Association of high-flow nasal cannula (HFNC) therapy with intensive care unit admission and mechanical ventilation use in respiratory syncytial virus–positive bronchiolitis in pediatric patients: a retrospective single-center cohort study

Abstract

Background

High-flow nasal cannula (HFNC) therapy is increasingly used in pediatric patients with respiratory syncytial virus (RSV)–positive bronchiolitis to manage respiratory distress. Despite widespread adoption in both intensive care unit (ICU) and non-ICU settings, evidence regarding its impact on critical clinical outcomes remains inconsistent. This study aimed to evaluate the association between HFNC use and ICU admission and mechanical ventilation (MV) use in children with RSV-positive bronchiolitis.

Methods

We conducted a retrospective cohort study of pediatric patients under two years of age admitted with polymerase chain reaction–confirmed RSV bronchiolitis at a tertiary care hospital in Palestine between January 2021 and December 2023. Patients receiving HFNC were compared with those managed using standard low-flow oxygen therapy. Primary outcomes were ICU admission and MV use. Secondary outcomes included co-infection rates, antibiotic use, laboratory and radiological findings, and markers of disease severity. Multivariable logistic regression models were used to identify independent predictors of ICU admission and MV, adjusting for age, HFNC use, chronic illness, co-infection, and C-reactive protein (CRP).

Results

A total of 712 patients were included. HFNC recipients were younger (mean age 4.07 vs. 7.48 months) and had a higher prevalence of chronic medical conditions compared with non-HFNC patients. In unadjusted analyses, HFNC use was associated with higher rates of ICU admission (16.1% vs. 6.5%; p = 0.006) and MV (14.5% vs. 5.7%; p = 0.007). Co-infection was strongly associated with ICU admission and MV (p < 0.001). In multivariable analysis, HFNC use was independently associated with lower odds of ICU admission (OR 0.393, 95% CI 0.174–0.885; p = 0.024) but was not significantly associated with MV (OR 0.445, 95% CI 0.187–1.055; p = 0.066). Increasing age was associated with higher odds of MV (OR 1.161, 95% CI 1.041–1.294; p = 0.007). CRP levels were not associated with either outcome.

Conclusions

HFNC therapy was more frequently used in younger and higher-risk pediatric patients with RSV bronchiolitis. After adjustment for confounders, HFNC use was associated with lower odds of ICU admission but was not independently associated with MV.

Graphical Abstract

Introduction

Bronchiolitis is defined as the infection of the respiratory bronchioles with resultant airway obstruction, hyperinflation of the alveoli, increased airway resistance, atelectasis, and ventilation-perfusion mismatch [1]. Respiratory syncytial virus (RSV) is a leading cause of bronchiolitis and pneumonia in infants and young children. Globally, it accounts for a significant burden of lower respiratory tract infections, especially in children under 12 months of age [2].

Viral bronchiolitis is a predominant cause of hospital admissions and mortality among this age group, often leading to acute respiratory failure. Patients present with a wide spectrum of symptoms ranging from mild to severe respiratory distress, necessitating variable levels of medical intervention [3]. RSV infections in children under five years of age have a considerable global impact, with an estimated 33 million cases annually, including 3.2 million hospitalizations and 120,000 deaths [4]. Of these, approximately 60,000 deaths occur in hospital settings. RSV shows a distinct seasonality, with peak incidence in the fall and winter months, and a slightly higher prevalence in male infants [4].

Several risk factors contribute to the progression of RSV bronchiolitis to severe disease. These include prematurity, low birth weight, maternal smoking, passive smoke exposure, male gender, absence of breastfeeding, overcrowding, daycare attendance, indoor pollution, and family history of atopy [5]. The disease commonly affects infants between three and six months of age and presents initially with symptoms of an upper respiratory tract infection. This progresses to lower respiratory tract involvement, manifesting as cough, tachypnea, respiratory distress, and auscultatory findings such as wheezes and crackles. Fever, though less common, occurs in about one-third of cases [6]. The symptoms typically resolve within two weeks, but a subset of patients continues to experience respiratory symptoms for up to three weeks [3]. Management of RSV bronchiolitis primarily involves supportive care, with oxygen therapy and nutritional support being the cornerstone interventions [7].

