Including older people in clinical researchBMJ 2005; 331 doi: https://doi.org/10.1136/bmj.331.7524.1036 (Published 03 November 2005) Cite this as: BMJ 2005;331:1036
- Marion E T McMurdo (), head of ageing and health,
- Miles D Witham, clinical lecturer,
- Neil D Gillespie, senior clinical lecturer
- Ninewells Hospital and Medical School, Dundee DD1 9SY
- Ninewells Hospital and Medical School, Dundee DD1 9SY
Benefits shown in trials in younger people may not apply to older people
There are more old people alive today than at any time in history. Older people are, quite rightly, “the core business of the NHS.”1 The need to be able to draw on the results of good quality research to inform best practice in the specific management of older people is compelling. So we might expect that researchers would have eagerly embraced the participation of older people in clinical trials. Yet this is not the case. What do clinicians and researchers have to do to redress the serious bias against older people in clinical research?
The evidence that older people are being excluded from clinical research is widespread. Although the world is facing a global increase in the prevalence of diabetes mellitus, and by 2030 it is estimated that over 48 million older people in developed countries will have diabetes, the mean age of participants in the United Kingdom Progression of Diabetes Study was only 53 years. Guidelines on older people with diabetes have acknowledged the dearth of evidence, basing recommendations on expert opinion and extrapolation from research in younger populations.2 A systematic review of cancer clinical trials reported that patients aged over 65 were under-represented, with only a quarter to a third of potentially eligible older patients enrolled.3 Research on cardiovascular disease tells a similarly depressing story as 40% of the 593 randomised trials of coronary artery disease published between 1991 and 2000 had explicit exclusions on the basis of age.4 Analysis of a national cohort of 20 388 older Medicare beneficiaries with heart failure showed that only 18%, 13%, and 25% met the enrollment criteria for the SOLVD, MERIT-HF, and RALES trials respectively.5
In 1997 it was reported in the BMJ that one third of research papers published in four major medical journals excluded older people without justification.6 We repeated this analysis for papers published in the same journals in 2004, and found that although matters have improved a little, almost 15% of papers still unjustifiably excluded older people, and that fewer than 5% of the papers published were specific to older people (see table on bmj.com).
The under-representation of older people in clinical research is clear, but what might the explanation be? The first possibility is that investigators seeking clean science and low drop out rates are deterred by the prospect of enrolling participants with comorbidities and multiple medicines, which are common features of late old age. Such individuals clearly have many risk factors for drug interactions, side effects, death, and admission, all of which may fail to flatter trial findings. Age, protocol restrictions on comorbidities, and physicians' perceptions have been identified as important contributors to the exclusion of older patients from cancer trials.3 The narrow eligibility and exclusion criteria used in many clinical trials achieve a homogeneity of study population, and so reduce unexpected variations in drug efficacy and adverse events.7 Clearly a tension exists between the validity and the generalisability of trial findings, but this practice leaves a yawning chasm between patients in the real world and patients who participate in clinical studies. This greatly disadvantages older people, to whom the evidence does not apply.
The second possibility is that investigators view the older population as vulnerable and so in need of protection from researchers. This, of course, undermines autonomy. Sadly, the growing regulatory burden on researchers has provided a further disincentive to research in under-studied groups.8 Yet, when approached, older people report a high rate of willingness to consider participation in clinical trials.9
The final possibility is that investigators are daunted by the hard work involved in enrolling and maintaining older, frail people in clinical trials. Screening may be a prolonged process, and many older patients wish to discuss the trial with relatives, carers, and other health professionals before enrolling. Transport and mobility problems present another set of barriers to participation, especially among poorer, socially isolated patients—who ironically are most likely to suffer from ill health. Obtaining informed consent can be more challenging, as cognitive, hearing, and visual impairments may need to be surmounted.
It cannot be assumed that the benefits to younger patients enrolled in trials can be extrapolated to frail older people.10 Several strategies could potentially improve the current situation. Firstly, regulatory authorities should ensure that the population likely to be prescribed a particular drug once licensed actually participates in its evaluation.11 Secondly, those funding and commissioning research should favour proposals in typical populations which incorporate end points that are meaningful and relevant to older people. Both ethics committees and funders should recognise as unethical applications proposing to apply arbitrary and unjustified age restrictions.12 Funders could ensure that grant funding includes resources to address issues like access, transport, social networks, and the management of treatment-related events. Thirdly, protocols should be specifically designed to stratify by age or use modified exclusion criteria to allow the inclusion of patients with comorbid disease and multiple medication.3
Promotion of the benefits of research in older people is required by engaging with the public as a whole, older people, funding bodies, researchers, and clinicians. The greatest burden of disease in the developed world falls on older people. It is time that research activity reflected this.
Competing interests None declared.
Table appears on bmj.com