Noninvasive Ventilation for Acute Respiratory Failure in Pediatric Patients: A Systematic Review and Meta-Analysis : Pediatric Critical Care Medicine

Journal Logo

Review Articles

Noninvasive Ventilation for Acute Respiratory Failure in Pediatric Patients: A Systematic Review and Meta-Analysis

Boghi, Daniele MD1; Kim, Kyung Woo MD2; Kim, Jun Hyun MD, PhD2; Lee, Sang-Il MD2; Kim, Ji Yeon MD, PhD2; Kim, Kyung-Tae MD, PhD2; Ambrosoli, Andrea MD1; Guarneri, Giovanni MD3; Landoni, Giovanni MD3,4; Cabrini, Luca MD5,6

Author Information
Pediatric Critical Care Medicine 24(2):p 123-132, February 2023. | DOI: 10.1097/PCC.0000000000003109


Noninvasive ventilation (NIV) is a well-established treatment for acute respiratory failure (ARF) in adult patients (1), being superior in terms of survival and other relevant outcomes in many conditions when compared with standard oxygen therapy (2–4). ARF is a common problem also in the pediatric age, accounting for a relevant percentage of hospital admission and deaths (5). Consequently, NIV has also been evaluated in children. Few narrative reviews have described its potential role in this population (5, 6).

NIV was considered a promising tool in both hypoxemic and hypercapnic ARFs due to airway or parenchymal lung disease, but the lack of strong evidence was underlined. The 2021 Spanish Guidelines on NIV, a consensus of several national scientific societies, included a section for pediatric NIV use: based on two single-center, small observational studies, NIV was recommended in any pediatric patient with ARF who presents without contraindications and in pediatric patients at risk of extubation failure (7).

Two systematic reviews with meta-analysis were published on NIV use in the pediatric age; both considered only papers from low- to middle-income countries and did not focus on randomized controlled trials (RCTs) only (8, 9). Mandelzweig et al (8) concluded that NIV use can be effective, but the most recent one (9) concluded that data are inconclusive and context-dependent; therefore, available systematic reviews have conflicting results. Furthermore, no meta-analysis including only RCTs has ever been published for NIV application in pediatric age.

To better evaluate the best evidence from all countries on this topic, we performed a systematic review of RCTs investigating NIV efficacy in preventing intubation in children affected by ARF.


We used the Cochrane methodology and Preferred Reporting Items for Systematic Reviews and Meta-Analyses to conduct a systematic review and meta-analysis of randomized controlled trials (RCTs). This protocol was preregistered in the International Prospective Register of Systematic Reviews (CRD42022300669). Two trained investigators independently searched Pubmed and EMBASE, and the Cochrane Central Register of Clinical Trials. was also searched for studies that had recently been completed but had not yet been published, up to July 31, 2022. We also scanned the references of the retrieved articles in order to find further RCTs. A third senior author resolved disagreements between the two investigators. In Supplementary Material 1 (, search techniques for PubMed and EMBASE are reported. The search strategies were created to retrieve any RCT related to NIV use in pediatric patients. We considered as NIV both Continuous Positive Airway Pressure (CPAP) and Pressure Support (PS) or Bilevel Positive Airway Pressure (BiPAP) if delivered noninvasively; high-flow nasal cannula (HFNC) was not considered as an NIV technique. There were no language restrictions. Articles were initially reviewed for title and abstract, then retrieved as a complete article if they fulfilled the inclusion criteria. The following Participants, Intervention, Comparison, Outcomes, and Study Design criteria had to be met by eligible studies: Population: pediatric (i.e., >28 d after birth but <18 yr of age) patients affected by ARF, also including the postoperative setting and the postextubation setting; Interventions: NIV; Comparison intervention: any comparator; Outcome: intubation rate; Study design: randomized controlled trials. Exclusion criteria are as follows: trials performed exclusively or mainly in the neonatal age (<28 d after birth); chronic respiratory conditions (i.e., a persistent—usually for more than 1 year—or recurring respiratory disease); RCTs not reporting the outcomes of interest; and RCTs comparing only two NIV techniques. RCTs in which cross-over between the two investigated techniques was applied as a rescue measure were included, whereas those in which the cross-over was adopted after a predefined treatment period (preplanned cross-over) were excluded.

