Authors: Mohamad El-Khatib (Department of Anesthesiology and Pain Medicine, American University of Beirut Medical Center, Beirut, Lebanon), Nizar El Bcherawi (Department of Anesthesiology and Pain Medicine, American University of Beirut Medical Center, Beirut, Lebanon), Frida Atallah (Department of Anesthesiology and Pain Medicine, American University of Beirut Medical Center, Beirut, Lebanon), Marc Moukarzel (Department of Anesthesiology and Pain Medicine, American University of Beirut Medical Center, Beirut, Lebanon), Thuraya HajAli (Department of Anesthesiology and Pain Medicine, American University of Beirut Medical Center, Beirut, Lebanon), Joanne Khabsa (Clinical Research Institute, American University of Beirut, Beirut, Lebanon), Hassan Moukalled (Department of Anesthesiology and Pain Medicine, American University of Beirut Medical Center, Beirut, Lebanon), Lynn Sibai (Department of Anesthesiology and Pain Medicine, American University of Beirut Medical Center, Beirut, Lebanon), Patrick Maroun (Department of Anesthesiology and Pain Medicine, American University of Beirut Medical Center, Beirut, Lebanon), Christian Raphael (Department of Anesthesiology and Pain Medicine, American University of Beirut Medical Center, Beirut, Lebanon)
Source: The Cochrane Database of Systematic Reviews
Authors: Mohamad El-Khatib, Nizar El Bcherawi, Frida Atallah, Marc Moukarzel, Thuraya HajAli, Joanne Khabsa, Hassan Moukalled, Lynn Sibai, Patrick Maroun, Christian Raphael
This is a protocol for a Cochrane Review (intervention). The objectives are as
To assess the benefits and harms of high‐flow nasal cannula (HFNC) use, versus conventional oxygen therapy or other non‐invasive ventilation, for respiratory support in children for indications other than acute bronchiolitis.
Respiratory distress is a common presentation in children in pediatric emergency and critical care settings. It occurs when the respiratory system cannot meet the metabolic demands for oxygenation or ventilation, and is characterized by tachypnea (i.e. a respiratory rate greater than normal, resulting in abnormally rapid and shallow breathing), use of accessory muscles (recruitment of additional neck and chest muscles to assist breathing), and thoracic retractions (inward pulling of the chest wall during inspiration) [1]. Respiratory support may be required due to a variety of underlying disease including upper respiratory tract conditions, such as croup, epiglottitis, anaphylaxis or angioedema, bacterial tracheitis, and foreign body aspiration (acute conditions that may cause upper airway obstruction); lower respiratory tract diseases, such as asthma and bronchiolitis (a common infection of the small airways); lung parenchymal disorders (conditions affecting lung tissue and gas exchange), such as pneumonia and other respiratory infections; neuromuscular disorders; dysfunction of the respiratory control centers; and other causes including trauma, injury, post‐surgical interventions, and upper airway obstruction [2].
In these clinical contexts, impaired gas exchange may occur from hypoventilation (inadequate ventilation to meet gas exchange needs), airway obstruction, alveolar collapse (closure of small airways and alveoli), or increased metabolic demand. Timely initiation of appropriate respiratory support is critical to prevent progression to respiratory failure, reducing the need for invasive and non‐invasive mechanical ventilation, and improving clinical outcomes [3]. Respiratory support can be provided non‐invasively, through oxygen therapy, continuous positive airway pressure (CPAP), or bilevel positive airway pressure (BiPAP); or invasively, via endotracheal intubation and mechanical ventilation.
Children with severe respiratory distress and hypoxemia (low blood oxygen levels) often require mechanical ventilation, which can result in trauma to the lungs and airways, collectively termed ventilator‐induced lung injury (VILI). In addition to VILI, intubation and invasive ventilation are associated with other complications, including ventilator‐associated pneumonia, laryngeal‐tracheal damage (injury to upper airway structures), and systemic effects of increased sedative drug use, which can contribute to neuropathy and myopathy (nerve and muscle dysfunction), prolonged recovery, and the need for cardiovascular support with vasoactive infusions (intravenous medications to support blood pressure and circulation). These factors increase morbidity and the overall cost of care [3, 4, 5]. Therefore, non‐invasive respiratory support provides respiratory assistance through techniques designed to enhance alveolar ventilation, improve oxygenation, and reduce the work of breathing without requiring an invasive tracheal tube [3].
