Simpson's paradox visualized: The example of the Rosiglitazone meta-analysis

BMC Medical Research Methodology

Volumn 8, Issue , 2008, Pages

  Rücker, Gerta ^a   Schumacher, Martin ^a  

^a UNIVERSITY OF FREIBURG   (Germany)

Author keywords

[No Author keywords available]

Indexed keywords

ROSIGLITAZONE; 2,4 THIAZOLIDINEDIONE DERIVATIVE;

ARTICLE; CLINICAL RESEARCH; CONTROLLED STUDY; HEART INFARCTION; HYPOTHESIS; PLOTS AND CURVES; SIMPSON PARADOX; STATISTICAL PARAMETERS; CHEMICALLY INDUCED DISORDER; CHILD; HUMAN; LEUKEMIA; META ANALYSIS; POWER SUPPLY; RANDOMIZED CONTROLLED TRIAL; SAMPLE SIZE; STATISTICAL ANALYSIS; STATISTICS;

CHILD; DATA INTERPRETATION, STATISTICAL; ELECTRIC POWER SUPPLIES; HUMANS; LEUKEMIA; META-ANALYSIS AS TOPIC; MYOCARDIAL INFARCTION; RANDOMIZED CONTROLLED TRIALS AS TOPIC; SAMPLE SIZE; STATISTICS AS TOPIC; THIAZOLIDINEDIONES;

EID: 46749104060     PISSN: None     EISSN: 14712288     Source Type: Journal    
DOI: 10.1186/1471-2288-8-34     Document Type: Article

References (28)

1. Notes on the theory of association of attributes of statistics. G Yule, Biometrika 1903 2 121 134 10.1093/biomet/2.2.121
2. The interpretation of interaction in contingency tables. E Simpson, Journal of the Royal Statistical Society 1951 13 238 241
3. Good for women, good for men, bad for people: Simpson's paradox and the importance of sex-specific analysis in observational studies. SG Baker BS Kramer, Journal of Women's Health and Gender-Based Medicine 2001 10 9 867 872 10.1089/152460901753285769 11747681
4. Ecological bias, confounding, and effect modification. S Greenland H Morgenstern, Int J Epidemiol 1989 18 269 274 10.1093/ije/18.1.269 2656561
5. Confounding and Simpson's paradox. SA Julious MA Mullee, BMJ 1994 309 1480 1481 7804052
6. Ignoring a covariate: An example of Simpson's paradox. D Appleton J French M Vanderpump, The American Statistician 1996 50 4 340 341 10.2307/2684931
7. Simpson's paradox: An example from hospital epidemiology. R Reintjes A de Boer W van Pelt JM de Groot, Epidemiology 2000 11 81 83 10.1097/00001648- 200001000-00017 10615849
8. Simpson's paradox in meta-analysis. JA Hanley G Theriault, Epidemiology 2000 11 5 613 10.1097/00001648-200009000-00022 10955419
9. A cautionary tale. A Edwards, Significance 2007 2 47 48 10.1111/j.1740-9713.2007.00223.x
10. Meta-analysis, Simpson's paradox, and the number needed to treat. DG Altman JJ Deeks, BMC Medical Research Methodology 2002 2 3 10.1186/1471-2288-2-3 11860606
11. Simpson's paradox and calculation of number needed to treat from meta-analysis. CJ Cates, BMC Medical Research Methodology 2002 2 1 10.1186/1471-2288-2-1 11860604
12. Pooling, meta-analysis, and the evaluation of drug safety. M Lievre M Cucherat A Leizorovicz, Current Controlled Trials in Cardiovascular Medicine 2002 3 6 10.1186/1468-6708-3-6 11991807
13. The transitive fallacy for randomized trials: if A bests B and B bests C in separate trials, is A better than C? SG Baker BS Kramer, BMC Medical Research Methodolology 2002 2 13 10.1186/1471-2288-2-13
14. Randomized trials, generalizability, and meta-analysis: graphical insights for binary outcomes. SG Baker BS Kramer, BMC Medical Research Methodology 2003 3 10 10.1186/1471-2288-3-10 12809566
15. Chances of Simpson's paradox. J Jeon H Chun J Bae, Journal of the Korean Statistical Society 1987 16 117 125
16. Effect of Rosiglitazone on the risk of myocardial infarction and death from cardiovascular diseases. SE Nissen K Wolski, NEJM 2007 356 24 2457 2471 10.1056/NEJMoa072761 17517853
17. Beta blockade during and after myocardial infarction: An overview of the randomized trials. S Yusuf R Peto J Lewis R Collins P Sleight, Progress in Cardiovascular Diseases 1985 27 335 371 10.1016/S0033-0620(85)80003-7 2858114
18. Uncertain Effects of Rosiglitazone on the Risk for Myocardial Infarction and Cardiovascular Death. GA Diamond L Bax S Kaul, Annals of Internal Medicine 2007 147 8 578 581 17679700
19. Fixed vs random effects meta-analysis in rare event studies: The Rosiglitazone link with myocardial infarction and cardiac death. J Shuster L Jones D Salmon, Statistics in Medicine 2007 26 4375 4385 10.1002/sim.3060 17768699
20. Letter to the Editor. JR Carpenter G Rücker G Schwarzer, Statistics in Medicine 2007 [DOI: 10.1002/sim.3173]. 18069734
21. Much ado about nothing: a comparison of the performance of meta-analytical methods with rare events. MJ Bradburn JJ Deeks JA Berlin AR Localio, Statistics in Medicine 2007 26 53 77 10.1002/sim.2528 16596572
22. A Whitehead, Meta-analysis of controlled clinical trials Wiley 2002
23. Rosiglitazone and cardiovascular risk. M Bracken, N Engl J Med 2007 357 9 937 938 [Author reply 939-940]. 10.1056/NEJMc071602 17761600
24. Quantifying heterogeneity in a meta-analysis. JPT Higgins SG Thompson, Statistics in Medicine 2002 21 1539 1558 10.1002/sim.1186 12111919
25. Arcsine test for publication bias in meta-analyses with binary outcomes. G Rücker G Schwarzer JR Carpenter, Statistics in Medicine 2008 27 5 746 763 10.1002/sim.2971 17592831
26. Two Statistical Paradoxes in the Interpretation of Group Differences: Illustrated with Medical School Admission and Licensing Data. H Wainer LM Brown, The American Statistician 2004 58 2 117 123
27. Empirical evaluation shows the Copas selection model provides a useful summary in 80% of meta-analyses. JR Carpenter G Schwarzer G Rücker R Künstler, 2007
28. R Development Core Team, R: A Language and Environment for Statistical Computing R Foundation for Statistical Computing, Vienna, Austria 2006 http://www.R-project.org [ISBN 3-900051-07-0].