Elsevier

The Lancet

Volume 365, Issue 9454, 8–14 January 2005, Pages 176-186
The Lancet

Series
Subgroup analysis in randomised controlled trials: importance, indications, and interpretation

https://doi.org/10.1016/S0140-6736(05)17709-5Get rights and content

Summary

Large pragmatic trials provide the most reliable data about the effects of treatments, but should be designed, analysed, and reported to enable the most effective use of treatments in routine practice. Subgroup analyses are important if there are potentially large differences between groups in the risk of a poor outcome with or without treatment, if there is potential heterogeneity of treatment effect in relation to pathophysiology, if there are practical questions about when to treat, or if there are doubts about benefit in specific groups, such as elderly people, which are leading to potentially inappropriate undertreatment. Analyses must be predefined, carefully justified, and limited to a few clinically important questions, and post-hoc observations should be treated with scepticism irrespective of their statistical significance. If important subgroup effects are anticipated, trials should either be powered to detect them reliably or pooled analyses of several trials should be undertaken. Formal rules for the planning, analysis, and reporting of subgroup analyses are proposed.

Introduction

“The essence of tragedy has been described as the destructive collision of two sets of protagonists, both of whom are correct. The statisticians are right in denouncing subgroups that are formed post hoc from exercises in pure data dredging. The clinicians are also right, however, in insisting that a subgroup is respectable and worthwhile when established a priori from pathophysiological principles.”

A R Feinstein, 19981

Randomised controlled trials (RCTs) and systematic reviews are the most reliable methods of determining the effects of treatments.2, 3, 4, 5 However, when trials were first developed for use in agriculture, researchers were presumably concerned about the effect of interventions on the overall size and quality of the crop rather than on the wellbeing of any individual plant. Clinicians have to make decisions about individuals, and how best to use results of RCTs and systematic reviews to do this has generated considerable debate.6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 Unfortunately, this debate has polarised, with statisticians and predominantly non-clinical (or non-practising) epidemiologists warning of the dangers of subgroup analysis and other attempts to target treatment, and clinicians warning of the dangers of applying the overall results of large trials to individual patients without consideration of pathophysiology or other determinants of individual response. This rift, described by Feinstein as a “clinicostatistical tragedy”,1 has been widened by some of the more enthusiastic proclamations on the extent to which the overall results of trials can properly inform decisions at the bedside or in the clinic.23, 24, 25

The results of small explanatory trials with well-defined eligibility criteria should be easy to apply, but generalisability is often undermined by highly selective recruitment, resulting in trial populations that are unrepresentative even of the few patients in routine practice who fit the eligibility criteria.26 Recruitment of a higher proportion of eligible patients is a major strength of large pragmatic trials, but deliberately broad and sometimes ill-defined entry criteria mean that the overall result can be difficult to apply to particular groups,27 and that subgroup analyses are necessary if heterogeneity of treatment effect is likely to occur. Yet despite the adverse effects on patient care that can result from misinterpreted or inappropriate subgroup analyses (table 1), there are no reviews or guidelines on the clinical indications for subgroup analysis and no consensus on the implications for trial design, analysis, and interpretation of subgroup effects, and the CONSORT statement on reporting of trials includes only a few lines on subgroup analysis.28 This article discusses arguments for and against subgroup analyses, the clinical situations in which they can be useful, and rules for their performance and interpretation. Illustrative examples are taken mainly from treatments for cerebrovascular or cardiovascular disease but the principles are relevant to all areas of medicine and surgery.

Section snippets

Arguments against subgroup analysis

“… it would be unfortunate if desire for the perfect (ie, knowledge of exactly who will benefit from treatment) were to become the enemy of the possible (ie, knowledge of the direction and approximate size of the effects of treatment of wide categories of patient).”

S Yusuf et al, 19844

The main argument against subgroup analysis is that qualitative heterogeneity of relative treatment effect (defined as the treatment effect being in different directions in different groups of patients, ie,

Situations in which subgroup analyses should be considered

“The tragedy of excluding cogent pathophysiologic subgroup analyses merely because they happen to be subgroups will occur if statisticians do not know the distinction, and if clinicians who do know it remain mute, inarticulate or intimidated.”

A R Feinstein, 19981

Subgroup analyses should be predefined and carefully justified. Feinstein and others have emphasised the need for determination of pathophysiological heterogeneity, but there are three other indications for subgroup analysis (

Estimation and interpretation of subgroup effects

“Far better an approximate answer to the right question, which is often vague, than an exact answer to the wrong question, which can always be made precise.”

J W Tukey, 1962118

Conclusions

Large randomised controlled trials with broad eligibility criteria and high inclusion rates provide the most reliable data about the effects of treatments, but these should be designed, analysed, and reported in a way that allows clinicians to use the results as effectively as possible in routine practice. Subgroup analyses can be useful if there are widely differing risks of a poor outcome with or without treatment between specific groups, if there are important differences in pathophysiology

References (128)

  • MD Flather et al.

    Long-term ACE-inhibitor therapy in patients with heart failure or left-ventricular dysfunction: a systematic overview of data from individual patients

    Lancet

    (2000)
  • S Yusuf et al.

    β blockade during and after acute myocardial infarction: an overview of the randomized trials

    Prog Cardiovasc Dis

    (1985)
  • DW Taylor et al.