Data Extraction

The baseline characteristics, outcome data, and other pertinent information from the selected studies were collected separately by two trained authors, with a third author resolving disagreements. The primary outcome was the intubation rate. Secondary outcomes were mortality at the ICU, in hospital, and at the longest follow-up available, hospital and ICU length of stay, and any failure other than intubation.

The Revised Cochrane risk-of-bias tool (RoB 2) was used to assess the methodological quality of included RCTs (10). Those were assessed by two reviewers, with a third reviewer resolving disagreements. We classified each study as low, high, or with some concerns of risk of bias using Cochrane’s five items.

Data Analysis

The RevMan 5.3 software (The Nordic Cochrane Centre, The Cochrane Collaboration Copenhagen, Odense, Denmark), the metafor package (Version 2.4-0) for R (R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria), and STATA (Version 16; Stata Corporation LLC, College Station, TX) were used to conduct our analyses. We used a random-effects model because there were various types of NIV, and the patient population is diverse, so the treatment effect would not be identical among included studies (11). If the number of studies included was less than five, the Hartung, Knapp, Sidik, and Jonkman technique was used to reduce type I errors. If standard deviations were not reported, we used an average of the standard deviations from the studies, which reported standard deviations (12). If median and interquartile ranges (IQRs) were available from the articles, we imputed standard deviations using the Median-IQR method (mean imputation as median, sd imputation as third quartile to first quartile) (13). When only the ranges were available, we calculated sd by dividing it by 4 (14). The publication bias was checked by drawing funnel plots and using Egger test.


Our search strategy identified 2,795 records (Fig. 1). Major exclusions with reasons for exclusion are presented in Supplementary Table 1 ( Fifteen studies (15–29) and 2,679 patients were included in the analyses: NIV group (n = 1,268) and control group (n = 1,411).The analyzed studies included mainly patients affected by bronchiolitis, pneumonia, asthma, and postextubation failure (Supplementary Table 2, Comparators were represented by HFNC and standard oxygen “low flow” treatment (e.g., low-flow nasal cannula, mask, and hood). Only one study compared NIV with both HFNC and low-flow oxygen therapy (18).

Figure 1.:
Flowchart of study selection.

Primary Outcome

NIV showed significantly lower intubation rate than the control group (14 RCTs; NIV group, n = 106 of 945 [11.5%]; control group, n = 156 of 1,086 [14.5%]; risk ratio, 0.791; 95% CI, 0.629–0.996; p = 0.046; I2 = 0%; number needed to treat = 31) (Fig. 2). Visual inspection of the funnel plots and Egger test did not reveal a small study publication bias (p = 0.70) (Supplementary Fig. 1,; legend,

Figure 2.:
Forest plots of intubation rate. The control group consists of high-flow nasal cannula and low-flow oxygen (e.g., nasal cannula, mask, and hood). The study by Ramnarayan et al in 2018 (24) compared noninvasive ventilation (NIV) for an acute illness or after extubation (step-up or step-down), so we analyzed the groups separately (Ramnarayan P1 and Ramnarayan P2). RR = relative risk.

In subgroup analyses, NIV confirmed a lower intubation rate when studies with short duration of intervention (less than one hour) were excluded (15, 17–21, 23–29) (13 RCTs; NIV group, n = 105 of 927 [11.3 %]; control group, n = 156 of 1,067 [14.6%]; risk ratio, 0.78; 95% CI, 0.62–0.98; p = 0.03; I2 = 0%) (Fig. 3) and when studies allowing cross-over as rescue treatment were excluded (15, 17–21, 27–29) (nine RCTs; NIV group, n = 27 of 310 [8.7%]; control group, n = 58 of 383 [15.1%]; risk ratio, 0.57; 95% CI, 0.37–0.85; p = 0.007; I2 = 0%) (Supplementary Fig. 2,; legend,

Figure 3.:
Forest plots of intubation rate without studies with short time intervention (<1 hr). NIV = noninvasive ventilation, RR = relative risk.

