Data Extraction and Analyses
Data extraction from RCTs was focused on the outcome
frequency in the control group (control event rate, CER%)
and in the experiment group (experiment event rate,
EER%)1 and sample size in the control group and in the
test group(s). For parallel group RCTs, the frequency of
pneumonia, or lower respiratory tract infection, in the control
group (CER%) and in the experiment group (EER%),
were used to calculate the absolute risk reductions (ARRs;
ARR5CER%EER%), and numbers needed to treat
(NNTs) were calculated with 95% confidence intervals
(CIs) using standard formulas (95% CI5ARR 1.96
standard error).12,13 The NNT values (95% CIs for NNTs)
were obtained as reciprocals of ARR (NNT5100/ARR), as
previously described.12 To ensure the consistency of the
assessments throughout the study, two authors (EN, PS)
performed the data extraction independently, and any
disagreements were solved in consensus meetings. Pooling
data from individual studies (meta-analysis) was not
deemed suitable because of heterogeneous study designs,
quality of reporting methodological aspects, and trial
conduct.
To provide an overview of additional clinical studies in
this research area, the non-RCT studies that were identified
in the literature were scrutinized for the authors’ main conclusion(
s).