High-flow nasal cannula (HFNC) therapy delivers heated and humidified oxygen at high flow rates. Despite the increasing adoption of HFNC therapy for managing respiratory distress in pediatric patients with RSV-positive bronchiolitis, both inside and outside ICU, evidence regarding its impact on critical outcomes such as intensive care unit (ICU) admission and mechanical ventilation (MV) remains inconclusive [8]. Physiologically, HFNC may reduce work of breathing and improve respiratory parameters, such as respiratory and heart rates; however, evidence showing superiority over standard low-flow oxygen remains limited. Studies have showed that there’s no clear advantages of HFNC over low-flow oxygen or noninvasive ventilation [9]. Additionally, several studies have failed to show reductions in invasive MV, ICU admission, length of hospital stay, or duration of oxygen therapy when HFNC is used as an initial treatment compared with conventional oxygen therapy [10]. Despite its widespread adoption, there are no universally accepted international guidelines regarding HFNC initiation, escalation, or weaning in bronchiolitis, leading to substantial variability in clinical practice across institutions and healthcare systems.

Given the increasing use of HFNC in pediatric patients with RSV bronchiolitis despite inconsistent evidence, further evaluation of its clinical impact is warranted. The primary objective of this study was to assess the association between HFNC therapy and key clinical outcomes, specifically ICU admission and the need for MV, in pediatric patients with RSV-positive bronchiolitis. Secondary objectives included evaluating the relationship between HFNC use and co-infection rates, antibiotic utilization, and laboratory findings, and other markers of disease severity. By examining real-world outcomes in a tertiary care setting, this study aims to clarify the clinical implications of HFNC use in RSV bronchiolitis. Despite increasing use of HFNC in bronchiolitis, evidence regarding its impact on ICU admission and MV remains inconsistent, and real-world patient selection for HFNC is not well characterized. This study adds clinically applicable evidence by analyzing HFNC use in a large cohort of polymerase chain reaction (PCR)-confirmed RSV bronchiolitis, examining how patient characteristics, co-infection status, and inflammatory markers relate to HFNC initiation and escalation of respiratory support.

Materials and methods

Study design

This retrospective cohort study was conducted to evaluate the impact of HFNC therapy on clinical outcomes in pediatric patients with RSV-positive bronchiolitis. The study focused on key outcomes, including ICU admission rates and MV use, comparing patients treated with HFNC to those receiving standard oxygen therapy. Secondary outcomes included co-infection rates, antibiotic use, laboratory findings, and radiological results. The study period spanned from January 2021 to December 2023. The design, data collection, and reporting of this study adhere to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [11]. A completed checklist is available upon request.

Setting

The study was performed at a tertiary care hospital in Palestine. This hospital provides specialized care, including advanced respiratory and critical care support. Data were extracted from electronic healthcare records (EMRs) to ensure a standardized and comprehensive collection. In this hospital, HFNC therapy is typically initiated shortly after presentation to the emergency department (ER) when infants show persistent hypoxemia, significant work of breathing, or apnea. Initial stabilization is performed in the ER, after which patients are transferred to the pediatric ward where HFNC is commenced and monitored. HFNC is therefore generally initiated early during hospitalization rather than as a rescue intervention in the ICU. Patients who subsequently deteriorate despite HFNC therapy, such as worsening respiratory distress, progressive hypoxemia, or signs of respiratory failure, are transferred to the ICU for escalation of care, including consideration of MV. Standard low-flow oxygen was used for milder disease.

Participants

The study population included pediatric patients under two years of age who were diagnosed with RSV-positive bronchiolitis, confirmed through PCR testing. Eligible participants were required to have a first episode of RSV-positive bronchiolitis and complete medical records. Patients were excluded if their medical records were incomplete, if bronchiolitis was caused by pathogens other than RSV, or if they had received MV before admission. Additional exclusion criteria included the presence of severe chronic conditions that independently necessitated ICU admission or MV, such as advanced congenital heart disease or severe neuromuscular disorders. Patients who were transferred into or out of the hospital during their admission or who received experimental respiratory therapies beyond HFNC or standard oxygen therapy were also excluded. Baseline demographic characteristics, including age, gender, and the presence of chronic medical conditions, were documented for all eligible participants.

