of respiratory distress. This is important, as undifferentiated respiratory distress is common in the prehospital and PED settings, and having tools that identify children at risk for escalating respiratory support, irrespective of the underlying eti- ology, is important.

Recent publications have shown some clinical scoring tools and their components to be good predictors of respiratory failure on HFNC [10,16]. Both studies showed institutional-based respiratory assessment models to be accurate predictors of respiratory decompensation. It is worth noting that the CRS system used for this study is also an institu- tional scoring system. While CRS has not yet been validated for bronchi- olitis within our institution, it is commonly extrapolated to assess respiratory illness severity in other patients, especially those with bron- chiolitis and asthma. Nonetheless, further studies should be done to evaluate the precision of our own internal scoring tool at predicting need for intubation.

Our study also demonstrates a statistically significant decrease in the odds for intubation in patients receiving bronchodilators even after adjusting for age. While the American Academy of Pediatrics’ (AAP) practice guideline recommends against the use of bronchodilators and corticosteroids, trials of both drug classes remain common, particularly in critically ill children [24-29]. Some studies have found some benefit to specific symptoms, such as oxygenation and wheezing, especially when used on a trial basis or in the ICU setting, but do not otherwise in- fluence course of illness [26,30,31]. It is believed that wheezing in pa- tients with bronchiolitis is due to edema and debris in the airways

Table 1

Demographic and clinical descriptive comparison of HFNC & intubation patients (N = 2657).

Table 3 Bivariate (unadjusted) logistic regression models for breathing interventions and medica- tions associated with intubations in HFNC patients.

	No intubation within 48 h N = 2386 (89.8%) N (%) or median (IQR)	Intubation within 48 h of PED presentation N = 271 (10.2%) N (%) or median (IQR)	P-value				No intubation within 48 h N = 2386 (89.8%) N (%)	Intubation within 48 h of PED arrival N = 271 (10.2%) N (%)	OR	95% CI
Insurance						BiPAP	63 (2.6)	20 (7.4)	2.94	1.75-4.94
Private	859 (36.0)	80 (29.5)	0.004			CPAP	152 (6.4)	72 (26.6)	5.32	3.88-7.29
State Funded	1465 (61.4)	190 (70.1)				BiPAP/CPAP	210 (8.8)	91 (33.6)	5.24	3.92-7.00
Self-Pay/Other	62 (2.6)	1 (0.4)				Steroidsa	149 (6.2)	4 (1.5)	0.26	0.08-0.61
Race						Bronchodilatorsb	490 (20.5)	27 (10.0)	0.43	0.28-0.65
Caucasian	1698 (74.5)	196 (74.8)	0.24			Pneumonia	88 (3.7)	11 (4.1)	1.11	0.58-2.10

BiPAP: bilevel positive airway pressure; CPAP: continuous positive airway pressure; ED: emergency department; LOS: length of stay; CRS: clinical respiratory score; HFNC: high- flow nasal cannula; IQR: interquartile range RSV: respiratory syncytial virus infection.

African American	481 (21.1)	60 (22.9)
Othera	100 (4.4)	6 (2.3)
Ethnicity
Hispanic	1219 (52.7)	148 (56.1)	0.31
Non-Hispanic	1092 (47.3)	116 (43.9)
Gender
Male	1513 (63.4)	175 (64.6)	0.71
Female	873 (36.6)	96 (35.4)
Gestational age (weeks)	39.0 (37.0, 40.0)	38.0 (36.0, 40.0)	0.001
Age at presentation	6.95 (2.94, 12.7)	3.15 (1.35, 8.47)	b0.001
(months)
Age groups
b2 month	371 (15.5)	98 (36.2)	b0.001
2-b6 months	693 (29.0)	82 (30.3)
6-b12 months	659 (27.6)	50 (18.5)
12-b24 months	663 (27.8)	41 (15.1)
ED LOS	5.70 (4.20, 8.26)	3.91 (2.34, 6.54)	b0.001
Highest CRS in PED	5.0 (4.0, 6.0)	6.0 (4.0, 7.0)	b0.001
ED arrival to HFNC Start Time	137.50 (67.00, 348.75)	108.0 (46.0, 935.75)	0.81
ED Acuity Level	3.0 (2.0, 3.0)	2.0 (2.0, 3.0)	b0.001
Positive for RSVb	1053 (84.4)	193 (91.0)	0.01
BiPAP and/or CPAP	210 (8.8)	91 (33.6)	b0.001
Bronchodilatorsc	490 (20.5)	27 (10.0)	b0.001
Steroidsd	149 (6.2)	4 (1.5)	b0.001e

