Assessment of reporting biases
Funnel plots will be constructed if at least 10 studies are available
for the meta-analysis of a primary outcome.
Data synthesis
We will undertake analysis of the data using Cochrane RevMan
5.2 software (RevMan 2011).Wewill present the results with 95%
confidence intervals (95% CI) and report estimates for dichotomous
outcomes as RR, and continuous outcomes asMDor SMD.
Data synthesis will depend on the quality, design and heterogeneity
of the included trials. If data are not appropriate for pooling
or analysis, we will undertake a narrative overview, which will be
structured according to the study design that has been used.
Subgroup analysis and investigation of heterogeneity
If sufficient data are available and significant heterogeneity is identified,
we will undertake the following subgroup analysis: type of
setting (community, hospital, inpatient, outpatient). If substantial
statistical or clinical heterogeneity is detected either by the I2
statistic or by observation, and cannot be explained, we will not
combine study results but will present them in a descriptive summary.