Variables

The study examined several key variables. The primary dependent variables included ICU admission and the need for MV. The primary independent variables were age, type of respiratory support (HFNC versus standard oxygen therapy), and the presence of chronic medical conditions. Secondary variables included a range of laboratory findings such as white blood cell (WBC) count, C-reactive protein (CRP) levels, and arterial blood gas (ABG) results, as well as diagnostic findings like chest X-ray patterns, blood culture results, and lumbar puncture (LP) findings. Additionally, clinical outcomes such as co-infection status, antibiotic use, and the duration of supplemental oxygen therapy were analyzed.

Chest X-rays were interpreted by a consultant radiologist, and findings were confirmed by a second consultant radiologist. Co-infection was defined as the presence of an additional confirmed infection at the time of hospital presentation in patients with PCR-confirmed RSV bronchiolitis. Confirmation was based on gold-standard diagnostic testing appropriate to the suspected infection.

The ABG measurements were typically obtained early in the clinical course. ABGs results were categorized into clinically relevant groups based on standard reference ranges. Respiratory acidosis was defined as pH < 7.35 with PaCO₂ > 45 mmHg, metabolic acidosis as pH < 7.35 with normal or low PaCO₂, and respiratory alkalosis as pH > 7.45 with PaCO₂ < 35 mmHg. Hypoxemia was defined as PaO₂ < 60 mmHg or oxygen saturation < 92% at the time of sampling. Hypercapnia was defined as PaCO₂ > 45 mmHg in the absence of acidosis. Unremarkable ABG was defined as pH 7.35–7.45, PaCO₂ 35–45 mmHg, and absence of hypoxemia. In cases where multiple abnormalities were present, classification was based on the predominant physiological disturbance.

Data sources and measurement

Data were collected retrospectively from EMRs using a standardized protocol by trained physicians. Key data points included PCR-confirmed RSV infection, demographic and clinical characteristics, and laboratory parameters such as WBC counts and CRP levels measured on admission and during hospitalization. Radiological findings, including chest X-ray patterns such as infiltration, pneumonia, hyperinfiltration, and atelectasis, were recorded. Treatment interventions, including the initiation of HFNC therapy, antibiotic administration, and transitions to MV, were also documented.

Bias

Efforts were made to address potential sources of bias. Data collection protocols were standardized to minimize variability, and patients with chronic conditions that could independently necessitate ICU admission or MV were excluded to reduce confounding. The use of PCR testing ensured accurate classification of RSV infection, minimizing the risk of misclassification bias.

Study size

The study included a retrospective analysis of 712 pediatric patients, selected based on hospital admission records during the defined study period. Due to the retrospective nature of the study, no formal sample size calculation was performed, and all available eligible cases were included to maximize statistical power.

Statistical methods

Statistical analyses were conducted using IBM Statistical Package for the Social Sciences (version 29). Descriptive statistics were used to summarize patient demographics, clinical characteristics, and outcomes. Data were presented as counts and percentages (n [%]), except for variables indicated with an asterisk, which were expressed as mean and standard deviation (SD). Categorical variables were compared using chi-square tests, while continuous variables were analyzed using the Mann-Whitney test for non-normally distributed data. Normality of distribution was assessed using the Kolmogorov-Smirnov test. Logistic regression models were constructed to evaluate independent predictors of ICU admission and MV. Separate multivariable models were developed for each outcome. Covariates included in the regression models were selected a priori based on clinical relevance and prior literature and comprised age, HFNC use, presence of chronic medical conditions, CRP on admission, and co-infection status. Adjusted odds ratios (ORs) with 95% confidence intervals (CIs) were reported. Missing data were minimal and addressed through listwise deletion where applicable. A p-value of less than 0.05 was considered statistically significant for all analyses. Chronic medical conditions and co-infection were coded as binary variables (presence vs. absence) based on documentation in the medical record at the time of admission. The reference category in regression models was absence of the condition.