^a The “Other” category consist of Asian, Pacific Islander, and Native American.

^b Subgroup analysis among patients who had a positive viral test (n = 1459).

^c Bronchodilators: albuterol and/or levalbuterol.

^d Steroids includes: methylprednisolone, prednisolone, prednisone, dexamethasone.

^e P-value was calculated using the Fisher’s Exact Test.

from the Viral illness causing plugging of the small airways and not be- cause of bronchospasm or bronchial hypersensitivity as is the case with asthma [28]. However, there may be a subgroup of children that benefit from bronchodilation while experiencing the edema and debris associ- ated with bronchiolitis. For instance, children with atopy or a family

Table 2

Predictors for intubation 48 hours within PED arrival in HFNC patients.

BiPAP: bilevel positive airway pressure; CI: confidence interval; CPAP: continuous positive airway pressure; PED: pediatric emergency department; OR: odds ratio.

^a Steroids: methylprednisolone, prednisolone, prednisone and dexamethasone.

^b Bronchodilators: albuterol and levalbuterol.

history predominant of asthma, may benefit from bronchodilators [27,32]. It is possible that the burden of atopy is higher in our cohort, making children with bronchiolitis at our institution more responsive to those treatments. This is supported by the Centers for Disease Control and Prevention’s (CDC) evidence that asthma rates in children in Texas are higher than in the rest of the United States [33]. Further study should be dedicated to identifying patients more likely to be bronchodilator re- sponsive as this could change practice for these patients.

Our study did not find an increased risk of intubation with the sub- group of patients with RSV. Tristram and Welliver reported that 5% of patients with RSV are intubated [34]. Another study from Australia, which looked at 604 children with RSV bronchiolitis, reported that 25% required intubation [35]. Carroll et al. concluded that there is at least a 3.5-fold increased risk of intubation for bronchiolitis caused by RSV [27]. However, our study showed that while positive RSV testing re- sult was significant in the unadjusted analysis, it fell out from the re- gression model deeming that having this viral diagnosis does not influence intubation need. This conclusion is limited because not every patient was tested. In our sample, 1925 (72%) children had viral testing done. Overall, it is believed that bronchiolitis disease severity, in- cluding need for intubation, is worse in children with RSV [6] however our study does not support this belief.

Our analysis did however find an increased risk of intubation on ill- ness days 3 and 4 of RSV bronchiolitis. This result agrees with the per- ception of PED physicians who believe that children with bronchiolitis caused by RSV are at the peak for worse symptomatology around day 4 of illness. Fridman’s retrospective review of children admitted for RSV bronchiolitis also found that peak of illness, which was based on tachypnea, was on day 4 [37]. Other literature describes that the ex- pected course of RSV bronchiolitis presents to PEDs at its worst with the observed peak of symptoms on day 4-5 followed by improvement [29,33,38,39]. The model also showed that on day 6 of illness, one was statistically less likely to need mechanical ventilation. However, since we grouped children with symptoms of 1 week or more into one sub- group, it is difficult to determine its significance and furthermore

Table 4

Use of bronchodilators and intubation.

