Abstract

▪ The extent to which a clinician should believe and act on the results of subgroup analyses of data from randomized trials or meta-analyses is controversial. Guidelines are provided in this paper for making these decisions. The strength of inference regarding a proposed difference in treatment effect among subgroups is dependent on the magnitude of the difference, the statistical significance of the difference, whether the hypothesis preceded or followed the analysis, whether the subgroup analysis was one of a small number of hypotheses tested, whether the difference was suggested by comparisons within or between studies, the consistency of the difference, and the existence of indirect evidence that supports the difference. Application of these guidelines will assist clinicians in making decisions regarding whether to base a treatment decision on overall results or on the results of a subgroup analysis.

References

  • 1. Pocock SHughes M, and Lee R. Statistical problems in the reporting of clinical trials. N Engl J Med. 1987;317:426-32. CrossrefMedlineGoogle Scholar
  • 2. Effectiveness of intravenous thrombolytic treatment in acute myocardial infarction. Gruppo Italiano per lo Studio della Streptochinasi Nell'Infarcto Miocardico (GISSI). Lancet. 1986;2:397-402. CrossrefMedlineGoogle Scholar
  • 3. Lee TWeisberg MBrand DRouan G, and Goldman L. Candidates for thrombolysis among emergency room patients with acute chest pain. Ann Intern Med. 1989;110:957-62. LinkGoogle Scholar
  • 4. Tate D and Dehmer G. New challenges for thrombolytic therapy. Ann Intern Med. 1989;110:953-5. LinkGoogle Scholar
  • 5. Randomised trial of intravenous streptokinase, oral aspirin, both, or neither among 17,187 cases of suspected acute myocardial infarction: ISIS-2. Lancet. 1988;2:349-60. MedlineGoogle Scholar
  • 6. Yusuf SCollins R, and Peto R. Why do we need some large, simple randomized trials? Stat Med. 1984;3:409-20. CrossrefMedlineGoogle Scholar
  • 7. . Treatment of Early Breast Cancer. v 1. Worldwide evidence 1985-1990. New York: Oxford University Press; 1990:7-9. Google Scholar
  • 8. Sackett DHaynes R, and Tugwell P. Clinical Epidemiology: A Basic Science for Clinical Medicine. Boston: Little, Brown & Company; 1985. Google Scholar
  • 9. Chalmers TSmith HBlackburn BSilverman BSchoeder B, and Reitman D. A method for assessing the quality of a randomized control trial. Controlled Clin Trials. 1981;2:31-49. CrossrefMedlineGoogle Scholar
  • 10. Fletcher RFletcher S, and Wagner E. Clinical Epidemiology. 2d ed. Baltimore: Williams & Wilkins, 1988:185-6. Google Scholar
  • 11. Feinstein A. Clinical Epidemiology. Philadelphia: W.B. Saunders; 1985:306-7, 516-7. Google Scholar
  • 12. Goldman L and Feinstein A. Anticoagulants and myocardial infarction. The problems of pooling, drowning, and floating. Ann Intern Med. 1979;90:92-4. LinkGoogle Scholar
  • 13. Furberg C and Morgan T. Lessons from overviews of cardiovascular trials. Stat Med. 1987;6:295-306. CrossrefMedlineGoogle Scholar
  • 14. Horwitz R. Complexity and contradiction in clinical trial research. Am J Med. 1987;82:498-510. CrossrefMedlineGoogle Scholar
  • 15. Buyse M. Analysis of clinical trial outcomes: some comments on subgroup analyses. Controlled Clin Trials. 1989;10:187S-194S. CrossrefMedlineGoogle Scholar
  • 16. Bulpitt C. Subgroup analysis. Lancet. 1988;2:31-4. CrossrefMedlineGoogle Scholar
  • 17. Byar D. Assessing apparent treatment—covariate interactions in randomized clinical trials. Stat Med. 1985;4:255-63. CrossrefMedlineGoogle Scholar
  • 18. Shuster J and van Eys J. Interaction between prognostic factors and treatment. Controlled Clin Trials. 1983;4:209-14. CrossrefMedlineGoogle Scholar
  • 19. Kleinbaum DKupper L, and Morgenstern H eds. Interaction, effect modification, and synergism. In: Epidemiologic Research. New York: Van Nostrand Reinhold Company, 1982;403-18. Google Scholar
  • 20. Last J. A Dictionary of Epidemiology. 2d ed. New York: Oxford University Press; 1988:41, 66, 127. Google Scholar