Other comparisons of subgroups, such as NIV versus HFNC and NIV versus standard oxygen treatment, confirmed the magnitude and direction of findings (Supplementary Fig. 3,; legend, as did the sensitivity analyses (Supplementary Fig. 4,; legend,

The results of the risk of bias assessment are summarized in Figure 4. The quality of the studies was overall high when considering the impossibility of blinding the intervention.

Figure 4.:
Assessment of methodological quality of included studies based on the revised Cochrane risk-of-bias tool.

There was no small study publication bias by the Egger tests and with a visual inspection of funnel plots (Supplementary Fig. 1,; legend,

Secondary Outcomes

Mortality did not differ between groups: ICU mortality (six RCTs [15, 16, 23–26]; NIV group, n = 10 of 642; control group, n = 13 of 711; risk ratio, 0.78; 95% CI, 0.28–2.14; p = 0.62; I2 = 12%) (Fig. 5), inhospital mortality (six RCTs [15, 18, 21–24]; NIV group, n = 58 of 575; control group, n = 60 of 652; risk ratio, 0.64; 95% CI, 0.21–1.99; p = 0.45; I2 = 67%) (Fig. 5), longest follow-up mortality (eight RCTs [15, 16, 18, 21–25]; NIV group, n = 73 of 840; control group, n = 96 of 938; risk ratio, 0.65; 95% CI, 0.23–1.79; p = 0.40; I2 = 81%) (Fig. 5).

Figure 5.:
Forest plots of mortality. A, Forrest plot of ICU mortality noninvasive ventilation (NIV) vs control group. B, Forrest plot of hospital mortality NIV vs control group. C, Forrest plot of longest follow-up mortality NIV vs control group. The control group consists of high-flow nasal cannula and low-flow oxygen (e.g., nasal cannula, mask, and hood). The study by Ramnarayan et al in 2018 (24) compared NIV for an acute illness or after extubation (step-up or step-down), so we analyzed the groups separately (Ramnarayan P1 and Ramnarayan P2). RR = relative risk.

There was only a trend toward a reduced ICU and hospital length of stay in the NIV group when compared with the control group (Fig. 6). There were also no differences in ICU and hospital length of stay when NIV was compared with HNFC or standard oxygen therapy, separately (Supplementary Fig. 5,; legend,

Figure 6.:
Forest plots of ICU and hospital stay. The study by Ramnarayan et al in 2018 (24) compared noninvasive ventilation (NIV) for an acute illness or after extubation (step-up or step-down), so we analyzed the groups separately (Ramnarayan P1 and Ramnarayan P2). A, Forrest plot of ICU stay NIV vs control group. B, Forrest plot of hospital stay NIV vs control group. SMD = standard mean differences.

ICU stay was reduced in the NIV group when studies in postextubation patients were excluded: seven RCTs (16,18,24,25,27,28,30); SMD, 0.15 days; 95% CI, 0.02–0.29; p = 0.027; I2 = 0% (Supplementary Fig. 6,; legend,


In this first meta-analysis of RCTs focusing on NIV efficacy in ARF in the pediatric age, we found that NIV reduces intubation rate compared with standard oxygen therapy or HFNC, with findings strengthened when excluding trials that allowed cross-over to the comparator as a rescue treatment in case of failure and when trials applying interventions for less than 1 hour were excluded. No difference was found in mortality rates and ICU and hospital length of stay.