Age (months)

Factors	aORs	95% CI	P-value
Gestational age (weeks)	0.92	0.88-0.96	b0.001
Ages b2 months	Ref	-	-
Ages 2-5 months	0.42	0.27-0.66	b0.001
Ages 6-11 months	0.15	0.09-0.26	b0.001
Ages 12-b24 months	0.12	0.07-0.21	b0.001
Highest CRS in ED	1.52	1.36-1.71	b0.001
Post HF Pre ETT HR	1.03	1.02-1.04	b0.001
Post HF Pre ETT RR Days of illness	0.96	0.95-0.97	b0.001

Bronchodilators used (N = 517)

OR 95% CI P-value

<=2	Ref	-	-	Unadjusted			0.43	0.28-0.65	b0.001
3	1.33	0.82-2.16	0.24
4	1.14	0.69-1.89	0.60	Adjusted			0.65	0.42-1.00	0.05
5	0.51	0.26-1.02	0.06		b2	10	Ref
6	0.32	0.11-0.94	0.04		2-b6	68	0.46	0.33-0.63	b0.0001
>=7	1.27	0.61-2.64	0.52		6-b12	162	0.31	0.21-0.45	b0.0001

aOR: adjusted odds ratio; CI: confidence interval; CRS: clinical respiratory score; ETT: en- dotracheal intubation; HFNC: high-flow nasal cannula; HR: heart rate; Ref: reference

OR: odds ratio.

12-b24 277 0.27 0.18-0.41 b0.0001

group; PED: pediatric emergency department; RR: respiratory rate. Bronchodilators: albuterol and levalbuterol.

acute bronchiolitis and resp”>validate if the intubation risk is in fact from the viral status versus other complications of having a prolonged illness.

1. Limitations

As this was a retrospective chart review, classification bias could have occurred in the diagnosis of bronchiolitis at the time of the initial visit. Determining the clinical reason for intubation (respiratory distress versus apnea) can be difficult to ascertain retrospectively as there can be physician variability. However, we opted to use intubation as a surro- gate for disease severity rather than ICU admission, given that reasons for ICU admission may vary across institutions. We included patients with a diagnosis of pneumonia in the study because it is very difficult to discern the difference between viral and Bacterial pneumonia on chest x-ray. Our institution recommends against chest X-rays in all pa- tients with bronchiolitis however those who are sicker are more likely to get a chest x-ray (whether at intubation or not). These patients might have findings of pneumonia on chest x-ray that, otherwise, would be diagnosed as viral bronchiolitis if one was never obtained. While the physiology of bacterial pneumonia is different than viral pneumonia, it is difficult to ascertain with accuracy if the pneumonia was indeed viral or bacterial. We could not evaluate the prognostic ca- pacity of venous or other blood gases, as they were obtained in b4% of our cohort. In our study, there was a significant impact from the use of steroids and bronchodilators, which is not concordant with current AAP guidelines. As stated above, there is a high burden of asthma in Texas, which may contribute to our study population’s response to bronchodilators. While there is potential for confounding due to the ret- rospective design, we attempted to control for these with carefully de- fined inclusion criteria, large sample size, standardized data collection, and regression analysis. A prospective observational study should be the next step now that factors of interest have been identified.

1. Conclusions

In a large children’s hospital system where HFNC is frequently used, 10% of patients with bronchiolitis required intubation after

important intellectual content and statistical expertise, approval of revised manuscript

1. (SC): acquisition of the data, drafting of the manuscript and critical revision of the manuscript for important intellectual content, ap- proval of revised manuscript
2. (EAC): analysis and interpretation of the data, drafting of the manu- script and statistical expertise, approval of revised manuscript
3. (YS): acquisition of the data and drafting of the manuscript, approval of revised manuscript
4. (SDM): study concept and design and critical revision of the manu- script for important intellectual content, approval of revised manuscript

Declaration of competing interest

The authors report no conflicts of interest.

