|
Presentation of Evidence and its Characteristics in the Treatment Network
|
| • Presentation of network diagram to summarize identified evidence |
| • Reporting information reflecting the amount of information in the network, e.g. sample sizes,numbers of studies per comparison and the presence of multi-arm studies |
| • Presentation of information allowing readers to assess clinical and methodological heterogeneitywithin the treatment network: e.g. information tables listing effect modifiers across studies andcomparisons, These can include patient characteristics and risk of bias assessements |
|
Assessment of Assumptions for Network Meta-Analysis
|
| • Information to summarize evaluations of statistical heterogeneity within the treatment network |
| • Information and approach to summarize analyses to assess agreement of direct and indirectsources of evidence (and efforts to improve agreement if discrepancies are found) |
|
Presentation of Summary Treatment Effect Estimates and Related Measures
|
| • What estimates to report: all possible pairwise comparisons? Only those which are comparisonsagainst the chosen reference group or a treatment of primary focus? |
| • Should findings from traditional pairwise analyses also be provided? |
| • Presentation of summary estimates and corresponding uncertainty (i.e. credible/confidence intervals) |
| • Presentation of summary estimates from sensitivity and subgroup analyses |
| • Optimal use of tables and figures to most easily convey results to readers |
| • Presentation of treatment rankings and corresponding probabilities: Should they be included?What should be presented? |
|
Discussion and Conclusions
|
| • Commentary on the clinical and biologic plausibility of the observed findings |
| • Commentary relevant to any important concerns regarding the assumptions underlying thejoint synthesis that may play an important role in strength of interpretations drawn |
