Assessing the impact of case-mix heterogeneity in individual participant data meta-analysis: Novel use of I2 statistic and prediction interval

Case mix differences between trials form an important factor that contributes to the statistical heterogeneity observed in a meta-analysis. In this paper, we propose two methods to assess whether important heterogeneity would remain if the different trials in the meta-analysis were conducted in one common population defined by a given case-mix. To achieve this goal, we first standardize results of different trials over the case-mix of a target population. We then quantify the amount of heterogeneity arising from case-mix and beyond case-mix reasons by using corresponding I2 statistics and prediction intervals. These new approaches enable a better understanding of the overall heterogeneity between trial results, and can be used to support standard heterogeneity assessments in individual participant data meta-analysis practice.

[1]  J. Higgins,et al.  Cochrane Handbook for Systematic Reviews of Interventions , 2010, International Coaching Psychology Review.

[2]  R. Porcher,et al.  A novel approach for identifying and addressing case‐mix heterogeneity in individual participant data meta‐analysis , 2019, Research synthesis methods.

[3]  R. Porcher,et al.  Rethinking meta-analysis: assessing case-mix heterogeneity when combining treatment effects across patient populations , 2019 .

[4]  Zhiqiang Tan,et al.  Improved Estimation of Average Treatment Effects on the Treated: Local Efficiency, Double Robustness, and Beyond , 2018, 1808.01408.

[5]  T. Furukawa,et al.  Prediction intervals for random-effects meta-analysis: A confidence distribution approach , 2018, Statistical methods in medical research.

[6]  Jessica Gurevitch,et al.  Meta-analysis and the science of research synthesis , 2018, Nature.

[7]  Elias Bareinboim,et al.  Causal inference and the data-fusion problem , 2016, Proceedings of the National Academy of Sciences.

[8]  Jelle J Goeman,et al.  Plea for routinely presenting prediction intervals in meta-analysis , 2016, BMJ Open.

[9]  J. R. Carpenter,et al.  Multiple imputation for IPD meta‐analysis: allowing for heterogeneity and studies with missing covariates , 2015, Statistics in medicine.

[10]  Paul T. von Hippel,et al.  The heterogeneity statistic I 2 can be biased in small meta-analyses , 2015 .

[11]  P. V. von Hippel,et al.  The heterogeneity statistic I2 can be biased in small meta-analyses , 2015, BMC Medical Research Methodology.

[12]  Qingyuan Zhao,et al.  Entropy Balancing is Doubly Robust , 2015, 1501.03571.

[13]  Elias Bareinboim,et al.  External Validity: From Do-Calculus to Transportability Across Populations , 2014, Probabilistic and Causal Inference.

[14]  Paul T. von Hippel,et al.  Estimates of heterogeneity (I2) can be biased in small meta-analyses , 2014, 1410.2296.

[15]  Matthieu Resche-Rigon,et al.  Multiple imputation for handling systematically missing confounders in meta‐analysis of individual participant data , 2013, Statistics in medicine.

[16]  Elias Bareinboim,et al.  Meta-Transportability of Causal Effects: A Formal Approach , 2013, AISTATS.

[17]  Alex J. Sutton,et al.  Heterogeneity: Subgroups, Meta-Regression, Bias And Bias-Adjustment , 2011 .

[18]  Bryan S. Graham,et al.  Efficient Estimation of Data Combination Models by the Method of Auxiliary-to-Study Tilting (AST) , 2011 .

[19]  A B Haidich,et al.  Meta-analysis in medical research. , 2010, Hippokratia.

[20]  M. Kenward,et al.  Analysis of Incomplete Data Using Inverse Probability Weighting and Doubly Robust Estimators , 2010 .

[21]  David J Spiegelhalter,et al.  A re-evaluation of random-effects meta-analysis , 2009, Journal of the Royal Statistical Society. Series A,.

[22]  Els Goetghebeur,et al.  Estimation of controlled direct effects , 2008 .

[23]  J. Robins,et al.  Doubly Robust Estimation in Missing Data and Causal Inference Models , 2005, Biometrics.

[24]  M. Höfler,et al.  Causal inference based on counterfactuals , 2005, BMC medical research methodology.

[25]  S. Thompson,et al.  Quantifying heterogeneity in a meta‐analysis , 2002, Statistics in medicine.

[26]  P. F. Kauff.  Group , 2000, Elegant Design.

[27]  J. Pearl,et al.  Confounding and Collapsibility in Causal Inference , 1999 .

[28]  G. Smith,et al.  Meta-analysis: Potentials and promise , 1997, BMJ.

[29]  S. Pocock,et al.  INDANA: a meta-analysis on individual patient data in hypertension. Protocol and preliminary results. , 1995, Therapie.

[30]  ECiM 6BQ,et al.  Medical Research Council trial of treatment of hypertension in older adults: principal results. MRC Working Party. , 1992, BMJ.

[31]  B. Dahlöf,et al.  Morbidity and mortality in the Swedish Trial in Old Patients with Hypertension (STOP-Hypertension) , 1991, The Lancet.

[32]  T S Warrender,et al.  Randomised trial of treatment of hypertension in elderly patients in primary care. , 1986, British medical journal.

[33]  C. Bulpitt,et al.  Mortality and Morbidity Results from the European Working Party on High Blood Pressure in the Elderly Trial , 1985, The Lancet.

[34]  N. Naing.  Easy way to learn standardization : direct and indirect methods. , 2000, The Malaysian journal of medical sciences : MJMS.

[35]  C. Dollery,et al.  Medical Research Council trial of treatment of hypertension in older adults: principal results. , 1992 .

[36]  Prevention of stroke by antihypertensive drug treatment in older persons with isolated systolic hypertension. Final results of the Systolic Hypertension in the Elderly Program (SHEP). SHEP Cooperative Research Group. , 1991, JAMA.