High‐flow nasal cannula (HFNC) oxygen therapy, also referred to as heated, humidified high‐flow nasal cannula (HHHFNC), is a non‐invasive respiratory support commonly used in people (from preterm infants to adults) [6, 7]. It delivers a heated, humidified, and titrated mixture of air and oxygen through specialized nasal prongs via a system that includes a flow generator, oxygen‐air blender for fraction of inspired oxygen (FiO2) adjustment, heated humidifier, delivery circuit, and soft nasal interface, with optional pressure‐relief valves for safety [8]. In pediatric patients, flow rates are weight‐dependent; neonates aged one‐month‐old or less (< 4 kg) typically receive 5 to 8 L/min, infants between one month and under one‐year‐old (4 kg to 10 kg) receive 8 to 20 L/min, children between one year to under six years old (10 kg to 20 kg) receive 12 to 25 L/min, children between six to under 12 years old (20 kg to 40 kg) receive 20 to 30 L/min, and adolescents aged 12 years or older, or weighing more than 40 kg, receive 25 to 50 L/min, with cannula sizes adjusted accordingly [9]. HFNC is generally well‐tolerated, and may permit speaking, feeding, and interaction more easily. Also, it is reportedly better tolerated than other non‐invasive ventilation techniques, such as CPAP and BiPAP, improving the rate of compliance therapy and reducing the need for sedation [10]. Indications for HFNC include moderate to severe respiratory distress or hypoxemia from infectious and inflammatory lower respiratory tract conditions (e.g. pneumonia, bronchiolitis, asthma exacerbations), airway and structural abnormalities (e.g. tracheomalacia, an abnormal softness and collapse of the tracheal airway during breathing), peri‐extubation respiratory insufficiency, and preoxygenation prior to airway management. Contraindications include conditions associated with impaired ventilation or carbon dioxide retention (e.g. hypercapnic respiratory failure), anatomical factors that prevent adequate cannula fit (craniofacial anomalies such as choanal atresia, a congenital obstruction of the posterior nasal passages), and clinical conditions that increase aspiration or pressure‐related lung injury risks (excessive secretions, vomiting, nose bleeding, or pulmonary air leak) [11, 12, 13, 14].
HFNC exerts its physiological effects through several mechanisms. High flow rates wash out nasopharyngeal dead space, enhancing oxygen delivery and alveolar ventilation, while reducing CO2 retention and apneic episodes (periods of paused breathing). Matching or exceeding spontaneous inspiratory flow lowers upper airway resistance and decreases the work of breathing. Heated and humidified gas reduces mucosal injury, preserves mucociliary function, and minimizes metabolic demand. At sufficiently high flows, HFNC may also generate low levels of CPAP, helping to maintain airway patency (keeping the airway open) and overcome upper airway obstruction. The continuous high flow also generates a mild, variable level of positive end‐expiratory pressure (PEEP), particularly when the mouth is closed, contributing to increased end‐expiratory lung volume (the amount of air remaining in the lungs after exhalation), alveolar recruitment (opening of previously collapsed air sacs), and reduced atelectasis (lung collapse) [15, 16, 17]. HFNC may also induce favorable hemodynamic effects by modulating intrathoracic pressures, which can reduce right ventricular preload, an effect potentially beneficial in patients with cardiogenic pulmonary edema (fluid accumulation in the lungs due to heart failure) [18]. Moreover, the improved tolerability, speech compatibility, and reduced discomfort associated with nasal cannula enhance the suitability of HFNC in a wide range of pediatric clinical scenarios. An increasing number of studies have reported the use of HFNC in a range of pediatric indications beyond bronchiolitis, such as severe asthma [19], post‐extubation care in critically ill children [20], and severe pneumonia with hypoxemia [21].
Respiratory distress in children results from a wide range of conditions beyond acute bronchiolitis, many of which carry significant risks of morbidity and mortality. While HFNC therapy is well‐established for bronchiolitis [22], its efficacy and safety in other pediatric respiratory disorders (such as asthma exacerbations, pneumonia, and post‐extubation support) are less clearly defined. Emerging studies and randomized controlled trials (RCTs) suggest HFNC may improve clinical outcomes and reduce the need for invasive ventilation in these populations, but the evidence is inconsistent [23]. A comprehensive systematic review is therefore needed to systematically evaluate the role of HFNC oxygen therapy across diverse pediatric respiratory indications, enabling clinicians to make informed decisions and optimize care for critically ill children.
This Cochrane protocol supersedes a 2014 Cochrane review that evaluated HFNC therapy for respiratory support in children [24].
