Deficits and Remedy of the Standard Random Effects Methods in Meta-analysis

Abstract

The random effects model is often used in meta-analyses. A corresponding significance test based on a normal approximation has been established. Its type I error is derived in this article by theoretical considerations and computer simulations. The test can be conservative as well as unacceptably anti-conservative. The anti-conservatism increases with the increasing number of patients and the decreasing number of studies. A modification is proposed, which keeps the nominal level asymptotically as the number of patients approaches infinity. Simulations show that the modified test is often conservative, but its conservatism is small in those situations where the standard test is highly anti-conservative.

Keywords

Meta-analysis - Multi-center Study - Random Effects Model - Type I Error - Odds Ratio

• References

• 1 DerSimonian R, Laird N. Meta-analysis in clinical trials. Contr Clin Trial 1986; 7: 177-88.
• 2 Larholt K, Gelber RD. Heterogeneity in meta-analysis: A simulation study of fixed effect and random effect methods. Harvard School Publ Health, Dept Biostatistics; (unpublished), Boston: 1990
• 3 Böckenhoff A, Hartung J. Some corrections of the significance level in meta-analysis. Biometrical J 1998; 40: 937-47.
• 4 Larholt K, Tsiatis AA, Gelber RD. Variability of coverage probabilities when applying a random effects methodology for meta-analysis. Harvard School Publ Health, Dept Biostatistics; (unpublished), Boston: 1990
• 5 Follmann DA, Proschan MA. Valid inference in random effects meta-analysis. Bio-metrics 1999; 55: 732-7.
• 6 Hartung J. An alternative method for meta-analysis. Biometrical J 1999; 41: 901-16.
• 7 Petitti DB. Meta-analysis, decision analysis, and cost effectiveness analysis: methods for quantitative synthesis in medicine. Oxford: Oxford University Press; 1994
• 8 Whitehead A, Whitehead J. A general parametric approach to the meta-analysis of randomized clinical trials. Stat Med 1991; 10: 1665-77.
• 9 Lau J, Ioanides JPA, Schmid CH. Summing up evidence: one answer is not always enough. Lancet 1998; 351: 123-7.
• 10 Biggerstaff BJ, Tweedie RL. Incorporating variability in estimates of heterogeneity in the random effects model in meta-analysis. Stat Med 1997; 16: 753-68.
• 11 Koch A, Bouges S, Ziegler S, Dinkel H, Daures JP, Victor N. Low molecular weight heparin and unfractionated heparin in thrombosis prophylaxis after major surgical intervention: update of previous meta-analyses. Brit J Surgery 1997; 84: 750-9.
• 12 Ziegler S. Meta-Analyse im Modell mit festen und zufälligen Effekten. Dissertation University Heidelberg; 1999
• 13 Ziegler S, Victor N. Gefahren der Standard-methoden für Meta-Analysen bei Vorliegen von Heterogenität. Inform Biometrie Epidemiol Med Biol 1999; 30: 131-40.
• 14 Berlin JA, Laird NM, Sacks HS. et al. A comparison of statistical methods for combining event rates from clinical trials. Stat Med 1989; 8: 141-51.
• 15 Hardy RJ, Thompson SG. A likelihood approach to meta-analysis with random effects. Stat Med 1996; 15: 619-29.
• 16 Mosteller F, Chalmers TC. Some progress and problems in meta-analysis of clinical trials. Stat Sci 1992; 7: 227-36.
• 17 Victor N. The challenge of meta-analysis: Indications and contra-indications for meta-analysis. J Clin Epidemiol 1995; 48: 5-8.
• 18 Imhof JP. Computing the distribution of quadratic forms in normal variables. Biometrika 1961; 48: 419-26.