One-stage individual participant data meta-analysis models for continuous and binary outcomes: Comparison of treatment coding options and estimation methods.

Published on Aug 30, 2020in Statistics in Medicine1.783
· DOI :10.1002/SIM.8555
Richard D Riley70
Estimated H-index: 70
(Keele University),
Amardeep Legha4
Estimated H-index: 4
(Keele University)
+ 5 AuthorsD Burke10
Estimated H-index: 10
(Keele University)
: A one-stage individual participant data (IPD) meta-analysis synthesizes IPD from multiple studies using a general or generalized linear mixed model. This produces summary results (eg, about treatment effect) in a single step, whilst accounting for clustering of participants within studies (via a stratified study intercept, or random study intercepts) and between-study heterogeneity (via random treatment effects). We use simulation to evaluate the performance of restricted maximum likelihood (REML) and maximum likelihood (ML) estimation of one-stage IPD meta-analysis models for synthesizing randomized trials with continuous or binary outcomes. Three key findings are identified. First, for ML or REML estimation of stratified intercept or random intercepts models, a t-distribution based approach generally improves coverage of confidence intervals for the summary treatment effect, compared with a z-based approach. Second, when using ML estimation of a one-stage model with a stratified intercept, the treatment variable should be coded using "study-specific centering" (ie, 1/0 minus the study-specific proportion of participants in the treatment group), as this reduces the bias in the between-study variance estimate (compared with 1/0 and other coding options). Third, REML estimation reduces downward bias in between-study variance estimates compared with ML estimation, and does not depend on the treatment variable coding; for binary outcomes, this requires REML estimation of the pseudo-likelihood, although this may not be stable in some situations (eg, when data are sparse). Two applied examples are used to illustrate the findings.
📖 Papers frequently viewed together
2 Citations
14 Citations
64 Citations
#1Ian R. WhiteH-Index: 103
#2Rebecca M. TurnerH-Index: 28
Last. Georgia Salanti (University of Bern)H-Index: 76
view all 4 authors...
Differences between arm-based (AB) and contrast-based (CB) models for network meta-analysis (NMA) are controversial. We compare the CB model of Lu and Ades (2006), the AB model of Hong et al(2016), and two intermediate models, using hypothetical data and a selected real data set. Differences between models arise primarily from study intercepts being fixed effects in the Lu-Ades model but random effects in the Hong model, and we identify four key difference. (1) If study intercepts are fixed effe...
17 CitationsSource
#1Dean Langan (University of Manchester)H-Index: 11
#2Julian P T Higgins (UoB: University of Bristol)H-Index: 135
Last. Mark Simmonds (Ebor: University of York)H-Index: 25
view all 8 authors...
Studies combined in a meta-analysis often have differences in their design and conduct that can lead to heterogeneous results. A random-effects model accounts for these differences in the underlying study effects, which includes a heterogeneity variance parameter. The DerSimonian-Laird method is often used to estimate the heterogeneity variance, but simulation studies have found the method can be biased and other methods are available. This paper compares the properties of nine different heterog...
126 CitationsSource
#1Amardeep Legha (Keele University)H-Index: 4
#2Richard D Riley (Keele University)H-Index: 70
Last. D Burke (Keele University)H-Index: 10
view all 6 authors...
One-stage individual participant data meta-analysis models should account for within-trial clustering, but it is currently debated how to do this. For continuous outcomes modeled using a linear regression framework, two competing approaches are a stratified intercept or a random intercept. The stratified approach involves estimating a separate intercept term for each trial, whereas the random intercept approach assumes that trial intercepts are drawn from a normal distribution. Here, through an ...
14 CitationsSource
#1Hans-Peter Piepho (University of Hohenheim)H-Index: 72
#2Laurence V. Madden (OSU: Ohio State University)H-Index: 70
Last. Emlyn Williams (ANU: Australian National University)H-Index: 27
view all 5 authors...
Network meta-analysis can be implemented by using arm-based or contrast-based models. Here we focus on arm-based models and fit them using generalized linear mixed model procedures. Full maximum likelihood (ML) estimation leads to biased trial-by-treatment interaction variance estimates for heterogeneity. Thus, our objective is to investigate alternative approaches to variance estimation that reduce bias compared with full ML. Specifically, we use penalized quasi-likelihood/pseudo-likelihood and...
7 CitationsSource
#1Dan Jackson (University of Cambridge)H-Index: 35
#2Martin Law (University of Cambridge)H-Index: 9
Last. Ian R. White (UCL: University College London)H-Index: 103
view all 5 authors...
Comparative trials that report binary outcome data are commonly pooled in systematic reviews and meta-analyses. This type of data can be presented as a series of 2-by-2 tables. The pooled odds ratio is often presented as the outcome of primary interest in the resulting meta-analysis. We examine the use of 7 models for random-effects meta-analyses that have been proposed for this purpose. The first of these models is the conventional one that uses normal within-study approximations and a 2-stage ...
74 CitationsSource
#1Tim P. MorrisH-Index: 28
#1Tim P. MorrisH-Index: 27
Last. James R. Carpenter (Lond: University of London)H-Index: 59
view all 4 authors...