Ethics

The study protocol was approved by the Institutional Review Board (IRB) of the College of Medicine at Hebron University in Palestine (ID: ER.CM.March7/2024). As a retrospective study, the requirement for informed consent was waived in compliance with institutional and ethical guidelines. The data were extracted from the spreadsheet in an anonymous format and de-identified prior to analysis. All information was securely stored on a password-protected computer, ensuring complete confidentiality, adhering to the principles outlined in the Declaration of Helsinki.

Results

This study retrospectively analyzed 712 pediatric patients diagnosed with RSV-positive bronchiolitis to evaluate the impact of HFNC on clinical outcomes. Of the 750 patients initially screened, 38 were excluded due to incomplete medical records, leaving 712 for analysis. Reasons for exclusion included missing ABG data (20 patients) and unconfirmed RSV diagnosis (18 patients) (Fig. 1). Missing data were minimal (< 5%) and handled through listwise deletion. All included participants had complete data for key variables such as ICU admission, MV, and co-infection status. Notably, the HFNC group consisted of younger patients (mean age 4.07 months) compared to the non-HFNC group (mean age 7.48 months) (Fig. 2). Chronic conditions were more prevalent among those receiving HFNC (29.5%) versus those not receiving HFNC (17%) (Fig. 2), indicating that HFNC use is more common in patients with complex underlying medical issues (Table 1).

Fig. 1.

Study flowchart depicting patients’ selection. RSV, respiratory syncytial virus; PCR, polymerase chain reaction

Fig. 2.

Baseline differences between HFNC and non-HFNC groups. Panel A shows that patients receiving HFNC were significantly younger (mean 4.07 months vs. 7.48 months). Panel B shows that patients receiving HFNC had a higher prevalence of chronic conditions (29.5% vs. 17.0%). HFNC, high-flow nasal cannula

Table 1.

Baseline characteristics of pediatric patients with RSV-positive bronchiolitis, comparing HFNC and non-HFNC treatment groups

Characteristics		HFNC		Total
		Yes	No
Age*		4.07 (6.03)	7.48 (13.86)	7.23 (13.39)
CRP on Admission*		24.02 (36.75)	24.02 (39.44)	24.03 (39.07)
Mechanical Ventilation	Yes	9 (14.5%)	37 (5.7%)	46 (6.5%)
	No	53 (85.5%)	609 (94.3%)	662 (93.5%)
ABG	Respiratory acidosis	21 (33.9%)	121 (18.6%)	142 (19.9%)
	Metabolic acidosis	5 (8.1%)	23 (3.5%)	28 (3.9%)
	Respiratory alkalosis	1 (1.6%)	12 (1.8%)	13 (1.8%)
	Hypoxemia	15 (24.2%)	96 (14.8%)	111 (15.6%)
	Hypercapnia	3 (4.8%)	17 (2.6%)	20 (2.8%)
	Unremarkable	15 (24.2%)	271 (41.7%)	286 (40.2%)
	Not performed	2 (3.2%)	110 (16.9%)	112 (15.7%)
Lumbar Puncture	Positive	2 (3.2%)	8 (1.3%)	10 (1.4%)
	Negative	7 (11.3%)	61 (9.5%)	68 (9.7%)
	Not done	53 (85.5%)	570 (89.2%)	623 (88.9%)
Chronic Diseases	Yes	18 (29.5%)	109 (17%)	127 (18%)
	No	43 (70.5%)	534 (83%)	577 (82%)
Chest X-ray	Infiltration	15 (24.2%)	76 (12%)	91 (13.1%)
	Pneumonia	0 (0%)	12 (1.9%)	12 (1.7%)
	Hyperinfilation	0 (0%)	3 (0.5%)	3 (0.4%)
	Bronchiolitis changes	5 (8.1%)	6 (0.9%)	11 (1.6%)
	Atelectasis	3 (4.8%)	3 (0.5%)	6 (0.9%)
	Free	14 (22.6%)	85 (13.5%)	99 (14.1%)
	Not done	25 (40.3%)	450 (70.9%)	475 (68.1%)
Co-infection	Yes	15 (25%)	175 (27.6%)	190 (27.4%)
	No	45 (75%)	459 (72.4%)	504 (72.6%)
ICU Need	Yes	10 (16.1%)	42 (6.5%)	52 (7.4%)
	No	52 (83.9%)	602 (93.5%)	654 (92.6%)
Antibiotics	Single medication	17 (27.4%)	118 (18.3%)	135 (19.1%)
	Combination	9 (14.5%)	51 (7.9%)	60 (8.5%)
	No	36 (58.1%)	477 (73.8%)	513 (72.5%)

