Criteria for considering studies for this review
Types of studies
Randomised and quasi-randomised studies comparing palivizumab against placebo or no prophylaxis or another type of prophylaxis in the prevention of respiratory syncytial virus infection.
Types of participants
Infants and children (up to 18 years) of both sexes with diagnosis of cystic fibrosis made by either sweat test or genetic testing or clinical criteria, and of any disease severity.
Types of interventions
Prophylaxis with palivizumab (Synagis®) compared with either placebo or no prophylaxis or another type of prophylaxis. No limits were placed on setting, regimen or dose.
Types of outcome measures
Need for hospitalisation with respiratory syncytial virus (RSV) infection based on clinical diagnosis or validated laboratory diagnosis or both
Hospitalisation for RSV infection
Length of stay in hospital
Need for intensive care
Oxygen therapy for RSV infection
Need for oxygen therapy
Duration of oxygen therapy
forced expiratory volume in 1 second (FEV1) (absolute or per cent predicted for age, sex and height)
forced vital capacity (FVC) (absolute or per cent predicted for age, sex and height)
forced expiratory volume in 0.5 seconds (FEV0.5) measured by the raised volume rapid thoracic compression technique or thoracic gas volume (TGV) measured by whole body plethysmography in infants and young children
Nutritional status (weight, weight-for-age, weight-for-age Z-score, height, height-for-age, height-for-age Z-score, weight-for-height, body mass index (BMI))
Number of adverse events or number of children having adverse events
Number of acute exacerbations
Number of infections with Pseudomonas aeruginosa
Number of antibiotic courses
We accepted data from all provided time points for outcomes.
Search methods for identification of studies
No restrictions were placed on language.
We identified relevant studies from the Group's Cystic Fibrosis Trials Register using the terms: respiratory syncytial virus AND palivizumab.
The Cystic Fibrosis Trials Register is compiled from electronic searches of the Cochrane Central Register of Controlled Trials (CENTRAL) (updated each new issue of The Cochrane Library), quarterly searches of MEDLINE, a search of EMBASE to 1995 and the prospective handsearching of two journals - Pediatric Pulmonology and the Journal of Cystic Fibrosis. Unpublished work is identified by searching the abstract books of three major cystic fibrosis conferences: the International Cystic Fibrosis Conference; the European Cystic Fibrosis Conference and the North American Cystic Fibrosis Conference. For full details of all searching activities for the register, please see the relevant sections of the Cochrane Cystic Fibrosis and Genetic Disorders Group Module.
Date of most recent search: 17 March 2014.
Searching other resources
We searched the reference list of the eligible trial and existing review articles on palivizumab to identify additional relevant studies and trial reports. We contacted the drug manufacturer (MedImmune, Inc.) and authors to obtain information on ongoing or unpublished studies.
Data collection and analysis
Selection of studies
Two review authors independently screened the article identified by the search methods, first using the title and abstract and subsequently using the full-text. Disagreements regarding eligibility were resolved by consensus or by consulting with a third review author.
Data extraction and management
Two review authors independently extracted data and assessed the risk of bias in the eligible study using custom data abstraction forms designed for this review. We abstracted the information about study and participant characteristics, the intervention, and the outcomes directly into custom data abstraction forms. We corresponded with trial authors to address uncertainties in the identified study and to obtain additional information on missing data. We checked the data for accuracy and consistency, and resolved disagreements by consensus or by consulting a third reviewer. Finally, we entered the data into Review Manager software for analysis (RevMan 2011).
We accepted data from all provided time-points for outcomes.
Assessment of risk of bias in included studies
We employed the method described in The Cochrane Handbook for Systematic Reviews of Interventions to evaluate the risk of bias (Higgins 2011). We looked for adequacy of the random sequence generation; methods of concealment of treatment allocation; blinding of participants, personnel, and outcome assessors; completeness of outcome data for each main outcome; and presence of selective outcome reporting.
Measures of treatment effect
We calculated risk ratios and their associated 95% confidence intervals (CI) for each treatment group for dichotomous outcomes. We calculated odds ratios for adverse events.
No continuous outcomes were reported in the included study. If in future updates we identify studies reporting continuous outcomes, we will report the mean relative change from baseline or the mean post-intervention value as well as the difference in means between treatment groups and their associated 95% CI. We will also report the standard deviations; where standard errors are provided, we will convert these to standard deviations.
Unit of analysis issues
We have not included any cross-over trials as this study design is not appropriate in evaluating vaccines.
