Results from 3 trials examining the mortality risk of three treatments and placebo in patients with chronic obstructive pulmonary disease.

dat.woods2010

Format

The data frame contains the following columns:

authorcharacterfirst author / study name
treatmentcharactertreatment
rintegernumber of deaths
Nintegernumber of patients

Details

Count mortality statistics in randomised controlled trials of treatments for chronic obstructive pulmonary disease (Woods et al., 2010, Table 1).

Source

Woods, B. S., Hawkins, N., & Scott, D. A. (2010). Network meta-analysis on the log-hazard scale, combining count and hazard ratio statistics accounting for multi-arm trials: A tutorial. BMC Medical Research Methodology, 10, 54. https://doi.org/10.1186/1471-2288-10-54

Concepts

medicine, odds ratios, network meta-analysis

Examples

### Show full data set
dat.woods2010
#>          author   treatment r   N
#> 1     Boyd 1997  Salmeterol 1 229
#> 2     Boyd 1997     Placebo 1 227
#> 3 Calverly 2003 Fluticasone 4 374
#> 4 Calverly 2003  Salmeterol 3 372
#> 5 Calverly 2003         SFC 2 358
#> 6 Calverly 2003     Placebo 7 361
#> 7    Celli 2003  Salmeterol 1 554
#> 8    Celli 2003     Placebo 2 270

# \dontrun{

### Load netmeta package
suppressPackageStartupMessages(library(netmeta))

### Print odds ratios and confidence limits with two digits
settings.meta(digits = 2)

### Change appearance of confidence intervals
cilayout("(", "-")

### Transform data from long arm-based format to contrast-based
### format. Argument 'sm' has to be used for odds ratio as summary
### measure; by default the risk ratio is used in the metabin function
### called internally.
pw <- pairwise(treatment, event = r, n = N,
  studlab = author, data = dat.woods2010, sm = "OR")
pw
#>         studlab      treat1     treat2          TE      seTE event1  n1 event2  n2 incr allstudies
#> 1     Boyd 1997  Salmeterol    Placebo -0.00881063 1.4173252      1 229      1 227    0      FALSE
#> 2 Calverly 2003 Fluticasone    Placebo -0.60382188 0.6311772      4 374      7 361    0      FALSE
#> 3 Calverly 2003 Fluticasone        SFC  0.65457491 0.8692018      4 374      2 358    0      FALSE
#> 4 Calverly 2003 Fluticasone Salmeterol  0.28497571 0.7672979      4 374      3 372    0      FALSE
#> 5 Calverly 2003  Salmeterol    Placebo -0.88879759 0.6940644      3 372      7 361    0      FALSE
#> 6 Calverly 2003  Salmeterol        SFC  0.36959919 0.9158888      3 372      2 358    0      FALSE
#> 7 Calverly 2003         SFC    Placebo -1.25839679 0.8052894      2 358      7 361    0      FALSE
#> 8    Celli 2003  Salmeterol    Placebo -1.41751820 1.2270043      1 554      2 270    0      FALSE
#>          author  treatment1 treatment2 r1 r2  N1  N2
#> 1     Boyd 1997  Salmeterol    Placebo  1  1 229 227
#> 2 Calverly 2003 Fluticasone    Placebo  4  7 374 361
#> 3 Calverly 2003 Fluticasone        SFC  4  2 374 358
#> 4 Calverly 2003 Fluticasone Salmeterol  4  3 374 372
#> 5 Calverly 2003  Salmeterol    Placebo  3  7 372 361
#> 6 Calverly 2003  Salmeterol        SFC  3  2 372 358
#> 7 Calverly 2003         SFC    Placebo  2  7 358 361
#> 8    Celli 2003  Salmeterol    Placebo  1  2 554 270

### Conduct network meta-analysis
net <- netmeta(pw)
net
#> Number of studies: k = 3
#> Number of pairwise comparisons: m = 8
#> Number of observations: o = 2745
#> Number of treatments: n = 4
#> Number of designs: d = 2
#> 
#> Common effects model
#> 
#> Treatment estimate (sm = 'OR', comparison: other treatments vs 'Fluticasone'):
#>               OR      95%-CI     z p-value
#> Fluticasone    .           .     .       .
#> Placebo     1.81 (0.54-6.10)  0.96  0.3355
#> SFC         0.52 (0.09-2.85) -0.75  0.4514
#> Salmeterol  0.77 (0.19-3.08) -0.37  0.7078
#> 
#> Random effects model
#> 
#> Treatment estimate (sm = 'OR', comparison: other treatments vs 'Fluticasone'):
#>               OR      95%-CI     z p-value
#> Fluticasone    .           .     .       .
#> Placebo     1.81 (0.54-6.10)  0.96  0.3355
#> SFC         0.52 (0.09-2.85) -0.75  0.4514
#> Salmeterol  0.77 (0.19-3.08) -0.37  0.7078
#> 
#> Quantifying heterogeneity / inconsistency:
#> tau^2 = 0; tau = 0; I^2 = 0% (0.0%-89.6%)
#> 
#> Tests of heterogeneity (within designs) and inconsistency (between designs):
#>                    Q d.f. p-value
#> Total           0.57    2  0.7525
#> Within designs  0.56    1  0.4524
#> Between designs 0.00    1  0.9485

### Show forest plot
forest(net, ref = "Placebo", drop = TRUE,
  leftlabs = "Contrast to Placebo")


# }