Data are presented as n (%), except * presented as mean (SD). ICU intensive care unit, HFNC high-flow nasal cannula, CRP C-reactive protein, ABG arterial blood gas

Critically, the use of HFNC was associated with a significantly increased need for ICU admission (16.1% vs. 6.5%; p = 0.006) and MV (14.5% vs. 5.7%; p = 0.007) (Fig. 3). This sharp rise in the need for MV and ICU support highlights the potential severity of respiratory distress in HFNC patients, particularly in the context of acute bronchiolitis exacerbated by RSV. Furthermore, these findings suggest that HFNC is often utilized in patients with more severe disease courses, potentially as a last step before MV, thus reflecting its critical role in managing advanced respiratory compromise (Tables 1 and 2).

Fig. 3.

Unadjusted clinical outcomes by respiratory support. In unadjusted analyses, patients treated with HFNC had significantly higher rates of ICU admission (16.1% vs. 6.5%; p = 0.006) and mechanical ventilation (14.5% vs. 5.7%; p = 0.007) compared to those on standard oxygen therapy. HFNC, high-flow nasal cannula; ICU, intensive care unit

Table 2.

Clinical outcomes by HFNC vs. non-HFNC groups stratified by ICU need and mechanical ventilation

		ICU Need		P-value	Mechanical Ventilation		P-value
		Yes (n)	No (n)		Yes (n)	No (n)
HFNC	Yes (n)	10	52	0.006	9	53	0.007
	No (n)	42	602		37	609
Co-infection	Yes (n)	29	162	<0.001	25	165	<0.001
	No (n)	22	482		20	484

The chi-square test is used for statistical analysis. HFNC high-flow nasal cannula, ICU intensive care unit

ABG analysis showed a spectrum of abnormalities across both groups (Table 1). In the HFNC group, respiratory acidosis was observed in 33.9% of patients, metabolic acidosis in 8.1%, respiratory alkalosis in 1.6%, hypoxemia in 24.2%, and hypercapnia in 4.8%. In contrast, the non-HFNC group showed lower proportions of respiratory acidosis (18.6%), metabolic acidosis (3.5%), hypoxemia (14.8%), and hypercapnia (2.6%), while respiratory alkalosis was comparable between groups (1.8%). Unremarkable ABG findings were more common in the non-HFNC group (41.7%) compared to the HFNC group (24.2%). ABG testing was not performed in a higher proportion of non-HFNC patients (16.9%) than HFNC patients (3.2%).

Co-infections were slightly more frequent in the non-HFNC group (27.6%) compared to the HFNC group (25%), but co-infections were a decisive factor for worse outcomes across both cohorts. ICU admission occurred in 16.1% of patients receiving HFNC compared to 6.5% in the non-HFNC group (absolute risk difference: −9.6%). MV was required in 14.5% of the HFNC group versus 5.7% of the non-HFNC group (absolute risk difference: −8.8%). Patients with co-infections showed a significantly heightened need for ICU admission (p < 0.001) and MV (p < 0.001), highlighting the critical need for vigilant monitoring and aggressive management in these cases (Tables 1 and 2).

Antibiotic administration, both as monotherapy and in combination, was more prevalent in the HFNC group. This further suggests that these patients are presented with more severe or complex clinical pictures, potentially due to secondary bacterial infections or other complicating factors. This finding may also indicate more aggressive treatment strategies in patients requiring HFNC, aligning with the higher rates of ICU admission and MV observed (Table 1).

CRP levels on admission were not predictive of disease severity, as there was no statistically significant difference in CRP levels between patients who required MV and those who did not (p = 0.895), nor between the HFNC and non-HFNC groups (p = 0.212). This highlights the limited utility of CRP as a biomarker for assessing the need for more advanced respiratory interventions in this patient population (Table 3).