  • 21. Yusuf SWittes J, and Probstfield J. Evaluating effects of treatment in subgroups of patients within a clinical trial: the case of non-Q-wave myocardial infarction and beta blockers. Am J Cardiol. 1990;66:220-2. CrossrefMedlineGoogle Scholar
  • 22. Bailey K. Inter-study differences: how should they influence the interpretation and analysis of results? Stat Med. 1987;6:351-8. CrossrefMedlineGoogle Scholar
  • 23. Davis C and Leffingwell D. Empirical Bayes estimates of subgroup effects in clinical trials. Controlled Clin Trials. 1990;11:37-42. CrossrefMedlineGoogle Scholar
  • 24. Sackett D. Clinical diagnosis and the clinical laboratory. Clin Invest Med. 1978;1:37-43. MedlineGoogle Scholar
  • 25. Yusuf SPeto RLewis JCollins R, and Sleight T. Beta blockade during and after myocardial infarction: an overview of the randomized trials. Prog Cardiovas Dis. 1985;27:335-71. CrossrefMedlineGoogle Scholar
  • 26. Peto R. Why do we need systematic overviews of randomized trials? Stat Med. 1987;6:233-4. CrossrefMedlineGoogle Scholar
  • 27. Schneider B. Analysis of clinical trial outcomes: alternative approaches to subgroup analysis. Controlled Clin Trials. 1989;10:176S-186S. CrossrefMedlineGoogle Scholar
  • 28. Breslow N and Day N. Analysis of Case-Control Studies. v 1 of Statistical Methods of Cancer Research. Lyon: IARC Scientific Publications; 1980:122-59. Google Scholar
  • 29. Donner A. A Bayesian approach to the interpretation of subgroup results in clinical trials. J Chronic Dis. 1982;34:429-35. CrossrefGoogle Scholar
  • 30. Louis T. Estimating a population parameter values using Bayes and empirical Bayes methods. J Am Stat Assoc. 1984;79:393-8. CrossrefGoogle Scholar
  • 31. Beach M and Meier P. Choosing covariates in the analysis of clinical trials. Controlled Clin Trials. 1989;10:161S-175S. CrossrefMedlineGoogle Scholar
  • 32. A randomized trial of aspirin and sulfinpyrazone in threatened stroke. The Canadian Cooperative Study Group. N Engl J Med 1978;299:53-9. CrossrefMedlineGoogle Scholar
  • 33. Secondary prevention of vascular disease by prolonged antiplatelet treatment. Antiplatelet Trialists' Collaboration. Br Med J [Clin Res]. 1988;296:320-31. CrossrefMedlineGoogle Scholar
  • 34. Canner PHuang Y, and Meinert C. On the detection of outlier clinics in medical trials: I. Practical considerations. Controlled Clin Trials. 1981;2:231-40. CrossrefMedlineGoogle Scholar
  • 35. Lee DJohnson RBingham JLeahy MDinsmore R, and Goroll A. Heart failure in outpatients. N Engl J Med. 1982;306:699-705. CrossrefMedlineGoogle Scholar
  • 36. A randomized trial of propranolol in patients with acute myocardial infarction. I. Mortality results. Beta-blocker Heart Attack Trial Research Group. JAMA. 1982;247:1707-14. CrossrefMedlineGoogle Scholar
  • 37. Furberg C and Byington R. What do subgroup analyses reveal about differential response to beta-blocker therapy? Circulation. 1983;67:198-1101. MedlineGoogle Scholar
  • 38. Cook DWitt LCook R, and Guyatt G. Stress ulcer prophylaxis in the critically ill—a meta-analysis. Am J Med. [In press.] Google Scholar
  • 39. Guyatt GSullivan MFallen ETihal HRideout E, and Halcrow S. A controlled trial of digoxin in congestive heart failure. Am J Cardiol. 1988;61:371-5. CrossrefMedlineGoogle Scholar
  • 40. Jaeschke ROxman A, and Guyatt G. To what extent do congestive heart failure patients in sinus rhythm benefit from digoxin therapy? A systematic overview and meta-analysis. Am J Med. 1990;88:279-86. CrossrefMedlineGoogle Scholar
  • 41. Stieb DFrayha HOxman AShannon HHutchison B, and Crombie F. Effectiveness of Haemophilus influenzae type b vaccines. Can Med Assoc J. 1990;142:719-33. Google Scholar
  • 42. Kelton JHirsh JCarter C, and Buchanan M. Sex differences in the antithrombotic effect of aspirin. Blood. 1978;52;1073-6. CrossrefMedlineGoogle Scholar
  • 43. Hill A. Principles of Medical Statistics. 9th ed. London: Lancet; 1971:312-20. Google Scholar

This content is PDF only. To continue reading please click on the PDF icon.