Appendix A. Acute bronchiolitis and respiratory syncytial virus ICD codes

ICD-9

466.1 - Acute Bronchiolitis

466.11 - Acute Bronchiolitis due to respiratory syncytial virus

466.19 - Acute Bronchiolitis due to other specified organisms

079.6 - Respiratory syncytial virus ICD-10

J21.0 - Acute Bronchiolitis due to respiratory syncytial virus J21.1 - Acute Bronchiolitis due to Human metapneumovirus J21.8 - Acute Bronchiolitis due to other specified organisms J21.9 - Acute Bronchiolitis, unspecified

B97.4 - Respiratory syncytial virus as the cause of disease classified elsewhere

Appendix B. Clinical Respiratory Score (CRS)

Clinical Respiratory Score (CRS)

Assess Score 0 Score 1 Score 2

HFNC use, most within 6 h of PED presentation. Younger age at pre- sentation and higher maximum respiratory severity scores were sig- nificantly associated with respiratory failure. Patients presenting to

Respiratory rate

b2 mo: b50 2-12mo: b40 1-5 yrs: b30 N5 yrs: b20

b2 mo: 50-60 2-12mo: 40-50 N1-5 yrs: 30-40

N5 yrs: 20-30

b2 mo: N 60 2-12mo: N50 N1-5 yrs: N40

N5 yrs: N30

the PED on day 5 of illness had statistically significant lower rates

of intubation. In our study, the use of steroids and bronchodilators reduced the odds for intubation with this holding significant for bronchodilators after adjusting for age. While prospective investiga- tive research is needed, we believe that these factors can be used to identify children at high risk of respiratory failure requiring intuba-

Auscultation Good air movement,

scattered expiratory wheezing, loose rales/crackles

Depressed air movement, inspiratory and expiratory wheezes or rales/crackles

Diminished or absent breath sounds, severe wheezing, or rales/crackles, or marked prolonged expiration

tion at an earlier stage in management and appropriate disposition from pediatric emergency departments.

Prior presentations

Use of

accessory muscles

Mild to no use of accessory muscles, mild to no retractions, no nasal flaring on inspiration

Moderate intercostal retractions, mild to moderate use of accessory muscles, nasal flaring

Severe intercostal and substernal retractions, nasal flaring

Pediatric Academic Society Annual Meeting, 4/29/19, Baltimore, MD

Mental status Normal to mildly

irritable

Irritable, agitated, restless

Lethargic

Financial disclosures

No financial support was received for the investigation or manu- script development

Author contributions

Room air SPO₂ N95% 90-95% b90%

Color Normal Pale to normal Cyanotic, dusky SpO2: oxygen saturation.

Appendix C. Intubation status and days of illness in patients who tested positive for RSV using chi-square analysis (N = 1246)

1. (AS): study concept and design, acquisition of the data, analysis and interpretation of the data, drafting of the manuscript, critical revision of the manuscript for important intellectual content and statistical expertise, manuscript revisions
2. (LLG): acquisition of the data, analysis and interpretation of the data, drafting of the manuscript, critical revision of the manuscript for

No intubation within 48 h

N = 1053 (84.5%) N (%)

Days of illness^a

<=2	152 (14.5)	36 (18.8)	Ref	-
3	242 (23.1)	49 (25.7)	0.86	0.53-1.38

Intubation within 48 h of PED presentation

N = 193 (15.5%) N (%)

OR 95% CI

(continued)

No intubation within 48 h

N = 1053 (84.5%) N (%)

Intubation within 48 h of PED presentation

N = 193 (15.5%) N (%)

OR 95% CI

1. Franklin D, Babl FE, Schlapbach LJ, et al. A randomized trial of High-flow oxygen ther- apy in infants with bronchiolitis. N Engl J Med 2018;378:1121-31. https://doi.org/ 10.1056/NEJMoa1714855.
2. Nayani K, Naeem R, Munir O, et al. The clinical respiratory score predicts paediatric critical care disposition in children with respiratory distress presenting to the emer- gency department. BMC Pediatr 2018;18:339. https://doi.org/10.1186/s12887-018- 1317-2.