Table 3.

This table evaluates CRP levels on admission across different patient groups, including those requiring mechanical ventilation and those who did not

	Mechanical Ventilation		P-value	HFNC		P-value
	Yes	No		Yes	No
CRP on Admission*	31 (56)	24 (38)	0.895	24 (37)	24 (39)	0.212

CRP levels are also compared between HFNC and non-HFNC groups. Mann-Whitney Test is used for statistical analysis, and Kolmogorov-Smirnova is employed for testing normal distribution testing. *: Data are presented as mean (SD). HFNC high-flow nasal cannula, CRP C-reactive protein

In multivariable logistic regression analyses (Table 4), HFNC use was independently associated with lower odds of ICU admission (OR 0.393, 95% CI 0.174–0.885; p = 0.024). Chronic medical conditions were statistically associated with reduced odds of ICU admission (OR 0.329, 95% CI 0.166–0.653; p = 0.001), as was co-infection (OR 0.283, 95% CI 0.146–0.549; p < 0.001). Age (p = 0.158) and CRP levels (p = 0.553) were not significantly associated with ICU admission (Fig. 4).

Table 4.

Multivariable logistic regression analysis identifying predictors of ICU admission and mechanical ventilation in pediatric patients with RSV-positive bronchiolitis

Characteristics	ICU Admission			Mechanical Ventilation
	Odds Ratio, (CI)		P-value	Odds Ratio, (CI)	P-value
Age	1.026, (0.99, 1.063)		0.158	1.161, (1.041, 1.294)	0.007
CRP on Admission	0.998, (0.991, 1.005)		0.553	0.996, (0.988, 1.003)	0.272
HFNC Need	0.393, (0.174, 0.885)		0.024	0.445, (0.187, 1.055)	0.066
Chronic Diseases	0.329, (0.166, 0.653)		0.001	0.279, (0.133, 0.584)	0.001
Co- infection	0.283, (0.146, 0.549)		< 0.001	0.28, (0.137, 0.571)	< 0.001

 Adjusted ORs with 95% CIs and corresponding p-values are presented for each variable included in the regression models. ICU intensive care unit, HFNC high-flow nasal cannula, CRP C-reactive protein, OR odds ratio, CI confidence interval

Fig. 4.

Adjusted predictors of ICU admission and mechanical ventilation in patients with RSV-positive bronchiolitis. Forest plots display adjusted ORs with 95% CIs derived from multivariable logistic regression models. Panel A shows predictors of ICU admission, and Panel B shows predictors of mechanical ventilation. Covariates included HFNC use, age, presence of chronic illness, and co-infection. The vertical dashed line represents an OR of 1.0 (no association). Estimates to the left of the line indicate lower odds of the outcome, while estimates to the right indicate higher odds. After adjustment, HFNC use was associated with lower odds of ICU admission but was not independently associated with mechanical ventilation. ICU, intensive care unit; HFNC, high-flow nasal cannula; OR, odds ratio; CI, confidence interval; RSV, respiratory syncytial virus

For MV, increasing age remained independently associated with higher odds of MV (OR 1.161, 95% CI 1.041–1.294; p = 0.007). Chronic medical conditions (OR 0.279, 95% CI 0.133–0.584; p = 0.001) and co-infection (OR 0.28, 95% CI 0.137–0.571; p < 0.001) were associated with lower odds of MV. HFNC use showed a trend toward reduced MV requirement but did not reach statistical significance (OR 0.445, 95% CI 0.187–1.055; p = 0.066). CRP levels were not significantly associated with MV (p = 0.272) (Table 4).

Discussion

In this retrospective single-center cohort of 712 pediatric patients with PCR-confirmed RSV bronchiolitis, the apparent reversal in the direction of association between unadjusted and adjusted analyses of the HFNC use is most likely explained by confounding by indication. In unadjusted analyses results, HFNC use was associated with higher rates of ICU admission (16.1% vs. 6.5%) and MV (14.5% vs. 5.7%); however, patients receiving HFNC were younger and had a higher burden of chronic medical conditions, both of which are established markers of increased disease severity and risk of deterioration. In clinical practice, HFNC is typically initiated in patients perceived to be at higher risk, meaning that the crude association reflects underlying patient severity rather than a harmful effect of HFNC itself. After adjusting for these baseline differences in multivariable analysis, HFNC use was associated with lower odds of ICU admission, suggesting that when comparing patients of similar risk profiles, HFNC may have a protective or stabilizing role. This phenomenon is well recognized in observational studies and highlights the importance of adjustment for confounding factors when evaluating treatment effects in non-randomized settings.

