Network Meta-Analysis Methods: Number Needed to Treat Approaches

    The number needed to treat in pairwise and network meta-analysis and its graphical representation

    Veroniki AA, Bender R, Glasziou P, Straus SE, Tricco AC. The number needed to treat in pairwise and network meta-analysis and its graphical representation. J Clin Epidemiol. 2019 Jul;111:11-22. doi: 10.1016/j.jclinepi.2019.03.007.


    [sidebar] [no-bottom-margin]


    Share Buttons [span][/span]

    [ssba-buttons] [/sidebar] [sidebar-content]
    • A NNT may be presented in a bar plot, Cates plot or forest plot for a single outcome, and a bubble plot, scatterplot or rank-heat plot for ≥2 outcomes
    • The Cates plot, bubble plot, and rank-heat plot do not depict the NNT uncertainty, which can impact interpretation. If uncertainty around NNT should be considered in decision-making, then a bar plot or a forest plot can be used
    • The Cates plot is suggested if the corresponding effect estimate is statistically significant and the confidence interval is not too wide
    • Caution is needed in NNT interpretation, as considerations such as selection of effect size and CER, and CER assumption across multiple comparisons, may impact NNT and decision-making. The graphs we propose are helpful to interpret NNTs calculated from (network) meta-analyses
    • The number needed to treat (NNT) is an absolute measure of effect used to communicate the effectiveness or safety of an intervention, and is relatively easy to understand
    • It specifies the average number of patients who need to be treated to achieve treatment success in one patient compared to placebo or another treatment
    • The NNT calculation using a relative effect measure, such as odds ratio and risk ratio, requires a control event rate (CER). However, different CERs, such as mean CER across studies, pooled CER in a meta-analysis, expert opinion-based CER, and a range of possible CER values may impact the magnitude and direction of the NNT
    • The graphical representation of NNTs from NMA is crucial to ease interpretation of results. We present six graphical approaches for NNT from NMA and discuss their properties



    Category :

    Date : 14 Sep 2020