4 260 (24.8) 56 (29.3) 0.91 0.57-1.45

5 213 (20.3) 23 (12.0) 0.46 0.26-0.80

6 104 (9.9) 11 (5.8) 0.45 0.22-0.92

>=7 78 (7.4) 16 (8.4) 0.87 0.45-1.66

CI: confidence interval; hrs: hours; PED: pediatric emergency department; OR: odds ratio; Ref: reference group.

^a Data on day of illness was missing for 20 children who were not intubated & 3 who were intubated

Appendix D. Vital signs comparison between HFNC and intubation patients (N = 2657)

1. Lebel MH, Gauthier M, Lacroix J, Rousseau E, Buithieu M. Respiratory failure and me- chanical ventilation in severe bronchiolitis. Arch Dis Child 1989;64:1431-7. https:// doi.org/10.1136/adc.64.10.1431.
2. Damore D, Mansbach JM, Clark S, Ramundo M, Camargo CA. Prospective multi- center bronchiolitis study: predicting intensive care unit admissions. Acad Emerg Med 2008;15:887-94. https://doi.org/10.1111/j.1553-2712.2008.00245. x.
3. Dadlez NM, Esteban-Cruciani N, Khan A, Douglas LC, Shi Y, Southern WN. Risk factors for respiratory decompensation among healthy infants with bronchiolitis. Hosp Pediatr 2017;7:530-5. https://doi.org/10.1542/hpeds.

2017-0034.

1. Mansbach JM, Piedra PA, Stevenson MD, et al. Prospective multicenter study of chil- dren with bronchiolitis requiring mechanical ventilation. Pediatrics 2012;130: e492-500. https://doi.org/10.1542/peds.2012-0444.
2. Kneyber MCJ, Brandenburg AH, De Groot R, et al. Risk factors for respiratory syncy-

No intubation within 48 h N = 2386 (89.8%)

N (%) or median (IQR)

Intubation within 48 h of PED presentation

N = 271 (10.2%)

N (%) or median (IQR)

P-value

tial virus associated apnoea. Eur J Pediatr 1998;157:331-5. https://doi.org/10.1007/ s004310050822.

1. Wang EEL, Law BJ, Stephens D. Pediatric Investigators Collaborative Network on Infections in Canada (PICNIC) prospective study of risk factors and outcomes in patients hospitalized with respiratory syncytial viral lower respiratory tract in- fection. J Pediatr 1995;126:212-9. https://doi.org/10.1016/s0022-3476(95) 70547-3.

Triage SBP	114.0 (103.0, 125.0)	112.0 (103.0, 126.0)	0.53
Triage DBP	66.0 (57.0, 75.0)	65.0 (55.0, 74.0)	0.09
Triage Temp	99.50 (98.60, 100.80)	99.30 (98.40, 100.60)	0.05
Triage HR	166.0 (152.0, 180.0)	167.5 (156.0, 182.0)	0.03
Triage RR	56.0 (48.0, 64.0)	60.0 (48.0, 67.0)	0.02
Triage SpO2	95.0 (93.0, 98.0)	95.0 (91.0, 98.0)	0.39
Post HF/Pre ETT HR	160.0 (145.0, 176.0)	174.0 (160.0, 186.0)	b0.001
Post HF/Pre ETT RR	52.0 (44.0, 60.0)	48.0 (28.0, 63.25)	b0.001
Post HF/Pre ETT SpO2	98.0 (96.0, 100.0)	100.0 (97.0, 100.0)	b0.001
Change in HRa	-5.0 (-23.0, 11.0)	4.0 (-15.0, 25.5)	b0.001
Change in RRb	-4.0 (-14.0, 6.0)	-9.5 (-26.25, 5.25)	b0.001