The multivariable results further suggest that the crude association reflects patient selection rather than a causal harmful effect of HFNC. After adjustment, HFNC was associated with lower odds of ICU admission, supporting the interpretation that HFNC was preferentially initiated in higher-risk patients but was not independently associated with worse outcomes. However, HFNC was not independently associated with reduced need for MV. This pattern aligns with the literature showing that HFNC frequently improves short-term physiologic comfort but does not consistently translate into lower ICU admission or intubation when used as initial therapy. A randomized trial in moderate-to-severe bronchiolitis found no clinically meaningful benefit of high-flow compared with low-flow oxygen in improvement scores, duration of oxygen, length of stay, or ICU admission for invasive ventilation [12]. Similarly, a systematic review and network meta-analysis concluded that experimental evidence does not show clear advantages of HFNC over conventional oxygen as initial therapy or over NIV as rescue therapy for major outcomes [9]. These findings support a cautious interpretation: HFNC may help reduce “treatment failure” defined as escalation from low-flow to high-flow, but may not reduce the need for invasive ventilation in settings where escalation pathways are already robust.

Importantly, evidence in non-bronchiolitis acute hypoxemic respiratory failure may also reflect system-level behaviors around HFNC. In the PARIS-2 randomized clinical trial (children 1–4 years without bronchiolitis), early high-flow oxygen did not reduce length of stay and was associated with longer hospitalization and higher ICU admission compared with standard oxygen [10]. The authors suggested potential explanations including slower weaning and clinician perception that high-flow implies higher illness severity, lowering thresholds for ICU escalation [10]. Those mechanisms could plausibly operate in bronchiolitis as well, especially in centers without standardized HFNC initiation and weaning protocols.

Real-world health system data also suggest HFNC practice variation can affect outcomes and resource use. A multicenter retrospective study examining hospital-level HFNC use outside the ICU reported that higher non-ICU HFNC usage was associated with longer length of stay and higher costs, even after adjustment [13]. This raises the possibility that greater HFNC use may be associated with longer hospitalization, potentially reflecting differences in weaning practices, monitoring requirements, or institutional escalation norms [13]. The substantial international variability in HFNC initiation and weaning is well described in surveys, with many centers relying on local protocols [14, 15].

In multivariable analysis, both chronic medical conditions and co-infection were associated with lower odds of ICU admission and MV. These findings are counterintuitive from a clinical perspective, as both factors are typically associated with increased disease severity [16]. Several explanations may account for this observation. First, this may reflect residual confounding and differences in clinical decision-making that were not fully captured in the available data. For example, patients with known chronic conditions may present earlier, receive closer monitoring, or be managed more proactively, potentially reducing the need for ICU escalation. Second, co-infection status may act as a proxy for diagnostic and management pathways rather than disease severity alone. Patients identified with co-infections may have received earlier antibiotic therapy or closer clinical observation, which could influence outcomes. Additionally, the definition of co-infection relied on confirmed diagnoses, which may preferentially capture patients who were more thoroughly investigated rather than those with more severe disease. Third, unmeasured markers of disease severity such as work of breathing, feeding intolerance, or clinician-assessed distress were not consistently available in the dataset and therefore not included in the model. These factors may have influenced both treatment decisions and outcomes, leading to residual confounding. Given the retrospective design and these potential sources of bias, these associations should be interpreted with caution and considered hypothesis-generating rather than indicative of a protective effect.

Antibiotic exposure often reflects clinician concern rather than proven bacterial infection. Current guideline documents emphasize that bronchiolitis management is mainly supportive and caution against routine antibiotics without clear bacterial evidence [16].