To assess the benefits and harms of high‐flow nasal cannula (HFNC) use, versus conventional oxygen therapy or other non‐invasive ventilation, for respiratory support in children for indications other than acute bronchiolitis.
We will follow the Methodological Expectations for Cochrane Intervention Reviews (MECIR) when conducting the review [25], and PRISMA 2020 for reporting [26].
We will include RCTs, quasi‐RCTs (which do not use true randomization to assign participants to groups but instead use alternative methods, such as date of birth or alternation), and cluster‐RCTs. We will not restrict eligibility by the year of publication and language. We will exclude conference abstracts.
We will include all pediatric patients up to 18 years of age, including neonates, regardless of the pharmacologic treatment given in conjunction with ventilatory support. The settings of interest are the hospital emergency department, general pediatric wards and pediatric intensive care units (PICU), as well as perioperative use of HFNC (as post‐extubation therapy). We will exclude patient populations that present with bronchiolitis, as the evidence regarding this indication is already covered by several systematic reviews [22, 27, 28]. We will only include studies of mixed adult and pediatric populations if the pediatric‐specific data are available, or if we can obtain these data from the study authors. Where such data are not reported separately, we will contact study authors to request the relevant pediatric information.
For this Cochrane review, we define high‐flow nasal oxygen as the delivery of heated and humidified oxygen, or an air–oxygen mixture, via nasal cannula at flow rates greater than 5 L/min. We will include studies in which HFNC is used without restrictions on the frequency or duration of therapy (as those are tightly linked to underlying condition and patient’s clinical evaluation), regardless of the FiO2, titration strategy, or concomitant pharmacological treatments, provided the intervention reflects usual clinical practice.
We will compare HFNC with other non‐invasive respiratory support modalities, including standard low‐flow oxygen therapy (e.g. nasal cannula or face mask), CPAP, or BiPAP.
**Stabilization without escalation of oxygen ** proportion of participants who achieve clinical stabilization without the need for escalation to higher levels of respiratory support or during the treatment period (e.g. non‐invasive ventilation or invasive mechanical ventilation)Time to ** time (in hours) from initiation of HFNC to achievement of clinical stabilization without escalation of respiratory supportPatient ** final patient disposition following emergency department management, including discharge home, hospital ward admission, or intensive care unit admission
Mortality: all‐cause mortality within 24 hours, 48 hours post‐randomization, and all‐cause in‐hospital mortality**Treatment ** occurrence of oxygen desaturation during HFNC application necessitating escalation to other forms of non‐invasive ventilation or intubation
The important outcomes include measures of cost‐effectiveness, underlying disease severity, and physiologic indicators of respiratory status. These outcomes provide contextual information on disease burden and efficiency of care delivery, and support interpretation of the intervention’s clinical effectiveness.
Length of hospital ** total duration of hospitalization, measured in days from admission to hospital dischargeLength of PICU ** duration of stay in the PICU, measured in days from PICU admission to PICU discharge
**Change in underlying disease ** worsening of the primary respiratory condition, defined as the need for escalation of treatment (e.g. asthma treatment)
Hypercapnia: increase in arterial carbon dioxide levels, defined as an elevation in PaCO2 measured on arterial blood gas analysis during HFNC applicationApnea: occurrence of temporary cessation of breathing or respiratory arrest during HFNC application, as documented clinically or by monitoringTachypnea: presence of rapid and shallow breathing during HFNC application, defined according to age‐appropriate respiratory rate thresholds used by individual studies**Gas ** ratio of arterial oxygen partial pressure (PaO2) to FiO2 (PaO2/FiO2), measured during HFNC application as an indicator of gas exchange efficiency
Nasal mucosal ** occurrence of nasal mucosal injury or ulceration attributed to HFNC nasal prongs, as reported by study authorsGastric ** occurrence of clinically or radiologically identified gastric distension or insufflation during HFNC therapyAspiration: occurrence of clinically suspected or confirmed aspiration events during HFNC therapyPneumonia: incidence of new‐onset pneumonia diagnosed clinically or radiographically, or both, during or following HFNC treatmentOrgan ** occurrence of new organ dysfunction or failure (e.g. respiratory, cardiovascular, renal) as defined by the study authorsInfectious ** occurrence of systemic or localized infections, including sepsis or bacteremia, diagnosed during the study follow‐up period
We will search the following databases from inception to present the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE (via Ovid), and Embase (via Elsevier). In addition, we will search the databases of ongoing trials, including the Clinical Trials register (clinicaltrials.gov), the EU Clinical Trials Register (www.clinicaltrialsregister.eu), and the World Health Organization (WHO) International Clinical Trials Registry Platform (www.who.int/tools/clinical‐trials‐registry‐platform). We will summarize information, such as trial identifiers and other relevant details about ongoing studies, in the ‘Characteristics of ongoing studies’ table.