Infants and children suffer from ARF more frequently than adults; furthermore, ARF is a common cause of cardiac arrest in children (30). Prompt and effective treatment is required. The studies analyzed in our systematic review included mainly patients affected by bronchiolitis, pneumonia, asthma, and postextubation failure; unfortunately, data are too limited to perform a reliable analysis on the efficacy of NIV in preventing intubation according to the cause of ARF. Two meta-analyses recently investigated NIV efficacy in asthma and bronchiolitis in children (31,32). Regarding NIV use for acute asthma in children, NIV compared with standard treatment had no benefit or harm in terms of mortality, serious adverse events, or improvement in asthma symptoms, but the analysis was based on only two RCTs (31). Tang et al (32) evaluated the effectiveness of only CPAP in acute bronchiolitis: CPAP reduced hospital length of stay compared with standard oxygen therapy and reduced the treatment failure compared with HFNC, but no other outcome was affected.

Even though NIV could be a promising treatment in the pediatric setting, it should be underlined that this technique requires a trained and committed multidisciplinary team: the success of NIV is strictly linked to the correct choice of the interface, setting of the ventilator parameters, prevention of skin lesions, monitoring of respiratory fatigue signs, and several other factors (33). Furthermore, light sedation is frequently required to treat children (34). Risk factors for NIV failure include age below 6 months, no improvement after 2 hours of treatment, apnea, and pneumonia (35). Acute respiratory distress syndrome (ARDS) is likely the most challenging condition for NIV: a high failure rate must be expected, up to 80% (5), in particular, if severity scores are elevated (2). More alarming, a recent large trial found that preintubation NIV in pediatric ARDS was associated with worse mortality rates (36). The finding was not confirmed in a more recent (but smaller) retrospective trial (37). Our findings cannot clarify the issue: in the wait for large RCTs on ARDS, NIV in this context should be applied very cautiously, and intubation should not be delayed. On the other hand, most studies did not observe any major complications (5).

Our systematic review and meta-analysis present some limitations. We found a relatively low number of RCTs, with different setting, population, and applied protocols. Hence, reliable analyses of relevant subgroups (i.e., analyses focused on different causes of ARF) were impossible. Therefore, our findings are to be considered preliminary, and large RCTs are required before firm conclusions can be drawn. In particular, CPAP and PS/BiPAP might have different efficacy in different conditions: our study focused on comparing NIV versus other forms of ARF treatments, excluding RCTs comparing two NIV techniques. Furthermore, data on the severity of ARF were often lacking, with inclusion criteria depending on clinician judgment; the definitions of standard oxygen therapy and NIV were also not always reported or different among trials. We excluded neonatal age, as this stage of development presents peculiar characteristics. We did not evaluate and analyze other potentially relevant aspects such as avoidance of nosocomial infections, reduction of adverse effects of sedation and muscle relaxants, and costs, as these data were rarely or never reported. Finally, a few articles reported mortality data, length of ICU, and hospital stay.


The present meta-analysis suggests that in infants and children, NIV applied for ARF reduces the intubation rate when compared with standard oxygen therapy or HFNC; these findings are strengthened excluding trials with the application of treatments shorter than 1 hour and above all when excluding trials allowing cross-over as rescue treatment. Mortality and length of stay were not different between NIV and controls. Larger studies are required to confirm that this can translate into a beneficial effect on survival.