Quantitative evidence synthesis through meta-analysis is central to evidence-based medicine. For well-documented reasons, the meta-analysis of individual patient data is held in higher regard than aggregate data. With access to individual patient data, the analysis is not restricted to a "two-stage" approach (combining estimates and standard errors) but can estimate parameters of interest by fitting a single model to all of the data, a so-called "one-stage" analysis. There has been debate about ...
17 CitationsSource
#1Ewelina Rogozińska (QMUL: Queen Mary University of London)H-Index: 17
#2Nadine Marlin (QMUL: Queen Mary University of London)H-Index: 12
Last. Shakila Thangaratinam (QMUL: Queen Mary University of London)H-Index: 44
view all 57 authors...
Background: Diet- and physical activity-based interventions in pregnancy have the potential to alter maternal and child outcomes. Objectives: To assess whether or not the effects of diet and lifestyle interventions vary in subgroups of women, based on maternal body mass index (BMI), age, parity, Caucasian ethnicity and underlying medical condition(s), by undertaking an individual patient data (IPD) meta-analysis. We also evaluated the association of gestational weight gain (GWG) with adverse pre...
147 CitationsSource
#1D Burke (Keele University)H-Index: 10
#2Joie Ensor (Keele University)H-Index: 17
Last. Richard D Riley (Keele University)H-Index: 70
view all 3 authors...
Meta-analysis using individual participant data (IPD) obtains and synthesises the raw, participant-level data from a set of relevant studies. The IPD approach is becoming an increasingly popular tool as an alternative to traditional aggregate data meta-analysis, especially as it avoids reliance on published results and provides an opportunity to investigate individual-level interactions, such as treatment-effect modifiers. There are two statistical approaches for conducting an IPD meta-analysis:...
187 CitationsSource
#1Doneal Thomas (McGill University)H-Index: 4
#2Robert W. Platt (McGill University)H-Index: 100
Last. Andrea Benedetti (McGill University)H-Index: 53
view all 3 authors...
Background Individual patient data meta-analyses (IPD-MA) are often performed using a one-stage approach-- a form of generalized linear mixed model (GLMM) for binary outcomes. We compare (i) one-stage to two-stage approaches (ii) the performance of two estimation procedures (Penalized Quasi-likelihood-PQL and Adaptive Gaussian Hermite Quadrature-AGHQ) for GLMMs with binary outcomes within the one-stage approach and (iii) using stratified study-effect or random study-effects.
10 CitationsSource
#1Mark Simmonds (Ebor: University of York)H-Index: 25
#2Julian P T Higgins (UoB: University of Bristol)H-Index: 135
Where individual participant data are available for every randomised trial in a meta-analysis of dichotomous event outcomes, “one-stage” random-effects logistic regression models have been proposed as a way to analyse these data. Such models can also be used even when individual participant data are not available and we have only summary contingency table data. One benefit of this one-stage regression model over conventional meta-analysis methods is that it maximises the correct binomial likelih...
52 CitationsSource
Cited By4
#1James Baxter (Edin.: University of Edinburgh)
#2Sarah Langhorne (Lond: University of London)
Last. Katherine E. Atkins (Edin.: University of Edinburgh)H-Index: 27
view all 9 authors...
Background HIV-1 null infections that are initiated by multiple founder variants are characterised by a higher viral load and a worse clinical prognosis, yet little is known about the routes of exposure through which multiple variant transmission is most likely, and whether methods of quantifying the number of founder variants differ in their accuracy. Methods We conducted a systematic review of studies that estimated founder variant multiplicity in HIV-1 infection, searching MEDLINE, EMBASE and...
#1Dyuti Coomar (University of Birmingham)H-Index: 3
#2Jonathan Hazlehurst (University of Birmingham)H-Index: 14
Last. Shakila Thangaratinam (QMUL: Queen Mary University of London)H-Index: 44
view all 17 authors...
Introduction null Mothers with gestational diabetes mellitus (GDM) are at increased risk of pregnancy-related complications and developing type 2 diabetes after delivery. Diet and physical activity-based interventions may prevent GDM, but variations in populations, interventions and outcomes in primary trials have limited the translation of available evidence into practice. We plan to undertake an individual participant data (IPD) meta-analysis of randomised trials to assess the differential eff...
#1Jennie Louise (University of Adelaide)H-Index: 11
#2Amanda Poprzeczny (University of Adelaide)H-Index: 3
Last. Jodie M Dodd (University of Adelaide)H-Index: 50
view all 16 authors...
BACKGROUND The impact of maternal obesity extends beyond birth, being independently associated with an increased risk of child obesity. Current evidence demonstrates that women provided with a dietary intervention during pregnancy improve their dietary quality and have a modest reduction in gestational weight gain. However, the effect of this on longer-term childhood obesity-related outcomes is unknown. METHODS We conducted an individual participant data meta-analysis from RCTs in which women wi...
#1Alexandra Wright-Hughes (University of Leeds)H-Index: 8
#2Rebecca Walwyn (University of Leeds)H-Index: 18
Last. David Cottrell (University of Leeds)H-Index: 31
view all 8 authors...
Introduction Up to 10% of adolescents report self-harm in the previous year. Non-fatal repetition is common (18% in 1 year), death from any cause shows a fourfold and suicide a 10-fold excess. Despite the scale of the problem, there is insufficient evidence for effective interventions for self-harm. Those who self-harm do so for a variety of different reasons. Different treatments may be more effective for subgroups of adolescents; however, little is known about which subgroups are appropriate f...