In the absence of sufficient data directly comparing multiple treatments, indirect comparisons using network meta-analyses (NMAs) can provide useful information. Under current contrast-based (CB) methods for binary outcomes, the patient-centered measures including the treatment-specific event rates and risk differences (RDs) are not provided, which may create some unnecessary obstacles for patients to comprehensively trade-off efficacy and safety measures.
Purpose
We aim to develop NMA to accurately estimate the treatment-specific event rates.
Methods
A Bayesian hierarchical model is developed to illustrate how treatment-specific event rates, RDs, and risk ratios (RRs) can be estimated. We first compare our approach to alternative methods using two hypothetical NMAs assuming a fixed RR or RD, and then use two published NMAs to illustrate the improved reporting.
Results
In the hypothetical NMAs, our approach outperforms current CB NMA methods in terms of bias. In the two published NMAs, noticeable differences are observed in the magnitude of relative treatment effects and several pairwise statistical significance tests from previous report.
Limitations
First, to facilitate the estimation, each study is assumed to hypothetically compare all treatments, with unstudied arms being missing at random. It is plausible that investigators may have selected treatment arms on purpose based on the results of previous trials, which may lead to ‘nonignorable missingness’ and potentially bias our estimates. Second, we have not considered methods to identify and account for potential inconsistency between direct and indirect comparisons.
Conclusions
The proposed NMA method can accurately estimate treatment-specific event rates, RDs, and RRs and is recommended.
Get full access to this article
View all access options for this article.
References
1.
EggerMSmithGDAltmanD. Systematic Reviews in Health Care: Meta-Analysis in Context (2nd edn). BMJ Publishing Group, London, 2001.
2.
GuyattGHSackettDLSinclairJC. Users’ guides to the medical literature. JAMA1995; 274(22): 1800–04.
3.
LiTPuhanMAVedulaSSSinghSDickersinK. Network meta-analysis-highly attractive but more methodological research is needed. BMC Med2011; 9: 79.
4.
HigginsJPTWhiteheadA. Borrowing strength from external trials in a meta-analysis. Stat Med1996; 15(24): 2733–49.
5.
CaldwellDMAdesAHigginsJ. Simultaneous comparison of multiple treatments: Combining direct and indirect evidence. BMJ2005; 331: 897–900.
6.
CiprianiAFurukawaTASalantiG. Comparative efficacy and acceptability of 12 new-generation antidepressants: A multiple-treatments meta-analysis. Lancet2009; 373: 746–58.
7.
CiprianiABarbuiCSalantiG. Comparative efficacy and acceptability of antimanic drugs in acute mania: A multiple-treatments meta-analysis. Lancet2011; 378: 1306–15.
8.
ElliottWJMeyerPM. Incident diabetes in clinical trials of antihypertensive drugs: A network meta-analysis. Lancet2007; 369(9557): 201–07.
9.
PahorMPsatyBMAldermanMH. Health outcomes associated with calcium antagonists compared with other first-line antihypertensive therapies: A meta-analysis of randomised controlled trials. Lancet2000; 356(9246): 1949–54.
10.
SongFLokeYKWalshT. Methodological problems in the use of indirect comparisons for evaluating healthcare interventions: Survey of published systematic reviews. BMJ2009; 338: b1147.
11.
PalmeriniTBiondi-ZoccaiGRivaDD. Stent thrombosis with drug-eluting and bare-metal stents: Evidence from a comprehensive network meta-analysis. Lancet2012; 379: 1393–402.
12.
DanielsJMiddletonLChampaneriaR. Second generation endometrial ablation techniques for heavy menstrual bleeding: Network meta-analysis. BMJ2012; 344: e2564.
13.
WangSYChuHShamliyanT. Network meta-analysis of margin threshold for women with ductal carcinoma in situ. J Natl Cancer Inst2012; 104(7): 507–16.
14.
AltmanDGDeeksJJSackettDL. Odds ratios should be avoided when events are common. BMJ1998; 317: 1318.