1. Nava S, Hill N: Non-invasive ventilation in acute respiratory failure. Lancet 2009; 374:250–259
2. Rochwerg B, Brochard L, Elliott MW, et al.: Official ERS/ATS clinical practice guidelines: Noninvasive ventilation for acute respiratory failure. Eur Respir J 2017; 50:1602426
3. Cabrini L, Landoni G, Oriani A, et al.: Noninvasive ventilation and survival in acute care settings: A comprehensive systematic review and metaanalysis of randomized controlled trials. Crit Care Med 2015; 43:880–888
4. Cabrini L, Landoni G, Bocchino S, et al.: Long-term survival rate in patients with acute respiratory failure treated with noninvasive ventilation in ordinary wards. Crit Care Med 2016; 44:2139–2144
5. Najaf-Zadeh A, Leclerc F: Noninvasive positive pressure ventilation for acute respiratory failure in children: A concise review. Ann Intensive Care 2011; 1:15
6. Viscusi CD, Pacheco GS: Pediatric emergency noninvasive ventilation. Emerg Med Clin North Am 2018; 36:387–400
7. Luján M, Peñuelas O, Cinesi Gómez C, et al.: Sumario de las recomendaciones y puntos clave del Consenso de las Sociedades Científicas Españolas (SEPAR, SEMICYUC, SEMES; SECIP, SENeo, SEDAR, SENP) para la utilización de la ventilación no invasiva y terapia de alto flujo con cánulas nasales en el paciente adulto, pediátrico y neonatal con insuficiencia respiratoria aguda grave. Med Intensiva 2021; 45:298–312
8. Mandelzweig K, Leligdowicz A, Murthy S, et al.: Non-invasive ventilation in children and adults in low- and low-middle income countries: A systematic review and meta-analysis. J Crit Care 2018; 47:310–319
9. Sessions KL, Smith AG, Holmberg PJ, et al.: Continuous positive airway pressure for children in resource-limited settings, effect on mortality and adverse events: Systematic review and meta-analysis. Arch Dis Child 2022; 107:543–552
10. Sterne JAC, Savović J, Page MJ, et al.: RoB 2: A revised tool for assessing risk of bias in randomised trials. BMJ 2019; 366:l4898
11. Nikolakopoulou A, Mavridis D, Salanti G: How to interpret meta-analysis models: Fixed effect and random effects meta-analyses. Evid Based Mental Health 2014; 17:64–64
12. IntHout J, Ioannidis JPA, Borm GF: The Hartung-Knapp-Sidik-Jonkman method for random effects meta-analysis is straightforward and considerably outperforms the standard DerSimonian-Laird method. BMC Med Res Methodol 2014; 14:25
13. Teresa Greco GB-Z, Rin AZ: How to impute study-specific standard deviations in meta-analyses of skewed continuous endpoints? World J Meta-Anal 2015; 3:215–224
14. Wan X, Wang W, Liu J, et al.: Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol 2014; 14:135
15. Basnet S, Mander G, Andoh J, et al.: Safety, efficacy, and tolerability of early initiation of noninvasive positive pressure ventilation in pediatric patients admitted with status asthmaticus: A pilot study*. Pediatr Crit Care Med 2012; 13:393–398
16. Cam BV, Tuan DT, Fonsmark L, et al.: Randomized comparison of oxygen mask treatment vs. nasal continuous positive airway pressure in dengue shock syndrome with acute respiratory failure. J Trop Pediatr 2002; 48:335–339
17. Cesar RG, Bispo BRP, Felix PHCA, et al.: High-flow nasal cannula versus continuous positive airway pressure in critical bronchiolitis: A randomized controlled pilot. J Pediatr Intensive Care 2020; 9:248–255
18. Chisti MJ, Salam MA, Smith JH, et al.: Bubble continuous positive airway pressure for children with severe pneumonia and hypoxaemia in Bangladesh: An open, randomised controlled trial. The Lancet 2015; 386:1057–1065
19. Fioretto JR, Ribeiro CF, Carpi MF, et al.: Comparison between noninvasive mechanical ventilation and standard oxygen therapy in children up to 3 years old with respiratory failure after extubation: A pilot prospective randomized clinical study. Pediatr Crit Care Med 2015; 16:124–130
20. Lal SN, Kaur J, Anthwal P, et al.: Nasal continuous positive airway pressure in bronchiolitis: A randomized controlled trial. Indian Pediatr 2018; 55:27–30
