Continuous positive airway pressure (CPAP) for acute bronchiolitis in children

  • Protocol
  • Intervention



This is the protocol for a review and there is no abstract. The objectives are as follows:

To assess the efficacy and safety of CPAP compared to no CPAP or sham CPAP in infants and children up to three years of age with acute bronchiolitis.


Description of the condition

Acute bronchiolitis is one of the most frequent causes of emergency department visits and hospitalisation in infants (Deshpande 2003; Handforth 2000). Bronchiolitis (inflammation of the small airways in the lung) is predominantly a viral disease and usually affects infants and children younger than three years of age. It is mostly caused by respiratory syncytial virus (RSV) (CDC 2010). About 2% to 3% of all infants require hospitalisation due to bronchiolitis in the USA (Meissner 2009). Bronchiolitis occurs more frequently in males infants who are not breast fed, and who live in crowded conditions (Meates-Dennis 2005).

Bronchiolitis typically presents with viral symptoms (sneezing, rhinorrhoea and fever) which gradually progress to paroxysmal cough, wheezing, respiratory distress and irritability. Chest findings are non-specific and include wheezing, with or without fine crackles. Although not required for diagnosis, chest X-ray may reveal hyperinflated lungs with patchy atelectasis. About 10% to 15% of patients hospitalised with bronchiolitis respond poorly to treatment and require intensive care management. Further, nearly half of these develop respiratory failure and need mechanical ventilation (Navas 1992). Although uncommon, bronchiolitis may cause mortality which ranges from 0.5% to 2% (Kabir 2003; Levy 1997). The mortality rate is higher in low-income countries.

The management of bronchiolitis mainly includes supportive measures like adequate fluid intake, antipyretics and humidified oxygen supplementation if hypoxia is present (Davison 2004). Nebulised adrenaline (Hartling 2011a; Hartling 2011b) and hypertonic nebulised saline (Zhang 2011) have been found to be beneficial in acute bronchiolitis. Other therapeutic options, such as corticosteroid therapies (Fernandes 2010), antibiotics (Spurling 2011), bronchodilators (Gadomski 2010), heliox inhalation therapy (Liet 2010), chest physiotherapy (Roqué i Figuls 2012), nebulised recombinant human deoxyribonuclease (Merkus 2001; Nasr 2001), ribavirin (Ventre 2007) and steam inhalation (Umoren 2011), have been tried with no definitive benefit in bronchiolitis.

Description of the intervention

Continuous positive airway pressure (CPAP) is the application of positive pressure to the airways of the spontaneously breathing patient throughout the respiratory cycle (Duncan 1986). It keeps the airways open. CPAP may be applied to infants using nasal prongs (NCPAP), nasopharyngeal tube (NP-CPAP) or infant nasal mask (NM-CPAP). It is administered with a commercially available circuit used in conjunction with a continuous flow source, or a ventilator. CPAP devices may include the function of providing a heated and humidified flow to the patient. The use of CPAP has been associated with some adverse effects which may include local and systemic effects, for example, nasal mucosal damage, nasal excoriation, scarring, pressure necrosis and septal distortion (Lee 2002; Robertson 1996), aspiration secondary to gastric insufflation (Kiciman 1998), pneumothorax (de Bie 2002) and a decrease in cardiac output due to impaired pulmonary blood flow.

How the intervention might work

In bronchiolitis the peripheral airways are most severely affected by inflammation. In infants with acute bronchiolitis expiratory resistance is found to be greater than inspiratory resistance suggesting dynamic narrowing of the airways on expiration (Wohl 1969). Acute bronchiolitis is associated with increased thoracic gas volume (air trapping) and total pulmonary resistance, and decreased dynamic compliance (Phelan 1968). Initially infants compensate for the increased physiological dead space by increasing respiratory rate resulting in increased minute volume. Gradually they becomes exhausted and minute volume falls with increase in partial pressure of carbon dioxide (PCO2) and hypoxaemia. From here, the infant may improve with oxygen supplementation or may progress to respiratory failure.

When CPAP is applied it increases the functional residual capacity of lungs which results in enlargement of the diameter of almost all airways including the peripheral airways. The widening of the peripheral airways allows deflation of over-distended lungs in bronchiolitis. The increase in airway pressure also prevents the collapse of poorly supported peripheral small airways during expiration. CPAP has been used in bronchiolitis and benefits have been noted in observational studies (Beasley 1981; Soong 1993). One of the advantages is that it may prevent the need for mechanical ventilation in infants with acute bronchiolitis.