To ensure that we only capture RCTs, we will apply the Cochrane highly sensitive search strategy for identifying RCTs and controlled clinical trials [29]. The detailed MEDLINE search strategy is in Supplementary material 1. Our MEDLINE strategy will be adapted for use in other bibliographic databases.
We will conduct a full updated search across all specified databases four weeks before submission of the final review draft for editorial approval. If we identify additional relevant keywords during the electronic or manual searches, we will update the search strategies accordingly and document all modifications. No language restrictions will be applied.
We will try to identify other potentially eligible trials or ancillary publications by searching the reference lists of retrieved included trials, related systematic reviews, or other reviews. We will also search for potentially eligible trials on the following
Society for Pediatric Anesthesia (pedsanesthesia.org);American Society of Anesthesiologists (www.asahq.org); andEuropean Society of Paediatric and Neonatal Intensive Care (www.espnic.eu).
We will import citations identified by the search to the Covidence software platform, where deduplication will be performed [30]. Before starting the selection process, two review authors will conduct calibration exercises to ensure the validity of the process. We will develop screening forms for both stages (title/abstract screening and full‐text eligibility assessment) to guide the review authors in the selection process. The review authors performing the screening process will not be blinded to study authors, journal of publication, or study outcomes.
Two review authors will independently assess the title and abstract of every retrieved citation for potential eligibility. The two review authors will then independently assess the full‐text articles for eligibility using a standardized, pilot‐tested, screening form. They will compare their results and resolve any disagreements by consensus, or with the help of a third review author when needed. Eligible articles published in languages other than English, Arabic, or French will be translated using DeepL software as needed.
We will report on the results of our search and screening process using a PRISMA flow diagram. We will list the articles excluded after full‐text assessment, and the reasons for their exclusion, in a ‘Characteristics of excluded studies’ table.
Before we start the data extraction process, we will conduct calibration exercises to ensure consistency across review authors. Two review authors will independently extract data from each included study using a standardized data extraction form, which we will pilot test on five studies. We will resolve any disagreements by discussion and by consulting a third review author as needed.
We will seek relevant missing information on the included trial(s) from the primary author(s) of the article, if required.
The following data will be extracted from the included population, intervention, control, and critical and important outcomes. We will also extract statistical data needed for the meta‐analysis.
Two review authors will independently assess the risk of bias of each included study. We will resolve disagreements by consensus, or by consulting a third review author. We will assess the risk of bias using the Cochrane RoB 2 tool [31], applied at the outcome level, and with the effect of interest being the effect of assignment to the intervention (i.e. intention‐to‐treat analysis). We will assess the following
bias arising from the randomization process;bias due to deviations from intended interventions;bias due to missing outcome data;bias in measurement of the outcome; andbias in selection of the reported result.
We will assess outcome reporting bias by comparing the outcomes listed in the trial protocol, registration record, and methods section, with the outcomes for which data are reported in the results section [32]. We will judge the risk of bias criteria as either ‘low risk’, ‘high risk’, or ‘some concerns’. These judgments will be informed by signaling questions provided in the RoB 2 guidance and an overall risk of bias will be derived for each outcome [33].
If any cluster‐RCTs meet the inclusion criteria of the review, we will base the assessment of their risk of the bias on the revised Cochrane risk‐of‐bias tool for cluster‐randomized trials (RoB 2 CRT), which includes one additional bias arising from the timing of identification and recruitment of participants. We will assess bias due to loss of clusters (e.g. entire clusters withdrawing from the study) as an additional consideration within the pre‐existing domain of ‘bias due to missing outcome data’.
We will assess the risk of bias for the following critical
stabilization without escalation of oxygen therapy;time to stabilization;patient's disposition after emergency;mortality within 24 and 48 hours post‐randomization, and all‐hospital mortality; andtreatment failure requiring escalation to non‐invasive ventilation or intubation.
The RoB 2 tool and its extension for cluster trials are described in Chapter 8 and Chapter 23 of the Cochrane Handbook for Systematic Reviews of Interventions [33, 34].