1. Green M, Brayer AF, Schenkman KA, Wald ER. Duration of hospitalization in previously well infants with respiratory syncytial virus infection. Pediatr In- fect Dis J 1989;8:601-4. https://doi.org/10.1097/00006454-198909000- 00007.
2. Opavsky MA, Stephens D, Wang EE. Testing models predicting severity of re- spiratory syncytial virus infection on the PICNIC RSV database. Pediatric In- vestigators Collaborative Network on Infections in Canada. Arch Pediatr Adolesc Med 1995;149:1217-20. https://doi.org/10.1001/archpedi.1995. 02170240035005.
3. McCallum GB, Morris PS, Wilson CC, et al. Severity scoring systems: are they inter-

DBP: diastolic blood pressure; PED: pediatric emergency department; ETT: endotracheal intubation; HFNC: high-flow nasal cannula; HR: heart rate; IQR: interquartile range; RR: respiratory rate; SBP: systolic blood pressure; SpO2: oxygen saturation.

PED: pediatric emergency department. SpO2: oxygen saturation.

^a Change in HR = “Post_HF_Pre_ETT_HR_Comb”- “HR_triage”.

^b Change in RR = “Post_HF_Pre_ETT_RR_Comb” - “respiratory_rate_triage”

References

Hasegawa K, Tsugawa Y, Brown DFM, Mansbach JM, Camargo Jr CA. Trends in bron- chiolitis hospitalizations in the United States, 2000-2009. Pediatrics 2013;132: 28-36. https://doi.org/10.1542/peds.2012-3877.

Kriett M, Jamieson SW. Diagnosis and management of bronchiolitis obliterans. Transplant Replace Thorac Organs 1996:547-55. https://doi.org/10.1007/978-0- 585-34287-0_59.

Wainwright C. Acute viral bronchiolitis in children- a very common condition with few therapeutic options. Paediatr Respir Rev 2010;11:39-45. https://doi.org/10. 1016/j.prrv.2009.10.001.

Lin J, Zhang Y, Xiong L, Liu S, Gong C, Dai J. High-flow nasal cannula therapy for chil- dren with bronchiolitis: a systematic review and meta-analysis. Arch Dis Child 2019; 104:564-76. https://doi.org/10.1136/archdischild-2018-315846.

Wing R, James C, Maranda LS, Armsby CC. Use of high-flow nasal cannula support in the emergency department reduces the need for intubation in pediatric acute respi- ratory insufficiency. Pediatr Emerg Care 2012;28:1117-23. https://doi.org/10.1097/ PEC.0b013e31827122a9.

Masclans JR, Roca O. high-flow oxygen therapy in acute respiratory failure. Clin Pulm Med 2012;19:127-30. https://doi.org/10.1097/CPM.0b013e3182514f29.

McKiernan C, Chua LC, Visintainer PF, Allen H. High flow nasal cannulae therapy in infants with bronchiolitis. J Pediatr 2010;156:634-8. https://doi.org/10.1016/j. jpeds.2009.10.039.

Schibler A, Pham TMT, Dunster KR, et al. Reduced intubation rates for infants after introduction of high-flow nasal prong oxygen delivery. Intensive Care Med 2011; 37:847-52. https://doi.org/10.1007/s00134-011-2177-5.

Kelly GS, Simon HK, Sturm JJ. High-flow nasal cannula use in children with respira- tory distress in the emergency department: predicting the need for subsequent in- tubation. Pediatr Emerg Care 2013;29:888-92. https://doi.org/10.1097/PEC. 0b013e31829e7f2f.

Abboud PA, Roth PJ, Skiles CL, Stolfi A, Rowin ME. predictors of failure in infants with viral bronchiolitis treated with high-flow, high-humidity nasal cannula therapy*. Pediatr Crit Care Med 2012;13:e343-9. https://doi.org/10.1097/PCC. 0b013e31825b546f.

Metge P, Grimaldi C, Hassid S, et al. Comparison of a high-flow humidified nasal can- nula to nasal continuous positive airway pressure in children with acute bronchiol- itis: experience in a pediatric intensive care unit. Eur J Pediatr 2014;173:953-8. https://doi.org/10.1007/s00431-014-2275-9.

nally valid, reliable and predictive of oxygen use in children with acute bronchioli- tis? Pediatr Pulmonol 2013;48:797-803. https://doi.org/10.1002/ppul.22627.