    Low-dose and high-dose acetylsalicylic acid for patients undergoing carotid endarterectomy: a randomised controlled trial. ASA and Carotid Endarterectomy (ACE) Trial Collaborators

    Lancet

    (1999)
  • HC Wijeysundera et al.

    Neurohormones and oxidative stress in nonischemic cardiomyopathy: relationship to survival and the effect of treatment with amlodipine

    Am Heart J

    (2003)
  • SF Assmann et al.

    Subgroup analysis and other (mis)uses of baseline data in clinical trials

    Lancet

    (2000)
  • RA Stallones

    The use and abuse of subgroup analysis in epidemiological research

    Prev Med

    (1987)
  • PM Rothwell et al.

    Prediction of benefit from carotid endarterectomy in individual patients: a risk modelling study

    Lancet

    (1999)
  • PM Rothwell et al.

    Effect of endarterectomy for recently symptomatic carotid stenosis in relation to clinical subgroups and the timing of surgery

    Lancet

    (2004)
  • S Yusuf et al.

    Effect of coronary artery bypass graft surgery on survival: overview of 10-year results from randomised trials by the Coronary Artery Bypass Graft Surgery Trialists' Collaboration

    Lancet

    (1994)
  • F Neumann et al.

    Treatment of Chlamydia pneumoniae infection with roxithromycin and effect on neointima proliferation after coronary stent placement (ISAR-3): a randomised, double-blind, placebo-controlled trial

    Lancet

    (2001)
  • PL Barratt et al.

    DNA markers predicting benefit from adjuvant flourouracil in patients with colon cancer: a molecular study

    Lancet

    (2002)
  • R Peto

    Statistics of cancer trials

  • R Collins et al.

    Large-scale randomised evidence: trials and overviews

  • S Yusef et al.

    Why do we need some large, simple randomized trials?

    Stat Med

    (1984)
  • G Davey Smith et al.

    Beyond the grand mean

    BMJ

    (1997)
  • RE Wittes

    Problems in the medical interpretation of overviews

    Stat Med

    (1987)
  • IonnidisJPA et al.

    Uncontrolled pearls, controlled evidence, meta-analysis and the individual patient

    J Clin Epidemiol

    (1998)
  • I Chalmers

    A patient's attitude to the use of research evidence for guiding individual choices and decisions in healthcare

    Clin Risk

    (2000)
  • S Senn

    Applying results of randomised trials to patients. N of 1 trials are needed

    BMJ

    (1998)
  • AD Oxman et al.

    A consumer's guide to subgroup analyses

    Ann Intern Med

    (1992)
  • D Black

    The limitations of evidence

    J R Coll Physicians Lond

    (1998)
  • JR Hampton

    Size isn't everything

    Stat Med

    (2002)
  • LR Caplan

    Evidence based medicine: concerns of a clinical neurologist

    J Neurol Neurosurg Psychiatry

    (2001)
  • JG Evans

    Evidence-based and evidence-biased medicine

    Age Ageing

    (1995)
  • JD Swales

    Evidence-based medicine and hypertension

    J Hypertens

    (1999)
  • T Fahey

    Applying the results of clinical trials to patients in general practice: perceived problems, strengths, assumptions, and challenges for the future

    Br J Gen Pract

    (1998)
  • DL Sackett et al.

    Finding and applying evidence during clinical rounds: the “evidence cart”

    JAMA

    (1998)
  • DL Sackett

    Applying overviews and meta-analyses at the bedside

    J Clin Epidemiol

    (1995)
  • J Ellis et al.

    Inpatient general medicine is evidence based

    Lancet

    (1995)
  • DG Altman et al.

    The revised CONSORT statement for reporting randomised trials: explanation and elaboration

    Ann Intern Med

    (2001)
  • A randomised trial of aspirin and sulfinpyrazone in threatened stroke

    N Engl J Med

    (1978)
  • WS Fields et al.

    Controlled trial of aspirin in cerebral ischaemia

    Stroke

    (1977)
  • Collaborative overview of randomised trials of antiplatelet therapy. I: Prevention of death, myocardial infarction, and stroke by prolonged antiplatelet therapy in various categories of patients

    BMJ

    (1994)
  • K Anastos et al.

    Hypertension in women: what is really known?

    Ann Intern Med

    (1991)
  • MRC trial of treatment of mild hypertension: principal results

    BMJ

    (1985)
  • F Gueyffier et al.

    Effect of antihypertensive drug treatment on cardiovascular outcomes in men and women

    Ann Intern Med

    (1997)
  • A Amery et al.

    Influence of antihypertensive drug treatment on morbidity and mortality in patients over the age of 60 years. European Working Party on High blood pressure in the Elderly (EWPHE) results: sub-group analysis on entry stratification

    J Hypertens Suppl

    (1986)
  • ClelandJGF et al.

    Is aspirin safe for patients with heart failure?

    Br Heart J

    (1995)
  • MP Anderson et al.

    Effects of alprenolol on mortality among patients with definite or suspected acute myocardial infarction: preliminary results

    Lancet

    (1979)
  • Multicenter International Study: Supplemental report: reduction in mortality after myocardial infarction with long-term b-adrenoreceptor blockade

    BMJ

    (1977)
  • Cited by (0)

    View full text