For dichotomous outcomes, we will use risk ratios (RRs) with 95% confidence intervals (CIs). For time‐to‐event outcomes, we will use hazard ratios (HRs) with 95% CIs. Continuous data will be expressed as mean differences (MDs) with 95% CIs when all studies report the outcome using the same scale, and as standardized mean differences (SMDs) when studies report the outcome using different scales.
If the included studies have reported rate data (i.e. counts measured for each participant along with observation time), we will pool rate ratios.
When studies report medians and interquartile ranges (IQRs), we will assume the median approximates the mean only when the data appears to be symmetrically distributed. In such cases, we will estimate the standard deviations using the guidance provided in Chapter 6 of the Cochrane Handbook for Systematic Reviews of Interventions [35].
The unit of analysis will be the individual participant. For cluster‐RCTs, if clustering has not been accounted for in the analysis, we will adjust for clustering effect using the reported intracluster correlation coefficient (ICC). If the ICC is not reported, we will attempt to obtain it from study authors. If unsuccessful, we will use an ICC from similar studies and conduct sensitivity analyses to assess the robustness of the results. For studies reporting repeated measurements of outcomes at multiple time points, we will either select the most clinically relevant time point or treat each time point as a separate outcome. If more than one intervention arm from a multiple‐arm study is eligible for inclusion in the same meta‐analysis, we will combine the relevant intervention arms to create a single pairwise analysis. We will omit groups from multiple‐arm studies that are irrelevant to our comparison of interest.
We will contact study authors to request any missing outcome data. We will analyze data according to the available‐case analysis approach, including only participants with complete outcome data. For studies reporting both intention‐to‐treat (ITT) and per‐protocol (PP) analyses, we will prioritize and extract ITT data.
In addition, we will assess the potential impact of missing data on the review findings by conducting a sensitivity analysis excluding studies with high levels of missing data.
We will assess outcome reporting bias by comparing the outcomes listed in the trial protocol, registration record, and methods section, with the outcomes for which data are reported in the results section [32]. In addition, we will create a funnel plot, and we aim to test the funnel plot asymmetry if we include more than 10 studies in the meta‐analysis. While constructing the funnel plot, we will plot the effect estimates against their standard errors, and we will also present a 95% confidence region based on a fixed‐effect model to simplify the interpretation.
For categorical outcomes, such as treatment failure (yes/no), escalation to non‐invasive ventilation (yes/no), or mortality (yes/no), we will calculate RRs with 95% CIs for each study and pool them using a random‐effects model to account for between‐study heterogeneity. For continuous outcomes, such as length of hospital stay, length of PICU stay, or physiological measures (e.g. PaO2/FiO2 ratio), we will conduct a random‐effects meta‐analysis using the inverse variance method.
For time‐to‐event outcomes, we will pool log hazard ratios (HRs) using a random‐effects model with the generic inverse variance method. We will conduct meta‐analyses using Review Manager (RevMan) [36].
The main comparison of interest is HFNC versus all other conventional non‐invasive oxygen therapy (including low‐flow nasal cannula, oxygen mask therapy, CPAP and BiPAP).
We will use a random‐effects model since clinical diversity is expected. The Restricted Maximum Likelihood (REML) estimator will be used to estimate between‐trial variance (τ²). Where at least three studies are included and heterogeneity is greater than zero, we will calculate CIs using the Hartung‐Knapp‐Sidik‐Jonkman method [37]; for pooled analyses of two studies or where τ² = 0, we will use the Wald‐type method [38].
We will evaluate heterogeneity using the I² statistic and τ², supplemented by visual inspection of forest plots. If meta‐analysis is not feasible due to substantial clinical or methodological diversity, we will provide a narrative synthesis.
Given that some outcomes may be reported within the same studies and potentially correlated (e.g. treatment failure and escalation to intubation), we will consider multivariate meta‐analysis to simultaneously analyze related outcomes while accounting for correlations. Multivariate analyses will be performed using the mvmeta package in R [39], applying a random‐effects model and REML estimator for τ². Since within‐study correlations are rarely reported, we will conduct sensitivity analyses by imputing correlations across a plausible range (0 to 1, increments of 0.1), assuming negative correlations are unlikely. If multivariate methods are not feasible, we will perform separate meta‐analyses for each outcome.
We will perform subgroup analyses to explore potential sources of clinical heterogeneity. Subgroups include age, sex, underlying respiratory condition and its severity, type of comparator, HFNC flow rate, and clinical care setting. We have described these subgroups, along with their rationale, in further detail below.
