Methodological quality of multiple treatments or network meta-analyses (MTAs) in type-2 diabetes mellitus (T2DM)




Poster session 2 Thursday: Evidence synthesis - methods / improving conduct and reporting


Thursday 14 September 2017 - 12:30 to 14:00


All authors in correct order:

Ooi CP1, Ho J2
1 Universiti Putra Malaysia, Malaysia
2 Penang Medical College, Malaysia
Presenting author and contact person

Presenting author:

Cheow Peng Ooi

Contact person:

Abstract text
Background: Traditional meta-analysis is limited to the direct comparison of two treatments. MTA has the advantage of being able to compare multiple treatments both by direct and indirect comparison. However, it is more methodologically complex, and hence more vulnerable to methodological risk compared to conventional pairwise meta-analysis. Lack of methodological rigour limiting the trustworthiness of the conclusion may mislead the evidence users.
Objective:To assess the quality of evidence available for MTAs in the interventions for T2DM.
Methods: We searched CENTRAL, MEDLINE and other resources for publications of MTAs in interventions for T2DM. We used ‘A Measurement Tool to Assess Systematic Reviews’ (AMSTAR) to assess the methodological quality of published MTA on interventions for T2DM. Each of the 11 AMSTAR items was summarised using descriptive statistics.
Results: We included 48 heterogeneous MTAs from 23 journals and one health provider published from 2011 to February 2017. A total of 2246 primary studies and 1 010 582 participants with T2DM were included with 12.5% data missing. 92% included pharmacological treatments and 8% were updates of existing meta-analyses (MAs) or MTAs. Only one article performed well for all AMSTAR items. The poor performing items included providing lists of studies (10%), appropriate conclusion in respect to quality of included studies (16%), assessment of publication bias (16%) and presence of pre-planned protocol (18%) (Table 1). Better performing items included reporting unpublished data (45%), number of data extractors (53%), conflict of interest (59%) and assessment of quality of primary study (61%). Detailed description of included studies (71%) and methodologies of MA and MTA (80%) were the best-performing items.
Conclusion: These methodological shortcomings may threaten the validity of conclusions and limit the use of these MTAs for clinical decision. We suggest assessment of methodological quality of MTA using validated tools, such as AMSTAR, complementing the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) Network Meta-analysis Extension statement for improving reporting quality.