15.
DeeksJ. When can odds ratios mislead? Odds ratios should be used only in case-control studies and logistic regression analyses. BMJ1998; 317: 1155–57.
16.
SackettDLDeeksJJAltmanDG. Down with odds ratios!Evid Based Med1996; 1: 164–66.
17.
DaviesHTOCrombieIKTavakoliM. When can odds ratios mislead?BMJ1998; 316: 989–91.
18.
BorensteinMHedgesLVHigginsJPTRothsteinHR. Introduction to Meta-Analysis. Wiley, West Sussex, 2011.
19.
PsatyBMLumleyTFurbergCD. Health outcomes associated with various antihypertensive therapies used as first-line agents. JAMA2003; 289(19): 2534–44.
20.
TrikalinosTAAlsheikh-AliAATatsioniANallamothuBKKentDM. Percutaneous coronary interventions for non-acute coronary artery disease: A quantitative 20-year synopsis and a network meta-analysis. Lancet2009; 373(9667): 911–18.
21.
LumleyT. Network meta-analysis for indirect treatment comparisons. Stat Med2002; 21(16): 2313–24.
22.
SalantiGAdesAEIoannidisJPA. Graphical methods and numerical summaries for presenting results from multiple-treatment meta-analysis: An overview and tutorial. J Clin Epidemiol2011; 64(2): 163–71.
23.
ChungHLumleyT. Graphical exploration of network meta-analysis data: The use of multidimensional scaling. Clin Trials2008; 5(4): 301–07.
24.
LuGAdesAE. Combination of direct and indirect evidence in mixed treatment comparisons. Stat Med2004; 23: 3105–24.
25.
LuGAdesAE. Assessing evidence inconsistency in mixed treatment comparisons. J Am Stat Assoc2006; 101: 447–59.
26.
LuGAdesAESuttonAJ. Meta-analysis of mixed treatment comparisons at multiple follow-up times. Stat Med2007; 26: 3681–99.
JonesBRogerJLanePW. Statistical approaches for conducting network meta-analysis in drug development. Pharm Stat2011; 10(6): 523–31.
29.
WhiteIR. Multivariate random-effects meta-regression: Updates to mvmeta. Stata J2011; 11(2): 255–70.
30.
ManzoliLVitoCDSalantiG. Meta-analysis of the immunogenicity and tolerability of pandemic influenza A 2009 (H1N1) vaccines. PloS One2011; 6(9): e24384.
31.
DiasSWeltonNJSuttonAJAdesAE. Nice DSU technical support document 2: A generalised linear modelling framework for pairwise and network meta-analysis of randomised controlled trials. Report, National Institute for Health and Clinical Excellence, London, UK, August 2011.
32.
DiasSWeltonNJSuttonAJAdesAE. Nice DSU technical support document 5: Evidence synthesis in the baseline natural history model. Report, National Institute for Health and Clinical Excellence, London, UK, August 2012.
33.
DiasSWeltonNJSuttonAJAdesAE. Nice DSU technical support document 6: Embedding evidence synthesis in probabilistic cost-effectiveness analysis: Software choices. Report, National Institute for Health and Clinical Excellence, London, UK, May 2011.
34.
IbrahimJGChuHChenMH. Missing data in clinical studies: Issues and methods. J Clin Oncol2012. 30(26): 3297–303.
35.
LittleRJARubinDB. Statistical Analysis With Missing Data (2nd edn). John Wiley & Sons, New Jersey, 2002.
36.
RubinDB. Inference and missing data. Biometrika1976; 63(3): 581–92.
37.
RubinDB. Multiple Imputation for Nonresponse in Surveys. Wiley Online Library, New York, 1987.
38.
SchaferJL. Analysis of Incomplete Multivariate Data. Chapman & Hall/CRC, New York, 1997.
39.
SalantiGHigginsJPTAdesAEIoannidisJPA. Evaluation of networks of randomized trials. Stat Methods Med Res2008; 17(3): 279–301.