1. Golan-Tripto I, Goldbart A, Akel K, Dizitzer Y, Novack V, Tal A. Modified Tal Score: validated score for prediction of bronchiolitis severity. Pediatr Pulmonol 2018;53: 796-801. https://doi.org/10.1002/ppul.24007.
2. Oakley E, Brys T, Borland M, et al. Medication use in infants admitted with bronchiolitis. Emerg Med Australas 2018;30:389-97. https://doi.org/10.1111/1742-6723.12968.
3. Florin TA, Byczkowski T, Ruddy RM, Zorc JJ, Test M, Shah SS. Variation in the man- agement of infants hospitalized for bronchiolitis persists after the 2006 American Academy of Pediatrics bronchiolitis guidelines. J Pediatr 2014;165:786-792.e1. https://doi.org/10.1016/j.jpeds.2014.05.057.
4. Condella A, Mansbach J, Hasegawa K, et al. Multicenter study of albuterol use among infants hospitalized with bronchiolitis. West J Emerg Med 2018;19:475-83. https:// doi.org/10.5811/westjem.2018.3.35837.
5. Carroll CL, Faustino EVS, Pinto MG, et al. A regional cohort study of the treatment of critically ill children with bronchiolitis. J Asthma 2016;53:1006-11. https://doi.org/ 10.1080/02770903.2016.1180697.
6. American Academy of Pediatrics Subcommittee on Diagnosis and Management of Bronchiolitis. Diagnosis and management of bronchiolitis. Pediatrics 2006;118: 1774-93. https://doi.org/10.1542/peds.2006-2223.
7. National Institute for Health and Care Excellence. Bronchiolitis in children: diagnosis and management. https://www.nice.org.uk/guidance/ng9; 2015, Accessed date: 18 October 2019.
8. Gadomski AM, Scribani MB. Bronchodilators for bronchiolitis. Cochrane Database Syst Rev 2014:CD001266. https://doi.org/10.1002/14651858.CD001266.pub4.
9. Sala KA, Moore A, Desai S, Welch K, Bhandari S, Carroll CL. Factors associated with disease severity in children with bronchiolitis. J Asthma 2015;52:268-72. https:// doi.org/10.3109/02770903.2014.956893.
10. Fernandes RM, Hartling L. Glucocorticoids for acute viral bronchiolitis in infants and young children. JAMA 2014;311:87-8. https://doi.org/10.1001/jama.2013.284921.
11. Lewis GT. Environmental defense fund. Asthma in Texas. http://blogs.edf.org/ texascleanairmatters/2016/08/01/asthma-in-texas/; 2016, Accessed date: 18 Octo-

ber 2019.

1. Tristram D, Welliver RC. Respiratory syncytial virus. In: Long SS, Pickering LK, Prober CG, editors. Principles and practice of pediatric infectious diseases. 2nd ed. New York: Churchill Livingstone; 2003. p. 213-8.
2. Pham H, Thompson J, Wurzel D, Duke T. Ten years of severe respiratory syncytial virus infections in a tertiary paediatric intensive care unit. J Paediatr Child Health 2019. https://doi.org/10.1111/jpc.14491.
3. Fridman M. Three days coming three days staying: the course of respiratory syncy- tial virus bronchiolitis. Open Forum Infect Dis 2016:1283. https://doi.org/10.1093/ ofid/ofw172.986.
4. Eiland LS. Respiratory syncytial virus: diagnosis, treatment and prevention. J Pediatr Pharmacol Ther 2009;14:75-85.
5. McGrath B. The Royal Children’s Hospital Melbourne. Bronchiolitis ward man- agement. https://www.rch.org.au/rchcpg/hospital_clinical_guideline_index/ Bronchiolitis_Ongoing_Management/, Accessed date: 11 October 2019.