Biostatistics and epidemiology
Applying trial evidence to older people: generalisability and subgroup analyses
Why trial results often don't transfer to frail, multimorbid older patients, how baseline risk changes NNT and NNH, time to benefit, NHMRC levels of evidence, and how to judge a subgroup claim, using ASPREE, HYVET and SPRINT.
Tier 2Medical sciences concept
Written with AI assistance and fact-checked against the sources below. For exam study, not clinical decisions. How notes are made
At a glance
- Most trials under-represent people over 75, and almost entirely exclude those with frailty, dementia, multimorbidity, polypharmacy or residential care. These are the patients geriatricians treat.
- Applying evidence to an older patient means asking four questions:
- Is the relative effect likely to be similar?
- What is this patient's baseline risk, and so the absolute benefit?
Loading the rest of the note…
Finished reading?
Sources
- NHMRC additional levels of evidence and grades for recommendations for developers of guidelinesAU/NZNational Health and Medical Research Council, 2009
- Effect of aspirin on disability-free survival in the healthy elderly (ASPREE)New England Journal of Medicine (McNeil et al., 2018;379:1499-508)
- Effect of aspirin on all-cause mortality in the healthy elderly (ASPREE)New England Journal of Medicine (McNeil et al., 2018;379:1519-28)
- Treatment of hypertension in patients 80 years of age or older (HYVET)New England Journal of Medicine (Beckett et al., 2008;358:1887-98)
- Intensive vs standard blood pressure control in adults 75 years or older (SPRINT subgroup)JAMA (Williamson et al., 2016;315:2673-82)
- Development of the Instrument to assess the Credibility of Effect Modification Analyses (ICEMAN)CMAJ (Schandelmaier et al., 2020;192:E901-6)
- CONSORT Harms 2022 statement, explanation, and elaborationBMJ (Junqueira et al., 2023)
- CONSORT 2025 statement: updated guideline for reporting randomised trialsBMJ (Hopewell et al., 2025)
- Is a subgroup effect believable? Updating criteria to evaluate the credibility of subgroup analysesBMJ (Sun et al., 2010;340:c117)
- Last updated
- 14 November 2026
What changed
- 14 November 2026 Turnstile bot check on sign-in and signed-out feedback