40.
WhiteIRBarrettJKJacksonDHigginsJPT. Consistency and inconsistency in network meta-analysis: Model estimation using multivariate meta-regression. Res Synth Methods2012; 3(2): 111–25.
41.
ChuHNieLChenYHuangYSunW. Bivariate random effects models for meta-analysis of comparative studies with binary outcomes: Methods for the absolute risk difference and relative risk. Stat Methods Med Res2012; 21(6): 621–33.
42.
GelmanA. Prior distributions for variance parameters in hierarchical models. Bayesian Anal2006; 1(3): 515–33.
43.
GustafsonP. The utility of prior information and stratification for parameter estimation with two screening tests but no gold standard. Stat Med2005; 24(8): 1203–17.
44.
GustafsonPHossainSMacnabYC. Conservative prior distributions for variance parameters in hierarchical models. Can J Stat2006; 34(4): 377–90.
45.
NatarajanRMcCullochCE. Gibbs sampling with diffuse proper priors: A valid approach to data-driven inference?J Comput Graph Stat1998; 7(3): 267–77.
46.
LunnDJThomasABestNSpiegelhalterD. A Bayesian modeling framework: Concepts, structure, and extensibility. Stat Comput2000; 10(4): 325–37.
47.
LunnDSpiegelhalterDThomasABestN. The BUGS project: Evolution, critique and future directions. Stat Med2009; 28(25): 3049–67.
48.
GelfandAESmithAFM. Sampling-based approaches to calculating marginal densities. J Am Stat Assoc1990; 85(410): 398–409.
BrooksSPGelmanA. General methods for monitoring convergence of iterative simulations. J Comput Graph Stat1998; 7(4): 434–55.
51.
GelmanARubinDB. Inference from iterative simulation using multiple sequences. Stat Sci1992; 7(4): 457–72.
52.
GoodmanSN. Toward evidence-based medical statistics. 1: The P value fallacy. Ann Intern Med1999; 130(12): 995–1004.
53.
YusufSPetoRLewisJCollinsRSleightP. Beta blockade during and after myocardial infarction: An overview of the randomized trials. Prog Cardiovasc Dis1985; 27(5): 335–71.
54.
HongHChuHZhangJCarlinBP. A Bayesian missing data framework for generalized multiple outcome mixed treatment comparisons. Research Report 2012-018, 2012. Minneapolis, MN: Division of Biostatistics, University of Minnesota.
55.
ZhangJYuKF. What’s the relative risk?JAMA1998; 280(19): 1690–91.
56.
ZegerSLLiangKYAlbertPS. Models for longitudinal data: A generalized estimating equation approach. Biometrics1988; 44(4): 1049–60.
57.
CaldwellDMWeltonNJDiasSAdesAE. Selecting the best scale for measuring treatment effect in a network meta-analysis: A case study in childhood nocturnal enuresis. Res Synth Methods2012; 3(2): 126–41.
58.
AchanaFACooperNJDiasS. Extending methods for investigating the relationship between treatment effect and baseline risk from pairwise meta-analysis to network meta-analysis. Stat Med2012; 32(5): 752–71.
59.
DiasSWeltonNJCaldwellDMAdesAE. Checking consistency in mixed treatment comparison meta-analysis. Stat Med2010; 29: 932–44.
60.
HigginsJPTJacksonDBarrettJK. Consistency and inconsistency in network meta-analysis: Concepts and models for multi-arm studies. Res Synth Methods2012; 3: 98–110.
61.
van ValkenhoefGTervonenTde BrockBHillegeH. Algorithmic parameterization of mixed treatment comparisons. Stat Comput2012; 22: 1099–111.
62.
JansenJP. Network meta-analysis of individual and aggregate level data. Res Synth Methods2012; 3: 177–90.
63.
SaramagoPSuttonAJCooperNJMancaA. Mixed treatment comparisons using aggregate and individual participant level data. Stat Med2012; 31: 3516–36.
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
