Bias in the one‐step method for pooling study results

Greenland, Sander; Salvan, Alberto

doi:10.1002/sim.4780090307

Cited by 148 publications

(66 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Colbourn et al 16 (p.45) argue that the Smail estimate is based on the Peto 'one-step' method, which is a fixed effect estimator that they assert produces '…seriously biased estimates when the true treatment effects are large'. 38,39 Our sensitivity analysis showed that when the alternative estimate used by Colbourn et al 16 was used there was a significant impact on the results, which made the relative ICERs, when compared with 'Do nothing', even more favourable.…”

Section: Discussionmentioning

confidence: 86%

Cost‐effectiveness of rapid tests and other existing strategies for screening and management of early‐onset group B streptococcus during labour

Kaambwa

Bryan

Gray³

et al. 2010

BJOG

View full text Add to dashboard Cite

Objective To determine the cost-effectiveness of alternative screening and prevention strategies, including rapid intrapartum testing, for prevention of early-onset neonatal group B streptococcus (GBS) infection in the UK. Design A decision model was developed to investigate the costeffectiveness of screening and prevention strategies for GBS. A strategy of doing nothing was also considered. Deterministic and probabilistic sensitivity analyses were carried out.Setting Two large UK based obstetric units.Participants Test accuracy data were obtained from a primary study of rapid tests at the onset of labour and risk factors from 1400 women.Main outcome measures Incremental health sector costs per case of early-onset GBS death avoided.Results Compared with a strategy of do nothing, the incremental cost-effectiveness ratio was £32,000 per Early-Onset GBS Disease avoided and £427,000 per Early-Onset GBS Death avoided for the strategy of providing routine intrapartum antibiotic prophylaxis to all women without prior screening; Based on their current sensitivity, specificity and cost, screening using rapid tests was dominated by other more cost-effective strategies.Conclusions The most cost-effective strategy was shown to be the provision of routine intrapartum antibiotic prophylaxis to all women without prior screening but, given broader concerns relating to antibiotic use, this is unlikely to be acceptable. In its absence, intrapartum antibiotic prophylaxis directed by screening with enriched culture becomes cost-effective. The current strategy of risk-factor-based screening is not cost-effective compared with screening based on culture.Keywords Cost-effectiveness, group B streptococcus, labour, rapid tests.Please cite this paper as: Kaambwa B, Bryan S, Gray J, Milner P, Daniels J, Khan K, Roberts T. Cost-effectiveness of rapid tests and other existing strategies for screening and management of early-onset group B streptococcus during labour.

show abstract

Section: Discussionmentioning

confidence: 86%

Cost‐effectiveness of rapid tests and other existing strategies for screening and management of early‐onset group B streptococcus during labour

Kaambwa

Bryan

Gray³

et al. 2010

BJOG

View full text Add to dashboard Cite

show abstract

“…The method provides a fixed-effects model estimate of the overall odds ratio. In sparse data situations and under certain conditions, the method has been shown to produce the least biased results with the most accurate confidence interval coverages (Bradburn et al 2007), but can also be quite biased in other situations (Greenland and Salvan 1990).…”

Section: Discussionmentioning

confidence: 99%

Conducting Meta-Analyses inRwith themetaforPackage

Viechtbauer¹

2010

J. Stat. Soft.

14,418

11,303

View full text Add to dashboard Cite

The metafor package provides functions for conducting meta-analyses in R. The package includes functions for fitting the meta-analytic fixed-and random-effects models and allows for the inclusion of moderators variables (study-level covariates) in these models. Meta-regression analyses with continuous and categorical moderators can be conducted in this way. Functions for the Mantel-Haenszel and Peto's one-step method for metaanalyses of 2 × 2 table data are also available. Finally, the package provides various plot functions (for example, for forest, funnel, and radial plots) and functions for assessing the model fit, for obtaining case diagnostics, and for tests of publication bias.

show abstract

“…To address this, alternative two‐stage approaches have been proposed, which use a different weighting scheme to the inverse variance method, such as the Peto method 65 and the Mantel–Haenszel method 66, which also avoid the use of continuity corrections when there are zero cells. The Peto method is considered to work well when intervention effects are small, event risks are <1%, and when there are balanced experimental and control group sizes within trials 10, 60.…”

Section: Key Reasons Why Meta‐analysis Results May Differ For the Onementioning

confidence: 99%

Meta-analysis using individual participant data: one-stage and two-stage approaches, and why they may differ

2016

View full text Add to dashboard Cite

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: one‐stage and two‐stage. The one‐stage approach analyses the IPD from all studies simultaneously, for example, in a hierarchical regression model with random effects. The two‐stage approach derives aggregate data (such as effect estimates) in each study separately and then combines these in a traditional meta‐analysis model. There have been numerous comparisons of the one‐stage and two‐stage approaches via theoretical consideration, simulation and empirical examples, yet there remains confusion regarding when each approach should be adopted, and indeed why they may differ.In this tutorial paper, we outline the key statistical methods for one‐stage and two‐stage IPD meta‐analyses, and provide 10 key reasons why they may produce different summary results. We explain that most differences arise because of different modelling assumptions, rather than the choice of one‐stage or two‐stage itself. We illustrate the concepts with recently published IPD meta‐analyses, summarise key statistical software and provide recommendations for future IPD meta‐analyses. © 2016 The Authors. Statistics in Medicine published by John Wiley & Sons Ltd.

show abstract

Bias in the one‐step method for pooling study results

Cited by 148 publications

References 10 publications

Cost‐effectiveness of rapid tests and other existing strategies for screening and management of early‐onset group B streptococcus during labour

Cost‐effectiveness of rapid tests and other existing strategies for screening and management of early‐onset group B streptococcus during labour

Conducting Meta-Analyses inRwith themetaforPackage

Meta-analysis using individual participant data: one-stage and two-stage approaches, and why they may differ

Contact Info

Product

Resources

About