Why it is important to do this review

Acute bronchiolitis is a common clinical condition in paediatrics, yet no specific treatment is available except for supportive therapy. CPAP is often used for its management on an empiric basis (i.e. based on personal experience without good evidence from literature). We aim to assess the role of CPAP for this condition and a systematic review is best suited for the purpose.


To assess the efficacy and safety of CPAP compared to no CPAP or sham CPAP in infants and children up to three years of age with acute bronchiolitis.


Criteria for considering studies for this review

Types of studies

Randomised controlled trials (RCT) and quasi-RCTS. Cross-over RCTs and cluster-RCTs will also be included.

Types of participants

Children up to three years of age with acute bronchiolitis. Bronchiolitis is usually diagnosed clinically, therefore studies including children with a clinical diagnosis of bronchiolitis will be eligible for inclusion. All infants, whether RSV-positive, negative or RSV not checked will be included.

Types of interventions

We will include CPAP with any pressure level and delivered by any type of device, by any mode (nasal prongs, face mask, etc.) and for any duration of time, compared to no CPAP or sham CPAP. We will include all studies with CPAP applied at any time after presentation. We will exclude studies which include high-flow nasal cannulae (HFNC) as there is separate review protocol for that. We will include trials where both arms receive similar management in all other respects.

Types of outcome measures

Primary outcomes
  1. Need for mechanical ventilation.

  2. Time to recovery (as defined by the included trial).  

Secondary outcomes
  1. Change in respiratory rate.

  2. Change in arterial oxygen saturation.

  3. Change in arterial pCO2 and partial pressure of oxygen (pO2).

  4. Hospital admission rate (from emergency department to hospital).

  5. Duration of emergency department stay.

  6. Duration of hospital stay.

  7. Need for intensive care unit admission.

  8. Adverse effects, for example, local nasal effects, pneumothorax and shock.

  9. Mortality.

Search methods for identification of studies

Electronic searches

We will search the Cochrane Central Register of Controlled Trials (CENTRAL) latest issue, MEDLINE (1966 to current date), EMBASE (1974 to current date), CINAHL (1981 to current date) and LILACS (1982 to current date).

We will use the following search strategy to search CENTRAL and MEDLINE. We will combine the MEDLINE search strategy with the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE (Lefebvre 2011). We will adapt the search strategy to search EMBASE, CINAHL and LILACS. We will use no date, language or publication restrictions.


1 exp Bronchiolitis/
2 bronchiolit*.tw.
3 Bronchopneumonia/
4 bronchopneumon*.tw.
5 respiratory syncytial viruses/ or respiratory syncytial virus, human/
6 Respiratory Syncytial Virus Infections/
7 (respiratory syncytial virus* or rsv).tw.
8 or/1-7
9 Respiratory Therapy/
10 Respiration, Artificial/
11 Positive-Pressure Respiration/
12 Continuous Positive Airway Pressure/
13 (positive pressure* adj5 (ventilat* or respir* or breath* or airway*)).tw.
14 positiv* airway* pressur*.tw.
15 (ppv or cpap or ncpap or nm-cpap or np-cpap).tw.
16 continuous distend* pressur*.tw.
17 or/9-16
18 8 and 17

Searching other resources

We will search and the WHO International Clinical Trials Registry Platform (ICTRP) for completed and ongoing trials. We will independently review the reference lists of included studies to identify additional studies, if any. One review author (KRJ) will contact corresponding authors of included trials to ask about any additional ongoing RCTs or unpublished trials.

Data collection and analysis

Selection of studies

Two review authors (KRJ, JLM) will independently assess the title and abstract of studies found through the search strategy to identify relevant studies for the review. We will retrieve the full text of the selected potentially relevant studies. We will independently assess the full text of identified relevant studies for inclusion as per the criteria mentioned. One review author (KRJ) will correspond with investigators, where appropriate, to clarify study eligibility if required. We will list excluded studies with reason(s) for exclusion. We will resolve any disagreements by discussion.

Data extraction and management

Two review authors (KRJ, JLM) will independently extract data using a pre-defined data collection form in accordance with the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), which will include the following data: source, eligibility, methods, participants and settings, interventions, outcomes, results, adverse effects and miscellaneous (funding source of the study, or potential conflicts of interest). We will resolve any disagreements by discussion.

Assessment of risk of bias in included studies

Two review authors (KRJ, JLM) will assess for risk of bias in included studies using the criteria described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). This will include a judgement for each specific feature of the study as 'low risk', 'high risk' or 'unclear risk' of bias ('unclear’ means either lack of information or uncertainty over the potential for bias). The specific features will be random sequence generation, allocation concealment, blinding of participants/personnel and blinding of outcome assessors, incomplete outcome data, selective reporting and other sources of bias. We will create a 'Risk of bias' graph figure and 'Risk of bias' summary figure using RevMan 2011.