Admission CRP was not associated with MV use, ICU admission and did not differ meaningfully by HFNC exposure. The lack of predictive value of CRP in our analysis is also consistent with the broader view that single inflammatory markers poorly discriminate bronchiolitis severity and bacterial co-infection in many settings. This aligns with the guideline emphasis to avoid over-testing and focus on clinical assessment and oxygenation/feeding status [16, 17]. Together, these findings highlight that escalation outcomes are likely determined by a combination of clinical trajectory, support thresholds, and coexisting conditions.

In ABG analysis, patients receiving HFNC showed higher rates of respiratory acidosis, hypoxemia, and hypercapnia compared with those managed with standard oxygen therapy. These findings are consistent with the clinical observation that HFNC is preferentially initiated in patients with more severe respiratory compromise. Respiratory acidosis was notably more frequent in the HFNC group, reflecting alveolar hypoventilation and possible early respiratory muscle fatigue. The prevalence of hypoxemia aligns with the known pathophysiology of bronchiolitis, in which hypoxemia typically occurs early due to ventilation–perfusion mismatch, while hypercapnia and respiratory acidosis develop later as a result of fatigue and worsening airway obstruction [1].

This study has several limitations, most notably the potential for residual and unmeasured confounding inherent to its retrospective, non-randomized design and the single-center setting, which may limit generalizability. Confounding by indication is likely, as patients with greater illness severity were more frequently treated with HFNC, potentially inflating ICU admission and MV rates in this group. Unmeasured confounding may also be present because several clinical markers of disease severity were not consistently recorded in the medical records. Multiple subgroup and secondary outcome analyses were performed without formal correction which may have influenced the observed associations. We were unable to perform a sensitivity analysis restricting to patients who started HFNC within the first 6–12 h, which may have affected outcomes. Measurement and classification bias may also be present, particularly for co-infection, radiographic findings, and antibiotic use. Strengths include a large cohort, PCR-confirmed RSV infection, and clinically relevant outcomes.

External validity should be interpreted in light of local resource availability, ICU capacity, referral patterns, and monitoring infrastructure in our tertiary care setting. Nevertheless, the overall interpretation is consistent with international evidence: HFNC is commonly used in higher-risk bronchiolitis and does not consistently reduce major clinical outcomes compared with conventional oxygen, particularly in the absence of standardized escalation and weaning protocols [9, 14–16, 18]. These findings suggest the importance of a systems-aware approach in which HFNC is incorporated into clear institutional pathways defining initiation criteria, monitoring requirements, escalation triggers, and structured weaning. Future studies should focus on prospective assessment of disease severity and the implementation of standardized HFNC protocols to better delineate its role and optimize resource use in RSV bronchiolitis.

In summary, this single-center retrospective cohort suggests that HFNC use was associated with lower odds of ICU admission but was not independently associated with MV after adjustment for measured confounders. These findings contribute to the existing literature on HFNC use in RSV bronchiolitis and support further prospective studies to clarify its role and optimize institutional practice pathways.

Authors’ contributions

Majd Oweidat: Conceptualization, data curation, writing – original draft, writing – review & editing, formal analysis, investigation, methodology, project administration, resources, validation, visualization, and supervision. Mohammed Alra’e: Data curation, writing – original draft, investigation, and resources. Ra’fat Allawi: Data curation, writing – original draft, project administration, and supervision. Muneer Marwani: Conceptualization, data curation, project administration, and supervision. Wasef Alhroub: Data curation, writing – original draft, investigation, and resources. Alaa Ahmed Elshanbary: Data curation, and formal analysis. Qassam Rae, Kata’eb Talahmeh, Mohammad Arafeh, Alaa K. Najjar: Data curation, and writing – original draft. Mohammed Alsabri: Data curation, writing – original draft, and supervision. All Authors have approved the final copy of the manuscript.

Funding

No funding was received.

Data availability

The datasets generated and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

The study protocol was approved by the Institutional Review Board (IRB) of the College of Medicine at Hebron University in Palestine (ID: ER.CM.March7/2024). The IRB also formally waived the requirement for informed consent, as the study was retrospective and used anonymized, de-identified data. All information was securely stored on a password-protected computer, ensuring complete confidentiality and adherence to the principles outlined in the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.