Dealing with missing data
We contacted the authors of the included trial to obtain missing data and to clarify uncertainties in the available abstract and poster. We requested information regarding sequence generation, masking (blinding), intention-to-treat analysis, baseline characteristics (gender, nutritional status, pulmonary function), and outcome data (nutritional status, pulmonary function, airway colonisations with Pseudomonas aeruginosa, oxygen therapy for RSV infection, acute exacerbations, antibiotic courses, and any other outcomes). We also sought clarification on the criteria used for classifying adverse events as 'any', 'related', or 'serious'.
Assessment of heterogeneity
If in future sufficient trials are included in the review, we will assess heterogeneity between study estimates using the chi-squared test (obtained from Forest plots) and the I2 statistic (Higgins 2003). We will consider a chi-squared P value of less than 0.10 indicative of heterogeneity. The chi-squared test must, however, be interpreted with caution since the test may be negative in the presence of heterogeneity, in studies with small sample sizes or where a small number of studies have been pooled together in the meta-analysis. In addition, where there are many studies, it has excessive power to detect clinically insignificant heterogeneity.
In order to quantify inconsistency across studies, we will calculate the I2 statistic. This statistic describes the percentage of total variation across studies that are due to heterogeneity rather than by chance. The values of I2 lie between 0% and 100%, and a simplified categorisation of heterogeneity that we plan to use is: heterogeneity might not be important (I2 value of up to 40%), heterogeneity may be moderate (I2 value of 30% to 60%), heterogeneity may be substantial (I2 value of 50% to 90%), and considerable heterogeneity (I2 value of 75% or above) (Deeks 2011).
Assessment of reporting biases
If sufficient trials are included in the review in the future and if meta-analysis is possible, we will assess reporting bias using funnel plots. Interpretation of funnel-plots is primarily subjective and we will take this into account. To eliminate this subjectivity, the rank correlation test is commonly used (Lau 2006). But, this test relies on the presence of a large number of studies in the analysis (at least 30). Thus, small number of studies is a hindrance in the interpretation of funnel-plots. In theory, the funnel-plot evaluates the presence of differences in study results by size of study (precision of results). Thus, between-study heterogeneity further limits the validity of conclusions made from funnel-plots. We will interpret asymmetry of funnel-plots with caution (Lau 2006).
The authors' choice of reported outcomes can be influenced by the results, potentially making published results misleading (Higgins 2011). We compared the 'Methods' section of the abstract and poster with the 'Results' section, and we also considered if an outcome commonly reported in related studies was not reported. We contacted the authors for additional data.
We entered the extracted data into RevMan 5 for data analysis using a fixed-effect model (RevMan 2011).
If sufficient trials are included in future updates, and if meta-analyses are appropriate, we will pool studies of the same design and assess effects of the interventions. We will consider performing analyses using both random-effects and fixed-effects models. We will use random-effects models if there is indication of heterogeneity between trials and we can not explain the source of heterogeneity. We will use random-effects models for analysis if the trials have small sample sizes or the number of trials is small, in which situation tests for heterogeneity may be underpowered (Higgins 2011). If there is no indication of heterogeneity after qualitative exploration of the included trials and the use of statistical tests (I2 statistic), we will conduct analyses using fixed-effect models.
We analysed dichotomous outcomes using risk ratio (RR), but used the odds ratio (OR) for the analysis of adverse events. When available in future updates of the review, we will use the difference in means for continuous outcomes. Where different scales of measurement have been used, we will calculate a standardised difference in means (SMD).
When possible, we plan to conduct separate meta-analyses for each type of control group i.e. placebo, no prophylaxis, and each individual other type of prophylaxis.
Subgroup analysis and investigation of heterogeneity
When appropriate during future review updates, we will perform subgroup analyses based on the following variables:
Gender (males and females)
Age group (0 to 2 years, 3 to 6 years, and 7 to 18 years)
Presence or absence of other risk factors for severe RSV infection (premature birth, CHD, chronic lung disease of prematurity)
Geographical setting (urban and rural)
Dose used (15 mg/kg and other doses)
Regimen used (five monthly regimen and other regimens)
Definition of outcomes (as provided in included articles)
Duration of follow-up (up to one month, one month to six months, and beyond six months)
Delta F508 zygosity (homozygous and heterozygous)
When appropriate during future review updates, we will also conduct sensitivity analyses to assess the impact of including and excluding in the meta-analysis:
1. trials with inadequate methods of random sequence generation;
2. trials with inadequate methods of allocation concealment
3. trials with inadequate methods of masking (blinding) of participants, personnel, or outcome assessors;
4. trials with incomplete outcome data;
5. trials with selective outcome reporting; and
6. unpublished trials.