21. Liu C, Cheng WY, Li JS, et al.: High-flow nasal cannula vs. continuous positive airway pressure therapy for the treatment of children <2 years with mild to moderate respiratory failure due to pneumonia [Internet]. Front Pediatr 2020; 8:590906
22. McCollum ED, Mvalo T, Eckerle M, et al.: Bubble continuous positive airway pressure for children with high-risk conditions and severe pneumonia in Malawi: An open label, randomised, controlled trial. Lancet Respir Med 2019; 7:964–974
23. Milési C, Essouri S, Pouyau R, et al.: High flow nasal cannula (HFNC) versus nasal continuous positive airway pressure (nCPAP) for the initial respiratory management of acute viral bronchiolitis in young infants: A multicenter randomized controlled trial (TRAMONTANE study). Intensive Care Med 2017; 43:209–216
24. Ramnarayan P, Lister P, Dominguez T, et al.: FIRST-line support for Assistance in Breathing in Children (FIRST-ABC): A multicentre pilot randomised controlled trial of high-flow nasal cannula therapy versus continuous positive airway pressure in paediatric critical care. Crit Care 2018; 22:144
25. Ramnarayan P, Richards-Belle A, Drikite L, et al.: Effect of high-flow nasal cannula therapy vs continuous positive airway pressure following extubation on liberation from respiratory support in critically ill children: A randomized clinical trial. JAMA 2022; 327:1555–1565
26. Ramnarayan P, Richards-Belle A, Drikite L, et al.: Effect of high-flow nasal cannula therapy vs continuous positive airway pressure therapy on liberation from respiratory support in acutely ill children admitted to pediatric critical care units: A randomized clinical trial. JAMA 2022; 328:162–172
27. Sarkar M, Sinha R, Roychowdhoury S, et al.: Comparative study between noninvasive continuous positive airway pressure and hot humidified high-flow nasal cannulae as a mode of respiratory support in infants with acute bronchiolitis in pediatric intensive care unit of a tertiary care hospital. Indian J Crit Care Med 2018; 22:85–90
28. Vitaliti G, Vitaliti MC, Finocchiaro MC, et al.: Randomized comparison of helmet CPAP versus high-flow nasal cannula oxygen in pediatric respiratory distress. Respir Care 2017; 62:1036–1042
29. Yañez LJ, Yunge M, Emilfork M, et al.: A prospective, randomized, controlled trial of noninvasive ventilation in pediatric acute respiratory failure*. Pediatr Crit Care Med 2008; 9:484–489
30. Rotta AT, Wiryawan B: Respiratory emergencies in children. Respir Care 2003; 48:248–258
31. Korang SK, Feinberg J, Wetterslev J, et al.: Non-invasive positive pressure ventilation for acute asthma in children. Cochrane Database Syst Rev 2016; 9:CD012067
32. Tang G, Lin J, Zhang Y, et al.: The effects and safety of continuous positive airway pressure in children with bronchiolitis: A systematic review and meta-analysis. J Trop Pediatr 2021; 67:fmaa128
33. Rolim D S, Galas FRB, Faria LS, et al.: Use of noninvasive ventilation in respiratory failure after extubation during postoperative care in pediatrics. Pediatr Cardiol 2020; 41:729–735
34. Abadesso C, Nunes P, Silvestre C, et al.: Non-invasive ventilation in acute respiratory failure in children. Pediatr Rep 2012; 4:e16
35. Pons-Òdena M, Medina A, Modesto V, et al.: [What are the most reliable predictive factors of non-invasive ventilation failure in paediatric intensive care units?]. An Pediatr (Engl Ed) 2019; 91:307–316
36. Kopp W, Gedeit RG, Asaro LA, et al.: The impact of preintubation noninvasive ventilation on outcomes in pediatric acute respiratory distress syndrome*. Crit Care Med 2021; 49:816–827
37. Kyle JM, Sturza JM, Dechert RE, et al.: Clinical outcomes of acute respiratory failure associated with noninvasive and invasive ventilation in a pediatric ICU. Respir Care 2022; 67:956–966

intubation; mortality; noninvasive ventilation; pediatric

Supplemental Digital Content

Copyright © 2022 by the Society of Critical Care Medicine and the World Federation of Pediatric Intensive and Critical Care Societies