We will assess subgroup comparisons using the test for subgroup differences in RevMan [36], as recommended in Chapter 10 of the Cochrane Handbook for Systematic Reviews of Interventions [38]. This approach will allow us to evaluate whether the effects of HFNC differ significantly across the following prespecified subgroups.
Age ** airway anatomy, respiratory mechanics, tolerance of HFNC, and risk of adverse events differ substantially by ageInfants less than 12 months oldToddlers from 12 months to less than three years oldPreschool children from three years to under six years oldSchoolchildren from six years to under 13 years oldAdolescents from 13 years to under 18 years oldUnderlying respiratory ** the physiological benefits and risks of HFNC vary by disease mechanism (e.g. airway obstruction vs parenchymal disease)Airway obstructionAsthma or reactive airway diseaseCroupPost‐extubation stridor (operative or non‐operative)Parenchymal lung diseasePneumonia (viral, bacterial, or atypical)Chronic parenchymal lung diseases such as cystic fibrosis, interstitial lung disease, or bronchopulmonary dysplasiaNeuromuscular or respiratory muscle weaknessNeuromuscular disordersRespiratory muscle weakness from other causesMixed or other acute hypoxemic respiratory failureCases where more than one mechanism contributes (e.g., pneumonia with airway obstruction)Acute respiratory distress syndrome (ARDS) if reported in pediatric studiesBaseline severity of respiratory ** treatment effects and failure rates are likely to differ with baseline severityMild respiratory distressModerate respiratory distressSevere respiratory distressType of ** the expected magnitude of benefit and harm differs depending on the comparator modalityConventional low‐flow oxygen therapy nasal cannulaface maskCPAPBiPAP**Clinical care ** monitoring intensity, escalation thresholds, and outcomes may vary by clinical settingEmergency departmentPediatric wardPICU
Although the main focus of this review is to evaluate the use of HFNC in pediatric populations, we recognize the importance of considering potential equity‐related differences. Age and sex are prespecified subgroups in our analyses and will serve as the primary equity‐related assessments. By examining outcomes across these subgroups, we can explore whether the effects of HFNC vary by age or sex, providing insight into the intervention’s applicability and effectiveness across diverse pediatric populations. In addition, we will extract and report on sex and study geographical location.
We will conduct sensitivity analyses to explore the influence of the following factors (when applicable) on effect sizes, following the guidance in Section 10.14 of the Cochrane Handbook for Systematic Reviews of Interventions [38].
Excluding studies with high risk of biasExcluding studies with a large amount of missing data
For the multivariate meta‐analysis, we will perform sensitivity analyses by imputing the within‐study correlation over the full plausible range from 0 to 1, in increments of 0.1.
We will assess and present the overall certainty of evidence for each specified outcome using the GRADE approach, which considers the following risk of bias, inconsistency, indirectness, imprecision, and publication bias [40]. Two review authors will independently evaluate the certainty of evidence for each outcome, and will resolve any disagreements through discussion or by consulting a third review author when needed.
For the comparison of interest (HFNC versus standard low‐flow oxygen therapy), we will present our findings in a summary of findings table for the critical stabilization without escalation of oxygen therapy, patient disposition, mortality, and treatment failure (requiring escalation to non‐invasive ventilation or intubation).
In the summary of findings table, we will include the following information for each selected time point at which the outcome was measured, method of measurement, number of studies, number of participants contributing data, and the certainty of the evidence. We will justify our decision(s) to downgrade the certainty of evidence with informative footnotes, in accordance with GRADE guidance [40].
We did not involve consumers in the protocol development, and we do not plan to involve them in participation at the review stage. In the absence of direct consumer input, we have addressed patient‐ and public‐relevant perspectives by prioritizing clinically important outcomes and relevant time points. We will ensure that the plain language summary is clear and accessible to lay readers, and will aim to interpret and present the findings with a patient‐ and family‐centered perspective.
Supplementary materials are available with the online version of this 10.1002/14651858.CD016323.
Supplementary materials are published alongside the article and contain additional data and information that support or enhance the article. Supplementary materials may not be subject to the same editorial scrutiny as the content of the article and Cochrane has not copyedited, typeset or proofread these materials. The material in these sections has been supplied by the author(s) for publication under a Licence for Publication and the author(s) are solely responsible for the material. Cochrane accordingly gives no representations or warranties of any kind in relation to, and accepts no liability for any reliance on or use of, such material.