Measures of treatment effect

We will analyse dichotomous outcomes by calculating the risk ratio (RR). We will express continuous outcome data as mean differences (MD). We will express the overall results with 95% confidence interval (CI).

Unit of analysis issues

We will include RCTs, quasi-RCTs, cross-over RCTs and cluster-RCTs. If we find any cluster-RCTs, they will be meta-analysed using the generic inverse variance method in RevMan 2011. We will add standard parallel-group trials into the same generic inverse variance meta-analysis.

Dealing with missing data

We will take the following steps if we encounter missing data.

  1. We will contact the trial investigators to request missing data whenever possible.

  2. We will perform sensitivity analyses to assess the effects of any assumptions if they are made.

  3. We will address the potential impact of missing data on the findings of the review in the Discussion section.

  4. We will calculate standard deviation (SD) from study statistics (for example, CI, standard errors, t values, P values or F values) for missing SDs of continuous outcome data. If SD calculation is still not possible, then we will impute from other studies in the meta-analysis as per the Cochrane Handbook for Systematic Reviews of Interventions recommendations (Higgins 2011).

We will extract data to allow an intention-to-treat (ITT) analysis, where possible, which aims to include all participants randomised into a trial irrespective of what happened subsequently. We will calculate and report the loss to follow-up if there is a discrepancy in the numbers randomised and the numbers analysed in each treatment group.

Assessment of heterogeneity

We will assess clinical and methodological heterogeneity before pooling. We will compare inclusion and exclusion criteria among included studies to assess clinical heterogeneity. We will carry out assessment for statistical heterogeneity visually by looking at forest plots, using a Chi2 test and using the I2 statistic. Using the Chi2 test, a low P value of < 0.1 (or a large Chi2 test statistic relative to its degree of freedom) will provide evidence of heterogeneity of intervention effects. We will interpret the value of the I2 statistic as follows: 0% to 40%, heterogeneity might not be important; > 40% to 60%, moderate heterogeneity; > 60% to 80%, substantial heterogeneity and > 80% to 100%, considerable heterogeneity.

Assessment of reporting biases

We will assess publication bias by funnel plots in RevMan 2011, if sufficient numbers of included studies are available.

Data synthesis

We will carry out meta-analyses using RevMan 2011. We will use a fixed-effect model for pooled data analysis. We will use a random-effects meta-analysis if there is important (more than 40%) statistical heterogeneity among studies. We will create a 'Summary of findings' (SOF) table using GRADEPro 2008 software. We will include the primary outcomes, hospital admission rate and adverse effects in the SOF table.

Subgroup analysis and investigation of heterogeneity

We will perform subgroup analyses for the following groups if sufficient data are available.

  1. CPAP with oxygen and CPAP with heliox.

  2. Different pressure levels of CPAP (below 6, 6 to 10, and > 10 cm water level (H2O)).

  3. Method of CPAP, via nasal prongs or with a face mask.

  4. Duration of CPAP (< 12 hours, 12 to 24 hours, > 24 hours).

  5. Trials with no CPAP and sham CPAP as comparator.

  6. RCTs and cross-over RCTs.

We will investigate the statistical heterogeneity by performing subgroup analyses and sensitivity analysis to determine the possible reason(s), for example, due to low quality trials.

Sensitivity analysis

We will perform sensitivity analyses to test the robustness of the results if a sufficient number of trials are found.

  1. Repeating meta-analysis after exclusion of studies with inadequate concealment of allocation.

  2. Repeating meta-analysis after exclusion of studies in which the outcome evaluation was not blinded.

  3. Repeating meta-analysis imputing missing data as best possible and worst possible outcome(s).

  4. Comparing the difference of pooling analysis results by using a fixed-effect model and a random-effects model.


We thank Sarah Thorning for adding the electronic search strategy to the protocol. We would like to acknowledge Prakesh Shah, Julia Walters, Edward Grandi, Elaine Beller and Lubna Al-Ansary for their comments on the protocol. We would also like acknowledge Liz Dooley for supporting us in drafting the protocol.

Contributions of authors

Dr Jat (KRJ) designed, drafted and collected material for the protocol.
Dr Mathew (JLM) reviewed the protocol and suggested important intellectual input.
Both the authors approved the final protocol.

Declarations of interest


Sources of support

Internal sources

  • Internet and library facility from Govt. Medical College Hospital, Chandigarh, India.

External sources

  • No sources of support supplied